My cousin (American citizen) is married to an Australian. He (the Australian) has actually lived here and legally worked here before. They moved to Australia for a bit and for some abstruse reason nobody understands, he now can no longer legally live and work here anymore. They've been sorting his paperwork trying to move back for like 3 years.
"I have heard that it may be difficult to travel after I have married an American. Is that true?
Yes. It may not seem logical, but you have to be very careful about travel if you marry an American or green card holder.
If you are in F, J, or tourist status, you are expected to have a home abroad and the intent to return there. When you apply for a visa stamp at a U.S. embassy or consulate, you have to prove ties to your home country. Marriage to a U.S. citizen makes proving these ties difficult, because the assumption is that you will want to immigrate to the United States. If you need a new visa stamp in order to return to the United States, the likelihood that your will visa will be denied is strong—even if you plan to continue as a full-time student."
I don't think that's it because he was like 40 years old at the time he moved over here. He worked for UPS and they had a child here. It was extremely apparent he wasn't a tourist or student. I don't know what kind of paperwork he had at the time.
Now, they did meet in Hawaii so maybe he had some kind of tourism visa when he was there.
The question is not "does DOGE have the power to fire these employees?", it's "can Elon and Trump find the people that can disable their badges and disable their computer, and will congress respond when they do?"
And all the things you want to be done to Make America Great Again will not be done because the regulations requiring that they have eighty-seven different kinds of permission will still exist but the people with the authority to sign the permission slips will be on an extended paid vacation.
You know you can just... do things without asking for permission right? Especially when you're immune to the law due to having a guaranteed pardon from the president.
You actually can't. Let's say you don't feel that you need a building permit? Fine, but the construction company still won't touch it. You decide to sell an untested drug? Will still have a serious problem getting pharmacists to carry it.
Also, what about when they come after you in 4.1 years after you actually didn't get that preemptive presidential pardon because Trump would rather play golf that day?
You're going to care about permission slips. Because "damn the permission slips, I know Donald and he'll for sure pardon me if some stupid bureaucrat tries to have me arrested" is going to be very much a minority opinion, and the things you are going to want to do or have done are going to require the help of people who are going to be more "show me your permission slip, so I know I won't be going to jail in three years when the Democrats retake power".
Queuing theory shows it's actually much worse than doubling the time, because a task doesn't just take the time it needs for someone to work on it, it also takes the time it needs for someone to be free to work on it. Basically once the amount of tasks slightly exceeds the number of bureaucrats, the expected time taken shoots off to infinity. To solve that, one needs excess bureaucrats.
> admissions 20 per hour, discharges 19 per hour, result misery.
This isn't really a result about queueing. If the amount of work to be done exceeds the amount of work you're able to do, then work is going to remain undone. How you organize the work will never make any difference. What you're losing the time to isn't queueing, it's incapacity.
Note that in your parent comment:
> once the amount of tasks slightly exceeds the number of bureaucrats, the expected time taken shoots off to infinity
this is a queueing result; what's necessary is that tasks outnumber bureaucrats, not that the time required to complete the tasks exceeds your available bureaucrat-hours.
The mathematical study of queuing definitely includes statements the time taken by tasks, without which it would be difficult to do any calculations about queue length. The fact that a queue will grow to infinity if the arrival rate exceeds the processing rate is an elementary result that you will find in most introductory queuing theory texts.
It's so elementary that actual elementary school students have no trouble giving a rigorous proof. You don't even need a queueing concept. If nothing ever queues, you get exactly the same result, although the identity of which work gets done may change.
The point we want people to grasp does involve queues, and is not intuitive: there is a discontinuity at arrival rate=processing rate, in queue length or waiting time. since waiting time is infinite at arrival rate>processing rate, it must tend to infinity as arrival rate tends to processing rate. This is important as it means that 100% utilisation is a bad goal. But it's a goal that is often aimed for in the name of efficiency.
The 'Micawber principle' is trivial as stated, but it allows people for whom talk of infinities and discontinuities would be offputting,to generate the correct intuition naturally. expenditure being too close to income is bad, (the original Micawber principle), so too is arrival rate approaching processing rate ("result misery").
> The point we want people to grasp does involve queues
Sure. But the point you keep talking about doesn't. There's no relation. Your annual expenditures don't change whether you pay for things before getting them *or afterward*.
> This is important as it means that 100% utilisation is a bad goal. But it's a goal that is often aimed for in the name of efficiency.
I don't think this is true anywhere except political rhetoric.
I think they are trying to say that even if the intake is less than the nominal capacity queues still grow non linearly as you approach 100% utilisation if the system has dependent events and randomness in process time.
A toy example:
A worker feeds product into a machine. The machine can't work unless fed product (dependent events). The machine can process 1 a minute and the human doing the loading averages loading 1 a minute but the exact time varies, sometimes 30 seconds, sometimes 90 (random process time)
When the worker is slow, the machine has nothing to do, so capacity is lost. The worker may feed the next products faster than average but the machine can't work faster to make up the lost time.
In one sense the machine and the worker really can process one product per minute. In another sense when you account for the lost capacity above they can't and the closer you get to trying to operate at 1/min the more the queue will build up.
The market will respond by not trying to approve as many drugs (because the wait isn't worth it). Which will prevent unifinte queues, but also the survival of thousands of patients.
This is it! People get sick, people go on holidays, a new programme is introduced which means a flood of new applicants to deal with -you need some slack in the system to handle it.
If you've calculated "on the volume of current work, we only need N admin staff in this office", then you're up the creek without a paddle when the number becomes "N - 1 or more", or the volume of work increases. Files *do* pile up on desks because the one person in the office dealing with this is out and there is no-one free to take over, due to the New Efficiency Drive which meant that when the experienced people retired, no new hires were taken on and trained in.
Yes, this is very good. I am merely married to a bureaucrat (but one who worked at OIRA, the beating heart of American bureaucracy!)
A few things I'd note:
- the first Trump admin already did the "2 for 1" approach to regulation https://federalnewsnetwork.com/agency-oversight/2018/01/trumps-regulation-czar-touts-success-of-two-for-one-executive-order/ -- this was painful at the time, and there was lots of cheating/use of technicalities to keep things from grinding to a halt. Last I heard people were floating "10 for 1" for this incoming administration. This is consistent with the hyperbolic nature of DOGE pronouncements, too--it's all likely to come down to earth, but not guaranteed to since there are a lot of insane tyros about to be put in power.
- decreasing regulatory page count is an ok metric I suppose, but in addition to clearing out unused cruft (that lottery reg was probably not doing much harm) and some justifiable wins about saloon doors, I would be very curious to see how much of this got done by aligning Idaho regs with external standards by reference. It's a lot shorter to say "use this trade association's safety code" than to write it all out! At past jobs I have spent time analyzing state legislative output and though I don't mean to knock Idaho, I think it's fair to say that small states' government is often surprisingly threadbare. Many are part-time, and it's totally normal for legislators to have to share staff https://ballotpedia.org/States_with_a_full-time_legislature. All this is to say that what Idaho's doing might be good or bad, but I suspect it's something others have thought of, and might well only be practical because of how small the state is.
- a surprising number of those legislatively mandated reports don't get written on time (or at all). It's something Congress throws into laws left and right, and agencies sometimes decide to accept the risk of their boss getting yelled at at an oversight hearing when there isn't enough time to get everything done.
- more work has been done to avoid unnecessary thrash than you might imagine--"economically significant" regulations receive higher levels of scrutiny, for instance
- it's not just nuisance lawsuits: advocacy groups have been doing their part to make rulemaking worse by encouraging the public to spam dockets with thousands of nearly-identical comments about hot button issues. bureaucrats are typically required to review and respond to every substantive comment in the course of the process. they can ignore the junk, but just sorting through it takes time--and of course a lot of it isn't "junk" per se but is prosaic and obvious, but still must be accounted for.
Regarding page count, I rewrote a procedural manual for my job not long ago. I *easily* could have made it 50% shorter... by using more jargon and less clear language, cutting out explanations and examples, and ignoring rare edge cases. I don't think that would made it better! But it would satisfy the demands of idiots that want to cut page counts in half.
Grinding things to a halt is the point, though. Our government has been mandated to do too many things -- more things than we can afford. Something will give. The only choice we have is in exactly what breaks and when it breaks. We have a choice now, but if we wait much longer, we won't have a choice. What cannot go on, will not go on.
I object to saying we can't afford it - it's not like the EPA or the FDA have massive slices of the budget. The FDA's budget is $7.2 billion - which isn't nothing, but it's about 0.1% of the budget. We absolutely could scrape up another $7 billion if we decided we wanted the FDA to get stuff done twice as fast, we just don't have the political will because the Republicans are obsessed with "starving the beast."
Heck, Trump's tax plan is expected to add 3 *trillion* dollars to the deficit over 10 years - how is it possible that we can afford that, but we can't afford a few billion a year to make the bureaucracy function smoothly?
The "twice as fast" is tricky for the FDA, though. They passed the Covid vaccines fast, now some people are arguing those same vaccines are killing healthy young people by inducing cardiac complications and that these dangerous vaccines were rushed through too fast without proper checking. There is some evidence of the mRNA vaccines being implicated:
We need to spend less. That means having fewer laws and less enforcement. That means fewer people, but it also means less harm to the people. It's not realistic to think that everything the government does is helpful. Plenty of examples of special-interest legislation.
Special interests legislation simply means "two factions compromised." There isn't any other kind of legislation. Regulations exist because they serve the interests of one or more factions of the public. Cutting regulation simply means choosing which factions get screwed.
"It gives the example of repealing a regulation about what kind of doors pharmacies can have."
I just get the feeling that this was downstream of the Americans With Disabilities Act, and that pharmacies were required to have wheelchair accessible doors, which meant that there had to be a new regulation put in place about "your door must be of this type". I also wonder if there was a lawsuit or some sort of activist campaign, because generally that is what causes such specific regulations to be passed.
A lot of bureaucratic inertia is caused by badly written regulations which are left up to interpretation *plus* ambulance chasing law firms being all too willing to go "Did the local council refuse to rebuild your entire house because you are entitled to have the door frames expanded by 0.5 cm? Call us for legal advice now, you may be entitled to compensation!" advertising.
This means that bureaucrats are *very* hesitant to make decisions which are not backed up in triplicate by CYA, as if you make a decision which grants something, this will cost money and moreover set a precedent for others to claim the same, which means the Department is unhappy with you. But if you decide to refuse the thing, there is the possibility of a lawsuit which may well go against you, create a precedent for others to claim the same, and cost even more what with damages and fees, which means the Department is unhappy with you.
I have seen people absolutely trying to game the system re: the width of doors being too small to let a wheelchair through, so that's why I have my suspicions about that particular piece of legislation.
Oh, the ADA is definitely gamed. You take an unscrupulous attorney, and a cooperative disabled person, and go looking for any business with any ADA violation. The disabled person attempts to enter and claims that their rights were denied by the business. Then the attorney sues them and immediately offers to settle for thousands of dollars.
It could be any business even if the disabled person would otherwise have no reason to enter.
Never going to be any lack of unscrupulous lawyers, either. If Vivek turned his sights on the legal profession, he'd probably have near-unanimous public support, but he went to law school himself, didn't he? 😀
Also associated with not transgressing social limits, which is why he put aside his wife Sita for rumours among the people as to her being unchaste when captured by Ravanna.
One time in history when a king really should have gone "Stuff you, I'm the king, I am above the law, and if you don't like it then lump it, peasants". Ah, well!
Too lazy to hunt the citation, but the Niskanen Center had a piece where a fellow pointed out that the raw number of federal employees had not grown much in the past 60 years. Growth has come from contractors bureaucrats hire, not bureaucrats per se.
So if anything we have been experimenting with the Vivek’s proposal already. Grow the scope and scale of the federal government without growing its headcount or capacity to deal with congress’ shirking of legislative responsibility to clarify statutes, and see what happens. Seems to be bad things to me!
Anyone with experience/knowledge of transit + infrastructure knows that it’s a fundamental mistake to hire consultants instead of just growing the bureaucracy to match needed expertise…
Instead of just hiring a core group of competent engineers, the NYC transit agency just outsources to consultants, who then hire contractors. No one is incentivized to care about efficiency or cost. Every party to the infrastructure work *but transit users* benefits from wasteful spending. So New York builds for 5-10x the cost of the Swedes
"Anyone with experience/knowledge of transit + infrastructure knows that it’s a fundamental mistake to hire consultants instead of just growing the bureaucracy..."
Not true if you are the owner of the outside consultant group being hired. \S
"No one is incentivized to care about efficiency or cost"
It's worse than that -- people are incentivized to inflate costs. I've worked on government funded research projects and from a manager's perspective it is worse to underspend the grant than to fail to meet the original goals. You're punished for being more efficient than you originally planned for.
Though I partially agree with you, I'll try to give the other side of the argument. I manage some government contracts, and when my management tells me to make sure all the money gets spent, part of what they really mean is "money is very tight, don't ask for more than you're going to use, because if you don't need it we can use it somewhere else."
Underspending is frowned upon because budget people at every level share this concern. Inflating the cost at the last minute to make sure you spend everything is a particularly underhanded and wasteful solution to keep from getting dressed down by the boss; the preferred solution is to make a good estimate before the project starts. Unfortunately it's very hard to carve out flexibility for something like a research project where costs are inherently unpredictable. The system demands a budget in advance, and demands you follow it.
That's not really my experience. There is concern that it "looks bad" if you don't spend everything because it means you can't properly estimate your work and costs, and that will reflect poorly when your next proposal is being reviewed. Or, even worse, for ongoing stuff your budget might get cut in the future.
I would agree the grant sin my world are all "spend all the money or you are a bad performer", they NEVER want money back. There is some minor pressure for cost reasonableness, but only on the front end. On the back end they just want all the money gone and to know that you vague accomplished what the plan was.
Yep, clawback. That's why at the end of the financial year, if there's anything left in the budget, you are encouraged to spend! spend! spend! since not alone will the excess be taken back, but next year your budget will be reduced.
And if next year you have extra unforeseen expenditure? Too bad, you managed to be too efficient for your own good. So better to buy stuff you don't need now (because you're not allowed to keep the saved amount on hand to cover anything else or for next year) rather than lose the money going forward.
It's one of those "looks good on paper" ideas that works out terribly in practice.
Obviously YMMV depending on the domain, but sometimes it’s not even that the contractor is particularly bad or inefficient. It’s that the government ends up spending almost as much effort “overseeing” the contractor (and the entire contract bid and award process) as they would if they just did the work themselves.
The bit I took from Edward Snowden's book about NSA contractors probably applies to bureaucrats in general.
There's rules on how many employees NSA can have and how much they can pay them, because these are federal employees. So what do you do when there's more work than employees, and you also want to keep your best people but can't pay them market rate?
After a stint as a proper NSA-employed agent, you let them switch to becoming contractors and hire them back at a higher rate. That's why Snowden was officially working for Booz Allen Hamilton, in a NSA facility with the same job and security clearance as before.
I guess they hire consultants when some regulation stops them from growing the bureaucracy or paying the internal experts as much as they'd like?
Yes, this is a very underrated factor: I'd be more keen to stop them hiring management consultants and contractors at vastly inflated wages than to arbitrarily cut headcount (especially if you're not cutting the actual legal risk or bureaucratic requirements first).
That's the problem with the top-down approach by politicians and government ministers versus the on-the-ground view of the people tasked with implementing the shiny new policies.
Not to mention that some of those aforesaid politicians and ministers have nice cosy relationships with the outside consultants, be it "they donate to my election campaign" or "my brother-in-law's cousin runs this".
"outsources to consultants, who then hire contractors. No one is incentivized to care about efficiency or cost."
You should see the saga of the Irish new children's hospital. This has been going on for decades at this point, and it's *still* not ready yet, and the fancy contractors who won the contract are still "yeah it's gonna be late and over-budget even more":
"The Chief Officer of the NPHDB, David Gunning, reported in July 2023 to the Oireachtas Committee on Health that the costs of merely building the hospital, excluding the costs of commissioning it for use, were now expected to approach €2.2 billion due to further overruns. Only 27 of the 3,000 rooms had been completed, and the expected building completion date had been revised to May 2024. By October 2023, that date had been revised again to late October 2024, and it was emphasised that the last date of the building phase would be the first date of an "operational commissioning phase", so the hospital would not actually open before April 2025. The government confirmed in February 2024 that the total sanctioned budget had now reached €2.24 billion, and that this would now be the "maximum allocation", with no further funding to be put towards the project. By May 2024, the date for "substantial completion" of building had been revised to February 2025, with a further six months before becoming operational. By October 2024, the completion date had again been revised to June 2025."
Bear in mind, they initially lumbered into action on this in 2006. Building didn't start until 2016.
A beautiful stew of "government has no frickin' clue what it wants; changes its mind about location several times; entities already on-site take lawsuits against putting new building on that site; lots of consultants, lots of reports, no action for years; finally decide they'll build it on this site, plus two satellite sites (already a change from "one all-purpose centre"); contractors start building and then Covid lockdown hits in 2020; now it's "yeah it'll be open in January February June well sometime" and the budget has spiralled from €650 million to €2 billion and that's probably not going to be the final cost".
That's not even counting "and when it's open, half the wards will be shut due to staffing limitations" which is what I anticipate to happen. Saw it with our own local regional hospital; fancy new cardiac ward got built after massive fund-raising effort, then - was closed for ages because there wasn't the funding in the budget to staff it. Thankfully, that has been sorted out now, but I'm going to believe in the shiny new promises when I see them fulfilled.
Yeah, a new building means a politician gets their name on it. Money for staff, or even cleaning the building? Not sexy, no publicity (unless it's bad publicity for spending too much), cut it.
Most of the military forces in the world are like that too. Lots of money for new tanks and planes, nothing for maintaining the ones they have.
You need consultants because often they are the only people who know what they are doing. The DEI hires in the bureaucracy can't actually get anything done, so you bring in the consultants to actually do the heavy lifting.
I agree a better system would be to pay those consultants less and actually make them federal employees, but many of them couldn't get jobs with the federal government because they don't meet the "desired" criteria (veterans/disabled/minorities/etc.).
If you don't think it is DEI related I would love to introduce you to the racial/gender demographics of the consultants versus the staff in the two main departments I know well. Advancement is a real non-starter for white males unless they are super exceptional, so where do they all end up? Consulting making more than the bureaucrats and doing their jobs for them. It is a pattern that is clear as day, but no one ever wants to talk about it. There is a huge brain drain where the top staff tend to leave after 3-5 years and go work at like Deloitte or some other big contractor because there is simply more money to be made and they will actually pay competent people whereas the government will even promote them much less pay them.
I work in a bureaucracy (though not the US federal one).
Headcount is restricted by budget, and which bucket their compensation comes from. It's a lot easier to write a justification for hiring a contractor "for 6 months" to fill a perceived need than to go through the process of hiring an employee. Many of those employees start out temporary as well.
Of course those contractor positions then get extended. Subject to budget, though budgeted money gets moved around to different buckets all the time.
The way a friend of mine in federal government explained it to me, it's hard to hire new people above about a ~$100k salary, and the maximum pay is well below corporate rates. If you need some sort of tech service, you have to hire a contractor from Booz Allen or wherever. Of course, making it easier to hire $250k/year government employees has its own problems.
Assuming the work exists and has to be done, you can hire a 100k employee. Or you can hire a contractor. But you'll pay 200k or so for that contractor.
Why? Well, when the work is done, you can stop paying the contractor. So you have to pay for their retirement/sick days/vacation somehow. Not to mention the contracting agency's cut. And the days the contractor is looking for another contract.
Of course, this assumes you're hiring the contractor for a specific project, as opposed to hiring the contractor because there's a hiring freeze on employees.
In my specific area there's a "sunshine" law, where employees above 100k compensation are on a publicly available list. 100k was specified in the law, so as the years go by (almost 30 now) inflation makes the list grow a lot.
People look at 100k and think it's a lot, so they put extra rules around it. I just looked and apparently 100k has lost roughly *half* its value since that law went into effect...
I've been travelling a lot recently, and I've noticed that fines in some countries have their own units that can then be centrally adjusted without needing to change the legislation. Researching a little bit, it seems to be mostly ex-Soviet countries, although I have seen it in Colombia as well.
To me this seems like a no-brainer, at least in a limited fashion, I'd be interested in understanding the drawbacks.
I'm aware that some places have income-based fines as well, but I think that's more of a political question, whereas keeping fines and thresholds inflation-adjusted seems pretty much strictly better than having to double them periodically? The main argument I can think against it is that fines getting eroded creates a sort of automatic sunset clause, but it has the inverse problem for thresholds.
inflation is entirely caused by the government? Or do you mean more that, fundamentally, inflation of significance is. I ask because my understanding of inflation was that it occurs when, society wide, people bid up the price of goods. Thus a federal reserve increase in the money supply causes inflation by increasing ability to spend, but a society wide increase in wealth or looser lending standards do as well.
The US government pays out a $255 death benefit through SS if your spouse dies. It hasn't been adjusted for inflation since the 1940s. Since the law is still on the books, we are still paying bureaucrats to process claims for this 1940s death benefit.
That's the case in Germany. The unit is "Tagessatz", which essentially is a day of your suspected salary for a day. This also has the benefit of automatically adjusting to your income, so that punishment hurts somewhat equally (though losing 10 days as a plumber isn't exactly the same a losing 10 days as Jeff Bezos, even though Jeff Bezos would be paying millions in theory, he would barely notice, whereas the plumber most likely does).
Australia does this too. Fines are written in the legislation as x penalty units, with the monetary value of a penalty unit adjusted each year in response to inflation. Note the exact value of penalty units varies between federal and state and between each state.
The context of the conversation was about Eric Snowden, and why the federal government hires Booz Allen contractors as sys admins in the first place. My friend's answer was that the government salary is uncompetitive, you couldn't get the talent you need.
Almost no one in the federal government at the professional level makes what they would in the private sector. As an extreme example, the Secretary of Defense, for example, makes $246k a year. I don't want to pretend to know what being SecDef entails, but I'm guessing they should make more than a 20-something project manager at a tech company.
Now expand this to the rest of the federal workforce. I'm not sure a lot of entry level engineers are motivated by the prestige of working for government as GS-7s.
I think what happened instead is, you only get people who are already independently wealthy and just want the prestige, at least when it comes to secdef. or at least, the whole system is biased in that direction.
I tend to think that's what's going to happen here if they do ax a bunch of civil servants through relocations, etc. They'll end up having to hire a ton of people back, except now they'll be employees of politically connected contractors and the government will be spending more on them with even less accountability or ability to control the work.
I wouldn't say less accountability, but they do have different incentives. Even contractors who always work full time on government projects are not part of civil service culture and are not aligned with the agency's mission the way federal employees should be (ideally). There are other advantages to using them though, like budget flexibility, scaling up and down rapidly, availability of niche expertise, and (at least in the cases I deal with) lower employee turnover. Whether it's actually cheaper to contract a job out or do it in house is very tough to figure out.
My initial impression is that 'civil service culture' is part of the problem, from my time evaluating fed-funded ngos and their interface with the gov't... but I don't actually know if i have a good reason for thinking this, and it could be pretty localized to my field. I am interested to hear why your impression is the opposite
I'm sure it varies a lot. The people I work with are generally quite capable, and are at least motivated in the sense that they want to achieve the things they're supposed to be working on, rather than just out to further their own careers at their employer's expense. On the other hand, there is a culture of being pretty relaxed and prioritizing work-life balance, rarely working overtime, and being willing to spend lots of time on things like training, seminars, meetings and so forth without worrying too much about it. We work the way we work and things get done when they get done. I think that's a big part of what people complain about with civil service, and although it's nice it does grate sometimes. When I was a contractor it was my manager's job to keep all that stuff away from the engineers and block anything that might distract us from getting real work done. On the whole I prefer the government's way of doing things.
Considering that the contractors do the same work as the bureaucrats, I think this is a difference without meaning. We just aren't on the hook for unionized employees and pensions. Its a tool to side step those requirements, they aren't fundamentally doing different work.
No, they don't. It's a fundamental misunderstanding that you can parachute in some MBAs from Deloitte for 6 months and expect them to have any understanding of the actual mission, constraints, intergovernmental issues, etc. that an agency contends with in time to accomplish anything useful.
I find folks who hold this opinion either haven't worked in a bureaucracy, haven't seen what happens when contractors try to 'do the job' of the real employees, or both.
Yeah, that's the "why can't we treat government like a business?" mindset which runs into problems due to the fact that governments are not for-profit. I hate this new idea of calling the people using government services as "clients" or "customers" because they're not; if they don't like/can't get the service they want, it's not like "Aldi didn't have the thing in stock, I'm going to Tesco instead" for them.
And it's also not similar in "okay this person is renting a local authority house, you can treat them like someone renting from a property management service" because you can't. Sometimes I wish you could, because (again) there are people gaming the system and not paying even the minimum rent, and practically daring the council to evict them because they know they can go to court and the judge will order the council to find housing for them.
But in the main, you can't charge "market rate rent" etc. to those people, and you are (technically at least) obliged to provide a range of services in maintenance and upkeep and accommodating needs for them, so it's not at all like "put an MBA in charge of the housing department and they'll get all the backlog of rent paid up on time".
In my experience the people from Deloitte or wherever are 9 times in 10 wildly better than the bureaucrats and more helpful even if they have only been there a week.
I don't know I get hired as a contractor all the time to basically do shit that employees should have done. It is like my main line of business. And how it generally goes is something that was someone's full time job to do, or say 40% of their full-time job, and that they have had a year or two to complete.
I get done in a couple months of working 10 hours a week.
One $20 million program I worked on the main lady working on it was making like 1-23 phone calls a month, and sending maybe 5 emails. Full time staff head of this project, accomplishing nothing. "What is happening with subgrantee X?" I don't know I haven't talked to them in a 6 months. There are only 30 subgrantees and your WHOLE job is managing them, you should be talking to them EVERY SINGLE WEEK.
Some of the places I get hired the people are overworked, and I legit feel bad of them. But many they are just incompetent/ignorant or lazy. Paying me $20k for two months to complete something that is 5 years overdue (yes I do a lot of things like that) when the full-time staff haven't gotten it done in that time seems like a steal.
That's the thing - if you don't/can't have the personnel in your organisation all in the name of Lean Efficiency, then you have to hire outside consultants for expert advice on tricky problems and large-scale projects, and of course this costs a minimum of twice as much, because now you're paying them market-rate fees.
Do federal agencies need to be capable of being sued by the general public? If they need more accountability couldn’t that be provided by greater legislator scrutiny? Someone with some basic incentive to make sure the agency is still able to function.
"Grow the government while taking away power from individuals to sue the government" does seem like a way to increase government capacity. It also seems out of step politically for both parties at this moment in time.
That's essentially what various forms of government immunity is all about. Prosecutors can't be sued for bringing charges. Judges can't be sued for ruling on cases.
If/when people are able to sue, it was often intentional on the part of Congress allowing it, and always should at least have been contemplated because Congress knows how to add immunity to lawsuits in bills.
Those immunity rules were mostly created by the Supreme Court, who apparently thinks that nobody could possibly enforce and administer the law if they were actually SUBJECT to it. It's like eating at a restaurant where the chef says he couldn't do his job if he had to eat the food they serve there.
In practice the rules get applied in pretty horrifying ways sometimes, as in https://en.m.wikipedia.org/wiki/Connick_v._Thompson where a man spent 18 years wrongfully imprisoned when there was exonerating evidence withheld and the Court insists it's important that nobody be liable for this.
Soldiers can and do get tried for murder, even during war. Murder is a crime under the Uniform Code of Military Justice. Your example isn't even a very good strawman.
No, but I also don't think that really has anything to do with what I said.
Connick wasn't a soldier, he wasn't making decisions while somebody shot at him, and his actions and the actions of his staff falsely imprisoned somebody for almost two decades. And he wasn't even being held criminally liable.
I think it's very clear that the state shouldn't be able to falsely imprison its own citizens and then refuse to so much as compensate them for it.
In your post, you suggested, "couldn’t that be provided by greater legislator scrutiny?" In practice, the legislature would direct the Executive to stand up an oversight organization to provide this scrutiny and provide reports to Congress. This is actually a common thing that they do! The result is that you grow the government by creating a new oversight organization. This growth occurs even if the oversight organization is good at its function.
But critically, inspectors general and the like probably don't slow things down the way the threat of a lawsuit does. They can be more judicious about holding government accountable for things that are actually bad while not picking a fight over stuff that's basically fine.
Agree with all of this. My point is just that creating offices like this is ADDING bureaucrats, which is the opposite of what everyone wants to do! :-D
I think such agencies make a lot of decisions which affect some people disproportionally. Say the FDA bans a certain food chemical (phthalates, lactose, whatever). This would certainly impact some food manufacturers, and it is good that the have the ability to sue.
However, if they only can get sued by persons disproportionally affected by the regulations, this will distort their incentives. If you have to prove that you have personally suffered negative health consequences of e.g. some mild carcinogen allowed by the FDA (lest they get sued by Big Carcinogen), and realistically you will not be able to prove that, because cancer is stochastic -- if the substance increased your cancer risk by a factor of two, that means that it is entirely plausible that you were not harmed by the substance despite getting cancer, and thus you would not have standing to sue.
I think that the court is the correct battleground if some people want to ban something, and some people want to allow it. However, courts take their time, and there are a zillon chemicals someone could either put intentionally into food or detect in food at any day, and having blanket rules which allow or forbid all substances not explicitly regulated would both have bad consequences. Thus having a regulatory agency which can make decisions (a bit) faster than the court system seems reasonable.
Private lobbying basically already fulfils the same function in representing special interests. It’s just the money goes to lawmakers instead of lawyers. Arguably only a dubious improvement.
So whenever I am dissatisfied with a decision of a regulatory agency, I should try to lobby the US congress to add a provision like "Also, the FDA must allow/ban phthalates" to the FDA mandate?
I am not sure what it costs to get special interest legislation the public does not care about passed in congress, but I assume that it is much more expensive than suing in federal court.
Also, I have slightly higher confidence in the court system to actually make a reasonable decision than in the outcome of lobbyists interacting with congress.
In a functioning Constitutional system with separation of powers, yes. To me, the core issue here is that the regulatory agencies have been given huge swaths of power to make their own rules, to be judge, jury, and executioner. In a dream world, all rulemaking would have to get passed as Congressional legislation and executive agencies would only have enforcement powers. All ALJs world be moved to Article 3 courts. And all regulations would sunset every X years.
It's fine to "distort incentives" that way. The FDA's regulations are not supposed to be driven by lawsuits (thus it's not the "correct battleground"). The Constitution gives Americans rights that the government can't violate without being sued, it doesn't grant us the right to sue the government to compel it to act.
All rights come with corresponding obligations to preserve those rights. The right to property means little if the police refuse to investigate crimes against you. A law protecting you from tainted drugs means little if the FDA is corrupt and not doing the inspections that it should. So the proper solution in this case is to get a judge to tell the offending branch of the government to do its job properly.
Did you know that the police don't have a legally enforceable duty to protect you or to investigate anything? You can't actually sue the police to have them do their job--and many times you're not allowed to successfully sue them when they did something wrong! So no, this just isn't at all how things work. Private causes of action are the exception, not the rule, when it comes to government.
No, "equal protection" does not mean that. It can give you the same access to the courts as anyone else, but those courts can still say the government isn't legally compelled to act (as in Castle Rock v. Gonzales).
My understanding is that "they only can get sued by persons disproportionally affected by the regulations, this will distort their incentives* is pretty close to the default rules of standing in federal courts. In order to bring a case before a federal court, you need to have a personal stake in the "case or controversy", in some direct and concrete way.
For more or less the reasons you describe, Congress has often created statutory causes of action allowing interest groups to sue regulatory agencies. And where they haven't, there are workarounds like shopping around for state governments to sue on behalf of the collected interests of their citizens, or finding an individual who is personally inconvenienced by something you already want to sue over and getting them to agree to letting you sue nominally on their behalf.
"I think that the court is the correct battleground if some people want to ban something, and some people want to allow it."
The problem isn't just that courts are slow, it's that they are not experts in the thing they're being asked to ban. The point of regulatory agencies is to hire experts who can make good decisions and, if necessary, defend them in court. If everything started out in court, the courts would have to hire all those experts instead. As it is, when regulations get challenged in court, the decision often reflects a lack of technical understanding because the judge just doesn't know that much about it.
> I think that the court is the correct battleground if some people want to ban something, and some people want to allow it. However, courts take their time, and there are a zillon chemicals someone could either put intentionally into food or detect in food at any day, and having blanket rules which allow or forbid all substances not explicitly regulated would both have bad consequences. Thus having a regulatory agency which can make decisions (a bit) faster than the court system seems reasonable.
Except the FDA does NOT regulate additives very much. Since 1997 (really since the 60's), they allow any company to "self determine" that any given additive is safe, and put it directly in the human food supply.
The FDA on food is the exact opposite of "kill thousands by having ridiculously high 10 year, billion dollar hurdles," in the sense they let food manufacturers put whatever they want into the food supply at pretty much any time.
Since 2000, there’s been only 10 "full process" approvals. There have been 766 new food chemicals since then, which means 756, or 98.7%, have been “self-determined” and put in our food supply with essentially zero vetting or oversight, or means of measuring impacts, or means of recourse if they’ve messed up, either via incompetence or malice.
This was reaffirmed in 2016. The US has ~10k food additives, vs Europe and the UK's ~2k.
“The FDA has established a GRAS Notification Program to help ensure that these ingredients are safe for the ways in which they will be used and to help industry meet its responsibility for ensuring the GRAS status of ingredients they intend to use in food. **This notification is not mandatory; however, the FDA strongly encourages manufacturers to contact the agency** and follow the available procedures for our oversight of GRAS conclusions by submitting a GRAS notice.”
(bolding mine)
“By 1997 FDA had tentatively concluded that it could no longer devote substantial resources to the GRAS affirmation petition process. The agency published a proposed rule outlining a GRAS notification process to replace the resource-intensive GRAS affirmation petition process.”
“The GRAS notification program provides a voluntary mechanism whereby a person may inform FDA of a determination that the use of a substance is GRAS, rather than petition FDA to affirm that the use of a substance is GRAS.”
This seems to be one of those things that changed in the 60s/70s. In the New Deal era the federal government could run rampant and there wasn't much anyone could do about it. Later you get Jane Jacobs opposing Robert Moses on freeway construction, Ralph Nader the crusading consumer-rights lawyer, the environmental movement creating myriad opportunities to sue, and of course the civil rights movement setting the norm that the courts are more reliable than the (constrained) ballot box. https://www.slowboring.com/p/community-meetings-arent-democracy
I think the short answer is 'yes' - it is important to allow some check on administrative power. Congress, for various and sundry reasons, seems incapable of performing that function, which leaves the judiciary.
That being said, Congress can (and should) provide numerous offramps that allow the process to be expedited. That could mean providing specificity over what types of things can be sued over, how many appeals they get, carving out specialized courts that only hear specific classes of cases, etc, etc.
If you're committed to the "agencies make regulations" model, it's really hard to hold them to account. If, for example, the SEC did something blatantly unlawful (eg. require all dealing in crypto to happen through OCC-chartered banks), but there was both a president who was hostile to crypto and no strong majority in favour, the only recourse would be a lawsuit.
You could argue that there should be a requirement that the person in question be "affected" by the regulation, but that skews all business regulations in favour of businesses, all environmental regulation in favour of producers etc. (as they're the only ones who can sue if the general public can't). This isn't necessarily going to trend Libertarian either; if only airlines could sue the FAA, they'd slowly reinvent the CAB by the back door.
Judicial review is a fundamental part of the US system of government. Technically Congress can say "this law or federal agency is exempt from judicial review", but there is zero appetite in Washington for that.
It's not a feature of all governments. Many parliaments are sovereign in which judges purely interpret law. They can't repeal or alter laws.
Sure, but as above comments note, the US in say the 60s did not have *this* level of judicial review of government decisions. So it's clearly not a permanent unalterable feature of the American system of government
I would say yes, under basic rule of law principles. The judiciary system is its own thing and no one should be above it (though recent events seem to suggest some think the POTUS should, which is already plenty controversial).
I really enjoyed this article on how bureaucracy interacts with technology, https://www.niskanencenter.org/culture-eats-policy/, with more details form the trenches on how what Scott identifies actually plays out in practice at various levels.
The only nominative Musk-related determinism I can think of, in English, would be that "musk" is a strong smell, something that would be generated when people really work hard and push themselves to the limit. Easier if you leave the confines of English, as in Finnish his first name means "of life" (so if you mix languages his name would be "The musk of life"), allowing for more advanced level of punnery.
A lot of people seem to view bureaucracy as like some sort of a game-equalizing mechanic in 4X games. Once a civilization is big or advanced enough, it starts generating "bureaucrats" that then create "bureaucracy" that then slows the civilization down. Of course it would be bureaucrats doing that, it's in the name - confectioners create confections, bureaucrats logically then create bureaucracy.
If our position is that bureaucrats would gladly make things as simple for themselves as possible to make their own lives easier and the lives of the people they're regulating easier - and the libertarian view of a typical bureaucrat spending their lives going "BwaHaHaHA! I can't wait for another day trying to make the life of some small businessman harder and drive them to despair!" has never gelled with the actual bureaucrats I know - then a solution for tech companies to make following regulations easier would presumably be developing more software for regulators and their clients to follow the regulations as easily as possible. A TurboBureaucrat instead of TurboTax, in a way. 'course there's probably a fair bit of that already in existence, there certainly is here in Finland.
Software and other technology would probably make a lot of agencies more efficient, but the funding and procurement process for those technologies is equally thorny and larded up with rent seeking.
Naturally, but it's not made easier if the natural process of "cutting bureaucracy" involves trying to wring the amount of available funding to be as low as possible.
"Fujitsu Europe's boss has admitted the firm has a "moral obligation" to contribute to compensation for sub-postmasters wrongly prosecuted as a result of its faulty IT software.
Paul Patterson said Fujitsu gave evidence to the Post Office that was used to prosecute innocent managers.
He added that the Post Office knew about "bugs and errors" in its Horizon accountancy software early on.
...Between 1999 and 2015, more than 900 sub-postmasters and postmistresses were prosecuted for theft and false accounting after money appeared to be missing from their branches, but the prosecutions were based on evidence from faulty Horizon software.
Some sub-postmasters wrongfully went to prison, many were financially ruined. Some have since died.
It has been described as the most widespread miscarriage of justice in British history, but to date only 93 convictions have been overturned and thousands of people are still waiting for compensation settlements more than 20 years on."
Horizon was developed by the Japanese company Fujitsu, for tasks such as accounting and stocktaking.
It was introduced by the Post Office in 1999.
Sub-postmasters quickly complained about bugs in the system after it falsely reported shortfalls - often for many thousands of pounds - but their concerns were dismissed.
The Horizon system is still used by the Post Office, which describes the latest version as "robust"."
So this system was making known errors, both the software company and the Post Office knew this, and yet they still prosecuted people on fake charges and, moreover, the executives in charge have pretty much gotten off scot-free:
The smell meaning comes from the musk deer, which secretes an odorous chemical called musk from its musk gland. The musk gland is so called because it looks like a testicle, and the Sanskrit word for testicle is musk, so called because they look like mice (???) and the PIE word for mouse is musc.
Muscles are also named because they reminded someone of mice!
AFAICT Elon Musk's surname isn't related to any of this, but rather comes from several layers of abbreviating the name of the town of https://en.wikipedia.org/wiki/Montfiquet, where presumably his ancestors were from. Still, I thought it was a neat connection.
Tech helps, but then we have to be willing to reap the benefits of the increased efficiency rather than go "great, now we can do paperwork ten times as fast! That allows us to mandate ten times more paperwork!".
Having worked in the State Dept foreign service, I think they could automate 50% of the work and have 50% less people working very easily. Everything there is slow and feels like it’s from the 1980s. Nothing is efficient and no one cares because it’s so hard to fire people. There is little incentive to perform at private sector levels.
Instead of streamlining, they do the opposite and hire more people, because people like having more coworkers and people like doing less. They also want to hire their spouses, so they actually make up fake jobs for their spouses. This is for morale, but from a tax payer perspective it’s pretty nuts, because each employee costs a lot.
It’s a very slack job full of long lunches and cocktail parties and maximizing per diems. I eventually quit because 1-I was bored/frustrated by the endless paperwork and bureaucracy. 2-it felt morally wrong to be wasting tax payer money like this.
I'm not shocked about the foreign service... I know some people who work for the foreign service of a different country and it's a similar vibe. No wonder half of the U.S.'s ambassadorships are largely political prizes handed out to major donors and supporters. But I will say the jobs for spouses thing is a legitimate problem in an agency that is asking its employees to move around the world every few years. It's very hard for a spouse to put together anything resembling a decent career if the foreign service doesn't do it for them. You can see universities being forced to do the same thing with spousal hire programs.
I know. The spouse jobs make sense for morale, but they’re part of the bloat and some can actually be in positions that impact speed—like office managers. They also have too many local hires for all the same reasons I’ve mentioned, and though the local hires are usually cheaper, they’re not incentivized to be efficient or innovative.
Plus the whole system of moving everyone every 2-3 years also makes things inefficient because you’re constantly spending all your time just learning a new language/culture/job and unpacking and packing and on vacation in between posts.
Also a lot of flying to places to look at things when sending a video or photo could have sufficed, but people like travel and perdiem so they go in person.
Actually there is so much waste I can’t even get into it, but it’s super ridiculous looking back on it.
Ideally, you'd have foreign service operatives who concentrate on one area of the world where there just a few languages and where their previous experience is actually useful. "Area of the world" could be very broad - like all of Latin America, where you'd only need two languages (Spanish and Portuguese), and some things you learn as a junior diplomat ("Third Secretary", in diplo-speak) in Colombia would be useful years later as an Ambassador to Argentina or whatever. Also, you'd build up set of contacts that could be really useful across the region.
Obviously, there are some countries that are much more distinctive in culture and language than others (however good your Vietnamese might be, it doesn't help a bit in learning Thai), but they still have important relations with each other and understanding something about the other countries that are important to where you are posted is of real value.
Just to pick a random British diplomat whose autobiography I read - Craig Murray (who became semi-famous for something to do with opposing torture of people suspected of terrorism, and I'm not being more precise because I forgot the details and I don't want to be sued) was Ambassador to Uzbekistan after working in first Poland and later Ghana as a junior diplomat, as well as working on UNCLOS (the law of the sea) when back in the UK between postings. Yes, they sent an expert on the law of the sea to one of the only two doubly-landlocked countries on earth. What do you learn in Poland or Ghana that is useful in Uzbekistan? What contacts could you have cultivated that are useful now?
And this does sort of happen sometimes, but there are also cases where it doesn’t. I remember meeting an officer who spoke fluent Mandarin and was frustrated that they wouldn’t post her in China and were instead making her learn French because they’d decided to send her to a small post in Africa instead of to China. Because sometimes they believe that if you speak the language too well, or understand the culture too well, then you’re too open to foreign influence.
There's also a prestige/priority dimension here. As you say there are plenty of regional specialists in the Foreign Service. But I've also known people who got into trouble--not career-ending trouble, but trouble nonetheless--and this changed the possibilities for their next post. There's also the question of hardship posts (some people make their careers early doing this, but don't want to be in Afghanistan forever) and differing levels and means of interaction between the diplomatic staff and hosting country.
On the spousal job question: my grandparents were in the Foreign Service during an era that predates this policy. Wives were subject to extremely rigid expectations that constituted an unacknowledged full-time job. My grandmother chafed against these (I shared some of her recollections here: https://tomlee.wtf/2011/02/21/a-foreign-service-wife/) and even, I later learned, wrote a series of sardonic articles about her experiences in Foreign Service Journal under the pseudonym "Mary Stuart". I don't know much about the system that superseded it, but the old ways were impracticably sexist.
Yes. Correct. Even when I entered the service in the mid-2000s the sexism lingered and I had multiple incidents that would now be metoo material. Also, the wives were still expected to constantly host like First Ladies, and there were fewer women than men in the upper levels of the service. It felt very colonial at times even in the 2000s. Also, I was working mainly in Africa, which probably contributed to things feeling more antiquated.
Thank you for the link - it was a fascinating read.
The one thing that struck me was her reference to the SATs - I wonder if she's got them mixed up with some other test, or if they were really that different back then? She describes them involving a composition; the SAT when I took it had an essay, but it was a novelty nobody seemed to take very seriously (except the students, of course) because it couldn't be objectively graded like everything else, and I believe they removed it shortly thereafter. Possible she meant a subject text, or they were really different back then? I don't know much about testing during or right after the war; my father wouldn't have taken it until well after. It's interesting it goes back that far.
Anyway, really interesting to see her perspective. She must have been a very strong-minded lady.
As a fellow federal bureaucrat (well, okay, I'm a mini-bureaucrat in training, a lowly GS-7 clerk), I wanted to echo on to a point here: most of the posters tend to think that bureaucracies should be resident centers of expertise regarding their field, staffed by autist wonks who are deeply read in their field (which is to say, people largely like the poster here). Fundamentally, though, expertise in the bureaucracy's mission isn't what gets your promoted in the bureaucracy, it's expertise in the bureaucracy's personnel system and also in knife-fights with other bureaucracies over various rice bowls and other resources. Tismy wonks don't do well in those kinds of battles has been my experience, so the people who posters think should be in charge tend to end up shunted aside into unimportant subelements.
Also in major impacts to efficacy of the bureaucracy, the need to "error-proof" decisions and actions (not just from external lawsuits but from internal admin grievances) is unpleasant.
In the government tech contracting world, e.g., contracting for NASA or DARPA, they move program managers to a different location every 3 years. I can't imagine any reason for doing this that would justify the massive waste it causes. In my experience, contracts under a million dollars were usually just throwing money away, because by the time you got through the 2-to-4 year process of writing the grant solicitation, choosing one grant proposal, doing the work, and submitting the final report, the person who'd wanted that work done, and who would get credit for it if it were used for anything, was gone; and the entire project would just be dumped in the trash, sometimes without anyone even reading the final report.
DARPA program managers are explicitly term-limited, that's part of the way the agency was designed to work. The theory is that it forces you to bring in fresh thinkers every few years, and that creates a culture of PMs who take risks rather than becoming careerists who just approve the same non-risky programs and contractors over and over.
Does that create inefficiency? Obviously. Does it have benefits? From my experience I tend to think it has huge benefits: DARPA has a great reputation for sponsoring high-risk/high-reward results. But I can't prove it. I just think it's important to recognize how easy it is to take a deliberate organizational feature and casually misconstrue it as arbitrary government efficiency. (To stress, I'm trying to start an argument about whether the policy is good or bad, just to point out how important it is to fully understand things before you start cutting.)
This is the problem with superficial discussions of government reform. At a superficial level you might conclude "this seems inefficient, let's fix it." With just a little bit more understanding you might find out the thing you're fixing is an enormous Chesterton's fence. We live in the most powerful and richest nation in the world and that's largely because our forebears took state capacity seriously and *didn't* organize things with serial number lotteries.
Ambassadors feel like an anachronism anyways - a relic of a time when communication delays meant that you had to give your representative in a foreign court a lot of autonomy and prestige to negotiate on your behalf. The low-level stuff can be done by State Department workers lower down in the bureaucracy, while the high-level stuff is just done directly between heads of state and their immediate cohort of people.
They're also useful because when your people are furious about something Country X did, you can "summon the ambassador" of Country X to make it look like you're actually doing something about it.
" jobs for spouses thing is a legitimate problem" - This is not a problem for the employer to solve. The employer can just ignore it, and employees will self-select in the direction of higher pay (meaning these jobs will probably need to pay a bit more), and those without attachments, and or those looking to spend less time with their spouses.
This is not unlike basically any other job-specific downside, like danger. You have to pay a bit more, and you get folks more willing to engage with whatever that danger is.
"This is not a problem for the employer to solve. The employer can just ignore it"
The trouble is CHANGING the status quo. As of now, the entire field is used to having their spouses taken care of. If this abruptly changes, employees will self-select out in the middle of their careers, and there will be a period of instability with the average quality of an American diplomat falling even lower.
This is a very thin view of the employment market that misunderstands why businesses ever try to offer side benefits or improve working conditions. Employees make employment decisions based on many factors, with money offering diminishing returns, like everything. Employers in turn make decisions based on what gets them the best employees at the lowest price.
Imagine you are trying to hire a highly-qualified career diplomat to a post in the Congo or a star PhD to a professorship in the middle of Kansas. This person is not married to a nobody. Their spouse is almost certainly highly educated and intelligent. The couple will make their decision together and the trailing spouse is probably not interested in being a trophy husband/wife interrupting their burgeoning career for a resume gap.
There is an amount of money that would probably convince the spouse to suck it up and move so that the couple can save money for a few years. But this amount of money would probable be higher than what both of them could have made together back home.
Now imagine you are the agency doing the hiring. Do you offer a $120,000 employee another $100,000 to convince their spouse to agree to the deal? Or do you hire the spouse into a $80,000 job that also needs to be filled, even though they provide only 75% of the value that the optimal candidate would have provided? This is a no-brainer.
I was also an FSO, long ago. The Foreign Service is a small bureaucracy in a historically low-productivity field and many of my former colleagues were (and are) hard-working, though you can question how useful much of their work is. Spouses get hired because (i) it’s cheap (ii) most FSOs are quite bright and they and their spouses could get decent jobs in the States if they stayed there. So unless you have a spouse who just wants to stay home and care for kids (but nannies are really cheap in most of the world) there’s going to be a demand for jobs for spouses.
Yes. Definitely correct on the brightness and hard working nature of the officers, but I can’t help but think the whole system needs an overhaul. Most employees were frustrated by the slowness and inefficiencies also. But then a lot also milked the system. Or this is what I saw… could have just been the specific posts I was at, but I sensed it’s widespread.
This is a great case study actually. State budgets are incredibly popular to cut, and it's also popular for each new Secretary to create a new dedicated special office to absorb the few personnel remaining, so personnel dedicated to traditional mission have been basically vanishing over time. Some career tracks pick any post they want because of chronic understaffing.
And yet also... the core remaining, without the resources to do effective diplomacy, are a certain breed. Agree maybe a third are self selected for the benefits and just riding out the bipartisan disdain for their work, another third are ground down by the impossibility of driving effective change and become cynical with the same result. And the last third are wonky savants who understand an unsettling amount about state capacity and driving policy changes for the benefit of the US and humanity. And this last group is responsible for a phenomenal number of US wins that nobody will ever report on or notice.
So it's both true that there's a lot of personnel wasting time and also that if you cut personnel, which we have already done, it directly reduces the department's ability to do really clever and effective diplomacy.
At what multiple do we fund the military relative to diplomacy? And then when we look to our bag of hammers to solve policy problems we see an awful lot of nails. Then the public wonders why we get entangled abroad so much.
Better firing authority would be great, but should also be joined by better hiring authority, and greater accountability for results. Those seem better than mass firing to me.
Is it a very slack job until the day there is a major war or other international crisis?
This seems like the type of agency where the amount of work is not constant, and you desperately want to have enough excess capacity to handle the surges quickly because they correspond with the most important things you ever do.
My sister is actually married to a language officer in the Danish military, and got hired as a spouse so they could live together. Seems kind of necessary thing to have if you ever want to keep your workers past their wedding.
Also, speaking from the same example: It might be that most of the time the foreign service / military / diplomats etc don't have enough to do because the world is mostly peaceful and not involved in too many international crises. But you still hire that many officers fluent in Russian, because the day one of your submarines accidentally bumb into a Russian sub under the polar ice you really are going to need all the diplomats you can get your hands on.
But that's the level that is *not* going to be fired, is the problem. Vivek may get it pushed through to fire the receptionist in the local office or the Grade III doing filing, but the diplomats' spouses won't be touched because that's way too hot a potato.
So the inefficiency and bloat continues, while service on the ground declines.
I suspect Vivek is thinking of a version of this. Each worker can actually process 1000 forms. They process 10 because that achieves the departments performance objectives. If half were fired, theyd groan and process 20. This might not apply at acute crunch times like the translator example, but would day to day
> I was surprised to see someone with such experience in the pharmaceutical industry say this, because it goes against my understanding of how the FDA works.
The obvious explanation is he is just lying through his nose to score points in some intra-MAGA power struggle. It's like elected Democrats who talk about corporate greed - in private most of them would probably admit that they know punishing people for wanting to have money is not actually a good way to solve market failures, but that's the rhetoric that polls well so that's the rhetoric you'll get in public.
I think it's unlikely to happen as-described. That said, I don't think it's necessarily the case that he feels that it's only in his interests to to propose measures which will not harm the effective operation of the government, or that proposals which do will not be implemented.
Trump says a lot of stuff that seems to be maximally offensive/unhelpful/impossible on the way towards achieving a much smaller goal closer to what he probably intended in the first place. Vivek was directly following Trump's style and approach when campaigning.
Best guess, Vivek wants to cut the federal government and bureaucrats specifically by significant amounts. But not 50% and not arbitrarily. He likely wanted to scare the agencies into doing a better job themselves, rather than facing an obviously unfair and highly disruptive "you lose your job because your SSN ends in an odd number" approach.
Having listened to him for X hours on podcasts (the person he pretended to be on the debates was fairly cringeworthy; he's completely capable of playing a part for personal gain), and how many idiosyncratic complicated thoughts he has that may or may not work but are integrated into a coherent world view that doesn't map to political or personal incentives for him to have, I have a lot of good faith skepticism about this.
I predict that if Scott and Vivek got on a podcast for three hours and talked, Vivek would either change his mind or pass a turing test of explaining why he disagrees in a way that sounds reasonable to a neutral audience.
'Craft an explanation for why your policies are internally consistent and not self-serving that is convincing to the average person' is not a very high bar for a smart and experienced person with inconsistent self-serving policies, though.
Where is the boundary between a post that can be counted as hyperbolic but making a worthwhile point, and one that is so ludicrous and daft that it should get the poster laughed out of the room? I would argue that Ramasvamy's suggestion is on the latter side of the divide.
Not only does removing every other bureaucrat at random halve the capacity just on a "people who can get stuff done" basis, it will probably reduce the capacity close to 0 for a while, because at least some of the people you throw out will be specialists with arcane knowledge necessary to keep the system running, and many will be the liaisons between different branches who now suddenly have no idea who to talk to when some issue related to another division comes up.
If you think some of the work done by bureaucrats is unnecessary, by all means, find out which part of the work that is, and optimize it away. Until you're willing to do that, making blanket suggestions like that doesn't really help.
Or perhaps that the current system is so fucked up that large changes even in random directions would be a significant improvement over carefully-directed incremental changes.
Yeah, that's why America, the rich country, has the fucked up systems, while Russia, the poor country, has the really strongly functional ones. Come on man. America's system of government is the most blatant Chesterton's Fence in the history of the world.
I've never been to Russia, but I bet there's enough corruption to make the system work fine. They'll have regulations about what kind of doors pharmacies can have, but everyone understands that when the inspector comes by to inspect your doors, you need to pay him off, not try to demonstrate to him your doors meet regs.
The US's problem is the abysmal lack of such salutary corruption, "the people's deregulation."
Tacitus said it well, "The more numerous the laws, the more corrupt the state needs to be."
It's been shown again and again that random number generators are *better* than experts in some domains. There are multiple examples in Kahnemann's book 'Thinking fast and slow', but I think the classic example is that managed funds tend to underperform index funds.
But he didn't get laughed out of the room. He's one of the leader's top lieutenants now.
Besides, I don't see how it's that stupid. If the bureaucracy is too understaffed to actually approve anything, they're also too understaffed to actually enforce their rules. It's a temporary measure until the department can be eliminated and replaced entirely.
Right, he didn't get laughed out of the room. Neither did that other dude who promised to have people on the way to Mars by 2022, and neither did their overlord who promised to end the Ukraine war on day 1. Somehow that points to a much deeper, much more serious problem than any failures of US bureaucracy.
There are a lot of people who think they'll personally benefit from living in a US without state capacity, even if the overall economy shrinks by 60%. Based on history I think most of them are wrong, but you can't argue with personal greed. There are other more capable people who see eliminating the civil service as a way to eliminate one check on untrammeled executive power; they fully intend to staff a bureaucracy, just one filled with politically-appointed loyalists.
The key is to probably establish a secondary bureaucracy that evaluates the first for efficiency. Every form that the bureaucracy requires should need to be justified by a secondary form explaining why it is necessary work, which will be evaluated by this new department.
This already exists, and is called OMB. Go read about the Paperwork Reduction Act and see if it has made government better or worse and report back on your findings.
Perhaps a third layer, so every action done by the OMB has to be justified by a form that explains the benefit. These forms could be evaluated by the new overseeing department.
It wasn’t meant literally or seriously. Its purpose was to show how good a troll he could be as part of his audition for VP.
He is a high honors graduate of the Andy Kaufman School of Performance Art.
Think of Andy during his intergender wrestling champion phase preparing to wrestle a woman in some southern backwater venue holding up a bar of soap and roll of toilet paper to taunt the audience by implying that they need remedial instruction in hygiene.
Edit:
As a side note Andy worked out the mechanics of that bit with another performance artist, Laurie Anderson, acting as a shill playing a random woman in the audience.
I mean, sure, if you want people to die of workplace accidents, building fires, collapsing biildings, traffic accidents, airplane crashes, food poisoning, counterfeit medicine, air and water pollution etc, go ahead, remove all regulations. But don't come crying to us after your children's funeral.
Most of the improvement in life expectancy and safety has more to do with technological and medical improvements rather than regulation. Of course, the regulators are happy to take credit.
Could it be possible that the regulations that are actually necessary to solve these problems consist of a small fraction of the regulatory code? and that the rest do nothing at best, and make things worse at worst? Were these things *really* that much worse 100 years ago?
How many regulations actually make these things better? Seems that regulations on nuclear energy have made air pollution worse, forest clearing protections have caused more houses to burn down, and medical regulation has resulted in preventable deaths due to medicines being delayed (see the covid vax - finished in January 2020!).
Airline deregulation resulted in many more crashes except, oh wait, it didn't. The price dropped like a stone and it got safer. Hmm. Maybe they're just full of shit?
Another example: if you could buy a new build Toyota Corolla with 1980 regulatory standards, it's likely to not be *that* much safer, but would probably cost ~$10K. Maybe the fuel efficiency would be a little lower, but I haven't even mentioned oil drilling and pipeline regulations. Would anyone buy anything else? That such a product is not allowed to exist is money stolen *from you*!
Yes, technology has improved. So has the capacity of the chemical and food industries to poison us with convenient but unfortunately toxic substances, the capacity of the construction industry to build skyscrapers that kill 10000 rather than 100 people when they collapse in an earthquake, the capacity of the financial industry to lure in more suckers and defraud them of their life earnings, etc. Products have become more complex, and so have their supply chains, exponentially increasing the potential for fraud and deception.
"Of course, the regulators are happy to take credit." Probably with some justification. People in other countries have access to the same technology what the US does, at least in principle. If you follow the news, you hear a lot more reports of people dying from accidents related to shoddy maintenance, shoddy construction practices and horrendous workplace safety in poorer, less-regulated countries (including countries that have regulations on the books, but are too corrupt or incompetent to enforce them) than in the US and Europe. Sure, those regulations cost time and money and effort. If you don't want to afford that, feel free to move to, I don't know, Bangladesh?
"Could it be possible that the regulations that are actually necessary to solve these problems consist of a small fraction of the regulatory code?" That is a possibility. It is also a genuinely hard question to answer, since it would require going through the code, figuring out the history and justification for each item and judging whether it's worthwhile. Again, I would support that effort, but I am not on board with assuming the percentage of pointless regulations is "~100%" from the start.
I am not a bureaucrat, but I work as a software developer, and I can tell you that we have a similar phenomenon of dealing with bloated, borderline incomprehensible code. Sometimes fits of "no one needs that stuff, let's just throw it out and start from scratch" pay off; often, they are followed by lots of bug reports from angry customers and moments of "oh, so that's what those lines of code were for."
I think there might be other factors in Bangladesh...mainly the poverty.
And I'm not convinced that regulations are what got us out of poverty. From 1750-1913 we went from the first steam engines through to Belle Epoque Europe and Gilded Age America. Basically, if it's a popular tourist destination, it was probably built before our current massive regulatory code. And during this span, were there industrial accidents, bridge collapses, and the like? Absolutely! During this span, things seemed to be improving in the safety department over time, with the exceptions being in areas that were extremely poor.
So, if the regulations block economic growth, it's very possible they're killing more than they're saving. While ~100% of the code being useless is likely hyperbole, do keep in mind it plays to the regulators' interest to keep the code as long and esoteric as possible. The analogy I would use is software developers writing unmaintainable disgusting spaghetti code on purpose to increase their job security.
Also, I have to call you out on: "the capacity of the financial industry to lure in more suckers and defraud them of their life earnings" - if this was a real motivation, they'd ban sports betting and other gambling...they haven't. Indeed, they've massively expanded these while making it basically impossible for normies to invest in early stage startups. You could have made a killing getting into Apple or Microsoft early - but you were locked out of getting into Facebook, Stripe, or SpaceX early. Was that really for your own good?
"And I'm not convinced that regulations are what got us out of poverty." No. Regulations are (part of) what brought security and health, after the worst of poverty had been taken care of. More money doesn't buy safety all by itself. Of course, while you're dirt poor, you're more willing to put up with horrible workplace conditions, pollution etc, but once people can afford nice houses and nice cars, they can also afford the regulations that make sure they can enjoy those houses and cars alive and healthy.
"if this was a real motivation, they'd ban sports betting and other gambling..." - it sure was a real motivation, and it worked. When's the last time you heard of a properly regulated, old-fashioned bank going under due to fraudulent dealings, and all their customers losing their money? When's the last time you heard of a large crypto company going under due to fraudulent dealings, and all their customers losing their money? What's the fundamental difference? Wasn't De-Fi supposed to be just like regular finance, just with the opportunity to make big money unburdened by those pesky regulations?
Now I don't know what led to the liberalization of sports betting, but I think it was a shitty idea. (The difference between banking and gambling, of course, is that gambling is always understood to be optional, but you need some way of participating in the financial system to receive money and pay your bills.)
Regarding the startups, I disagree with you as well. Most startups fail utterly. If you're so intimately familiar with a startup that you're convinced that they will be a massive success with good reason, and you know that you could handle the loss of the money, do the work, talk to them, do the due diligence, become an investor. If you're a somewhat informed investor, buy shares of a fund that invests in that startup. If you're a normie who just happened to hear a convincing sales pitch... then it's no better than gambling. (Also, opening up the field for normies would change the incentive structure for the worse. Again, look at the crypto space: lots of startup-like projects collecting investments from normies, many of them with some well-known name or other to back them up. What percentage of them end up not just as unfortunately failed businesses, but outright scams and rug-pulls?)
A false proposition implies any proposition, so it's true both that if ~100% of the work they do is in unnecessary then firing half of them will be fine and that if ~100% of the work they do is unnecessary then firing half of them will still be a disaster because it will unleash the wrath of Nyarlathotep.
That's an interesting choice of date, because the Triangle Shirtwaist Fire, the most central example of "why safety regulations exist," happened in 1911. Would I like to go back to an era where it was legal to lock your employees inside the building and let them burn to death? Gee, let me think about it!
Ah yes. Because some regulations actually are useful that means that they all must be useful.
It's like how after 9/11 they stopped future hijackings by just locking and reinforcing the cockpit doors, but somehow also got everyone to go along with an absurd system where idiots molest you and micromanage the amount of liquids in your bag and make you take your shoes off for no reason while we pay billions for the privilege.
Then someone says: "maybe we just just go back to the pre-9/11 airport security and just keep the doors locked" and people respond "Wow! You want to have another 9/11 to happen!"
I guess it's "common knowledge" that a lot of regulations are results not of lawsuits, or fear of lawsuits, but of inter-department warfare, where one group of bureaucrats try to grab more power by introducing newer, longer, stricter, more urgent forms. I haven't ever worked in government, or a big regulation-heavy company, so I don't know how true is that. But it's certainly the popular perceptions of bureaucracy. So maybe DOGE can target such regulations and be successful.
But in reality, Musk's goal is probably to eliminate most oversight altogether, and damn the consequences (as long as he gets to Mars before his death, he personally doesn't care about the rest of us; and after his death, Mars can build its own bureaucracy from scratch). He just can't say it out loud. So halving the number of people staffing various regulatory agencies might work: it will bring work to a halt, and so DOGE can point fingers and say "well, now those people are completely useless, they haven't approved (a single drug, a single rocket launch, whatever) in a year, let's get rid of them altogether!".
"But in reality, Musk's goal is probably to eliminate most oversight altogether, and damn the consequences" - bingo. It's one of the most brazen acts of corruption I have seen in a civilized country in recent memory. Dude bought himself a goverment position that can get rid of the regulations that he perceives as holding back his companies and also gets to have a say on government contracts for his companies and their competitors. Conflict of interest, anyone?
Well, as a non-American, I'm kind of interested to see how it turns out. I grew up on books like Heinlein's "The Man Who Sold The Moon", and maybe Musk will become "The Man Who Sold Mars". Living out those 50's libertarian tropes.
I remain confused about what "living on Mars" is going to accomplish. Even if it's possible to build anytime soon, a Mars colony is going to be dependent on critical materials shipped from Earth for all of Musk's lifespan and likely decades beyond. You definitely want to make sure that you have a stable political environment here on Earth. (In some sense you're even more vulnerable than someone living in a compound here on Earth, because every nation with space-faring weapons has a veto on your survival.)
I imagine it'd be enough for him just to get there, to prove that it's possible and maybe re-ignite interest in space exploration, or even a space race. I noticed he does talk about future Martian nation, but only rarely and vaguely, so I don't think he has a grand plan for that. Just go into space, become a hero, be buried on Mars, and the rest is for others to work out.
Besides, I don't think it is possible to achieve a "stable political environment here on Earth" in Musk's lifetime, anyway. Certainly not with his personal best abilities. So his choices are down to "forget about Mars, and concentrate on something more practical and boring" and "full steam ahead, and damn the Earth", and, well, he's the kind of person who chooses the later every time.
From a human perspective, there are various possible economic motivations (all pretty blue sky [red sky?] but something will probably work out), and many more scientific ones, but the biggest in the long-term is just the backup capacity of having more than one life-support system for large numbers of humans.
From Musk's perspective, there's the honour/hubris (take your pick) of the founder, plus someone has to ship all those critical materials and that's likely to be a good source of business for rockets for some time, plus as far as I know he's human and so should share those motivations as well.
There is no incentive in the current (well, incoming, but it seems like Trump is already in charge) administration for improving the United States ability to trade with other countries. The longshoremen bargain is best understood as purposefully slowing down our ports, as a very easy to implement tariff.
It would require an Act of Congress, although I think Trump could grant exemptions under circumstances (which would almost certainly be litigated if they were too generous).
From what I've heard, the reason why reform is so hard is not just opposition from the domestic shipyards remaining plus the unions, but from the Navy and its favored politicians in Congress - they really like the idea of having a US-made fleet of civilian cargo vessels that can be turned into a merchant marine in case of a conflict, and also are very afraid that the few remaining shipyards will go bust and wipe out the US's ability to make military ships.
I just checked, and the Jones Act creates a private cause of action. So Jones Act can just be enforced by the courts without the executive branch doing anything.
20% do the work, 20% are so bad they obstruct the 20%, and 60% are just waiting for retirement. Every GS-14/15 I have ever talked to over the past 40years. BTW- of course anyone who interacts with the bureaucrats says they were helpful. What sort of idiot says they are useless if you have to continually interact with them. Also GS 14/15s- do not irritate a bureaucrat in the 60%. You have just given them something to do until retirement- cause you problems.
I have a relative doing work for the government who is a true believer in the cause the department is doing, who thinks the government is good and trustworthy, and a significant portion of their job as they describe it to me is routing around the useless unfireable coworkers so they can accomplish their needed mission. Give her a magic button to fire all these people and she would.
This happens in every organization. Eventually you have a small percent who get most of the real work done, a large group that performs their minimal function to not get in trouble and a small percent who are actively detrimental. The large group is mostly just "putting their time in" "waiting for retirement" or whatever. Many may actually be reducing the productivity of the organization in various ways.
The problem with the federal government is that it's intentionally hard to fire people and the ones who can force that to happen lack ways to legibly determine who is good and who is not. So instead churn and the natural process of removing the unproductive is stalled out, gradually building the numbers of people just filling the ranks. Most of them are at least doing what they've been hired to do, so if their jobs are designed well they can at least keep filling out the forms and passing them along - in line with Scott's article. Better employees would be working on ways to reduce the needs for forms and streamlining the processes, etc.
That Musk was able to fire ~80% of Twitter's workforce without the company breaking speaks well to his ability here. But the federal government is much larger and has way more legal and cultural roadblocks to removing people. Musk also can't just ignore the complaints of people upset, because the government is required to be responsive to the public.
Federal civil servants aren't legally unfire-able. They could terminate the 20% if they wanted to, but when they stick around it's usually because their managers don't really give a shit either - don't want the hassle of retraining people or going through the termination process, etc.
Optimistically, resultant disruption of routines makes most of the remainder of the "actively detrimental" set easier to spot and remove, since half of the folks covering for them are gone, while those who turn out to be irreplaceable get promptly reinstated as soon as the "don't know what you've got 'till it's gone" effect kicks in.
The government will surely be an employer of choice in the future for high performing employees if they randomly delete half the current high performing employees.
Is there a Laffer curve for the bureaucratic state? Adding marginal bureaucrats from zero may improve a country's "output." Adding marginal bureaucrats to a bureaucrat-heavy country may increase sclerosis and decrease "output." In a plot of output vs. # of bureaucrats, the start of such a curve is positive in slope, and the far end is negative. A continuous curve must therefore have some maximum between those two points. Vivek is bolder than saying we're bureaucrat staffed beyond optimum, he's saying we're sufficiently beyond the max that reducing count by 2x improves output. Scott asserts otherwise. Scott, any thoughts as to what your curve looks like?
I think this is a very interesting question and I'd love to know about international analyses. Some countries are small and have small governments. You'd think that some government work would scale with country population but some wouldn't. Are these small countries doing less stuff than their bigger counterparts? Is the country worse as a result?
Even if that model held, don't overlook teh fact that it may be a ratchet.
IE just because adding 25% more bureaucrats at T1 makes you less efficient at T2, does not mean that firing the same number of bureaucrats at T3 will make you more efficient at T4. If the system found jobs for those extra bureaucrats to do and made important function route through them, removing them might just crash everything.
Which is not to say that you can't carefully unwind those effects to reverse the ratchet to the optimum level, but 'fire half of people at random' isn't going to do it.
A large contingent of the federal work provides services, like health care to veterans, screening passengers at airports, or helping the public navigate things like paying taxes or receive Social Security. Will these functions become more effective with fewer employees?
I think you may not have fully steelmanned the "thesis where red tape expands to consume the resources available to it".
What happens if with fewer bureaucrats they justify a regulation with a document only 500 pages long rather than 1000 pages long? When there's a lawsuit, does the judge just automatically say "not adequately justified!" and throw out the regulation? I don't think so. The judges will reduce the standard they expect.
At least they'll reduce the standard sometimes, when they personally think the regulation is needed. One might object that this gives too much power to judges, to selectively decide the justification was adequate. But don't they already have that power? If they don't like the regulation, they can say that 1000 pages wasn't enough, that they should have produced a 2000 page justification...
You seem to have misunderstood "number of pages." It is, in the post, a proxy for "a document adequate to convey appropriate information." Just halving the information, does not halve the threshold for "adequate." Otherwise, repeating the process, every document could be one letter long.
Maybe if the regulation is so important, they can justify it with fewer words? If the usefulness would be obvious and wouldn't require them to torture logic?
This is how laws were written in the 19th century. But presidents started using dubious logic to get around the law and/or extend their authority, so Congress started writing longer bills to hem the president in.
When someone complains about 500 page bills they're basically saying "I prefer to let the president decide everything." That's great when your party is in control, not so good when the other side gets a chance.
And the SCOTUS Chevron decision means Congress will have to write even longer bills than before.
So it's true that nothing is infinitely inelastic; if an agency has half as many people they will probably eventually cope by reduce the length of their reports somewhat, and if judges keep getting shorter reports they will probably cope by reducing their standards somewhat.
But the failure mode of this insight is thinking that only the things you *like* are elastic, and other things aren't. The number of drugs approved each year is elastic, that can go down. The number of mistakes made is elastic, that can go up. The amount the government pays out on lawsuits it fails to properly defend against is elastic, that can go up. Etc.
So yes, it's probably true that when you pull the legs out from under the entire system and cause everything related to it to malfunction, some of that malfunction will happen in the things you dislike and reduce them some amount. But a lot of the malfunction will also happen in the things you like.
I mean it just depends. Bureaucrats have pretty wide latitude to just ignore parts of the law if it is too burdensome to enact, and that happens a lot until there is eventually some lawsuit. And sometimes the lawsuits never happen and no one ever notices.
Just one example there is a requirement that all outside hiring for large swathes of grantees (I am going to be vague here) always need to first market/consider "low-income" people.
Yes this includes if you need an architect or a lawyer or a nuclear engineer. What does a low income nuclear engineer look like? Certainly not someone you would want to hire generally I would guess. And this is true even for say people building a structure. Why would you seek out low income ones? Also it sort of binds against the prevailing wage rules besides.
Anyway the rule is garbage and stupid, and so what happens is grantees promise to do X, some even stupidly try to do X, but the bureaucrats never ever check X and the grantees figure that out eventually , and soon everyone is just filling out forms and saying they do X, and the bureaucrats are saying they checked X, but no one is actually doing it because it is asinine.
That happens a lot. Maybe 10% of the time for any given rule +/-? Some small portion of the expertise I sell to government entities as a consultant is having the experience on what rules the federal government is serious about, and what rules the government just pretends to be serious about and/or has actively forgotten about.
Thanks for a great post. One thing it is missing is some glancing at what has already been tried at the federal level. Idaho's approach of changing the status quo from "regulate more" to "make sure the regulations really meet a bar" is kind of the approach Trump45 took to, at first glance, noticeable effect: https://www.cato.org/regulation/winter-2023-2024/was-trump-deregulator
It's also about ignorant politicians who have no experience with managing processes, so they write the laws without understanding the costs of them.
Look at the resume of your average politician. Barack Obama, Joe Biden, Donald Trump. Before they came to office, what large, high-volume organisation did they manage, or were even part of? And I can say the same about the UK. They come from law, journalism, trade unionism. None of them have worked in retail banking, Amazon, WalMart.
And if you don't work in these fields, you don't understand things like KISS. That in the era of computers, it's complexity, not scale that is the big cost. That if you have 1000 pages of rules, someone has to write the code for that, staff have to be trained in it, lawyers, accountants and so forth have to learn it all. A legal case will happen because there is a different tax on biscuits and cakes in the UK, and how is that defined? A better idea: tax both, or tax neither. Or even better, remove sales taxes entirely and just raise a tax elsewhere to make up the shortfall. Then you can fire the whole sales tax department.
>Barack Obama, Joe Biden, Donald Trump. Before they came to office, what large, high-volume organisation did they manage, or were even part of?
Trump: building and operating various hotels and casinos. I get the impression his involvement in the actual operation of the businesses (as opposed to promotion and financing) was relatively nominal, thought.
Biden: Had been a Senator for decades, including being chair or ranking minority member of various committees. It's quite possible for a back-bench legislator to remain ignorant of how the federal bureaucracy operates in response to legislative rulemaking, but it would require a particularly dense and slothful Senator to remain ignorant over the course of something like Biden's career.
Obama: back bench legislator, and before than an attorney at a civil rights law firm. You have a good point in his case.
Bush the Younger: Governor of Texas, and before that CEO of an oil exploration company and owner of a major league baseball team.
Clinton: Governor of Arkansas, and before that Attorney General of the same state.
Yes, and none of those are operating a large organisation with millions of customers. They were politicians, except Trump. And his hotels and casinos were not exactly a huge success, were they? He made his money in real estate which is mostly about large deals with small numbers of customers, not large scale operations.
I don't think you can simultaneously argue that high government office doesn't count as experience "operating a large organization with millions of customers" and that such experience is an important qualification for competently executing such office.
But supposing for the sake of argument that only private sector experience counts, Bush the Younger should get partial credit for the Texas Rangers, which aren't a huge organization (I think MLB teams have on the order of a few hundred employees between the players, coaching staff, ground crew, and front-office staff) but do have millions of customers.
On the subject of Trump's business record, I would be one of the last people to defend it, except to note that it exists.
Being a politician isn't operational experience. It's oversight. And I don't draw a line between private and public sector. Someone who worked in the PR department of a private business also lacks operational experience. I would take someone who managed a local authority over those people.
Or is your point that public sector institutional cultures are so severely and pervasively broken that experience as an agency director, state governor, or senior senator is usually experience at doing things the wrong way with entirely the wrong mindset?
If so, I'm more sympathetic to that line of argument even if I don't think I'm willing to endorse it entirely.
No. I've worked in parts of the public sector for some very good people, people who have operational experience. They would make better politicians because they understand what works and what doesn't and how to fix things better than politicians do.
They mostly know how to fix things using tools that aren't available to politicians. Their track record when they become politicians, is mediocre at best.
The State of Arkansas isn't a large organization with millions of "customers"? Not even Texas?
Not buying it.
Also worth noticing that Biden and Bush the Elder both spent eight years as literally the President's understudy, which should be about the best possible apprenticeship for that job.
100% a huge problem with the bureaucracy is the laws themselves which are passed, which sometimes literally don't make sense or are internally inconsistent.
Some of the COVID relief was like this. Where it was like "here is a big pile of money (sometimes billions) and a million rules for doling it out, and it needs to be out like in a couple months". Like brand new programs that are multiple billions and they are meant to be up and fully functional in say 60 days. So what happens is just LOTS of rules get ignored and fraud protections skipped. Because something has got to give.
If I had to guess, I'd say Vivek believes that bureaucrats can actually do twenty forms per year, but are only doing ten because there's no accountability and they're basically impossible to fire.
I guess there's nothing stopping bureaucrats from refusing to do more than ten forms per year even if their workforce is cut in half. You'd have to reform in other places.
This. The problem for Vivek is that they will simply refuse even if they could work more, and since Vivek volunteered to be the scapegoat through his mass layoffs he will have to take the blame. It's about incentives.
"The Washington Monument syndrome,[1] also known as the Mount Rushmore syndrome[2] or the firemen first principle,[3][4] is a term used to describe the phenomenon of government agencies in the United States cutting the most visible or appreciated service provided by the government when faced with budget cuts"
Ah, I see this gambit works the same the world over. In the UK, if it had a name, it would probably be called the Trooping of the Colour syndrome:
When the Defence Ministry is told to make yet more cuts, the military often responds that the Trooping of the Colour ceremony must be cancelled this year, as there simply won't be the manpower for it! Cue press outrage, and the cuts are reduced.
(Trooping the Colour is an annual ceremony performed in London, in which regimental colours are paraded before the troops of a household regiment, in the sovereign's presence. It's original purpose was to remind the troops of the colour flag's appearance, so they would recognise it in battle. But these days the ceremony is no more than a popular tourist attraction.)
Another trick used by Home Office staff in passive aggressive efforts to resist orders to crack down on illegal immigration is to scour the length and breadth of the country until they find a blond, blue eyed, Hitler youth looking young man to publicly eject from the country because his parents' visas were slightly out of date when they arrived from Australia or South Africa twenty years previously, or were missing a vital date stamp or comma!
Not without an extraordinary amount of effort, which will not be sustained across the Federal civil service. The government uses pretty much the exact opposite of "at will" employment. I've seen exactly one civil servant fired for being so lazy that he basically didn't do any useful work, and it took a solid year of his manager's focused effort documenting the laziness and arguing the case with the bureaucracy.
Apparently, efforts to analyze and optimize U.S. Federal bureaucracy is one of the largest sinks of time any money in the U.S. Govt. Mandates keep coming down requiring more and more documentation and analysis of cost and time. In all honesty, the same is true in my multi-billion dollar company (we are sort-of a startup). We have to get multiple quotes even to spend $1000. We probably end up wasting $5k in salary to protect against an excessive $1k quote.
I wouldn't treat # of bureaucrats and red tape as independent. The more bureaucrats there are, the more red tape the system can support. If you drastically cut staff, the first order effect is these orgs will be massively backlogged. The second order effect could be to figure out how to re-tape everything to work with reduced staff.
The right way to do this would be a first principles approach where you figure out the regulatory scope and sizing to achieve some goal. But as it stands, the regulatory offices operate on blind trust and are mostly immune to electoral politics. So I think there's a safe presumption that they are way overgrown and you can blindly downsize them by huge percentages. Short term breakage is preferred to some 5 year transition plan that somehow leaves everything larger and more convoluted than before.
Scott identified the problem in the article, which is that government bureaucracies aren't doing things for the fun of it, they are doing the things that the legislators told them to do. (Writing annual reports on non-clinical drug testing, analyzing all the kinds of pthaletes(sp?) and deciding which are poisonous, that sort of thing.) Fire half the people, and the remaining ones are still are required by law to write that annual report and analyze the phthaletes.
So the solution rests with Congress reducing what the bureaucracies are required to do.
Is that really in the actual laws? Or is it just an outgrowth of what the policy folks have decided to interpret the laws as allowing them to get involved in?
Doing a naive text search of the entire US code section 21 (the Food and Drugs section) reveals no uses of the word "phthalates", for example. And the only two uses of "non-clinical" appear in text saying that people who want exemptions for rare diseases can ask for recommendations as to what they should do to test them. Expanding that to "we have full control over non-clinical testing" is, well, not something you can blame on Congress. Except to the degree that Congress doesn't actually write specific rules and just says "yeah, the secretary will make rules on what this means."
If the report on phthalates contains ten pages of core functionality and 990 pages of "yes we thought of that, please don't sue" ass-covering, maybe cutting the workforce in half will mean they put out the same ten pages of core functionality and only three or four hundred pages of ass-covering... without actually getting themselves sued any more often, nor to meaningfully greater effect, because the resources available *to the people who'd be inclined to sue them* haven't actually increased.
Go read about the judicial history of the National Environmental Policy Act and report back on your findings and whether you think this would still work.
I think the key theoretical tool missing here is the distinction between line and staff.
In the example here of just filling out forms towards a 1000-form goal to approve a drug, that's clearly some kind of line function for the bureaucracy. They do the main frontline productive work in terms of the demands on the bureacratic institution, like the FDA. I think most criticisms here hold true for the line -- you better think about why they need to fill out those forms to begin with, etc.
But I think people tend to instead think of the bureaucracy as non-line, staff functions. Especially if you're in the private sector, where that's basically what you mean by bureaucratic overheads. And then they generalise that to the entirety of public bureaucracies, since they sort of parallell a staff function for society-at-large.
Not only do these staff functions easily bloat, capture leadership etc way beyond what's justified, but also impose even more work on the line, as a sort of bureaucracy-within-bureacracy. Please fill out this form. Please attend this seminar. Even something like IT does this: a lot of new IT systems offload work on the line IKEA-style (instead of a supporting admin staff doing meeting room booking, you have to do it yourself via a nifty system).
Recent years' discussion on DEI and HR departments is the culture war spearhead of this, but you can also discuss, idk, agencies' individual communicatins departments and so on.
You still could probably not just fire half of staff function personnel at random, there are still externally imposed limits and so on. But I think to a first aproximation "bureaucracy as bloated staff numbers" holds true for them in a way it doesn't for line, and this post seems to assume all public bureaucracy functions as line staff.
Obviously Vivek is being hyperbolic but let's take a real example of a country that is doing something similar to what DOGE wants to achieve.
In roughly a year, Milei's government has cut 7% of the government workforce, roughly 30,000 public sector jobs.
Is the country falling apart? In my personal experience, having spent 23 days in Argentina (leaving as I write this comment), the country seems to be doing OK.
A country like Argentina has a huge amount of people in unprofitable SOEs and obviously make-work government boards that it’s easy to find places to trim the fat. They aren’t exactly there to process millions of regulations.
In the US bureaucracy is huge because it’s essentially a massively parallel computer. The biggest civilian employers—over half of the workforce—are the VA, HHS, Dept. of the Army, Navy, Air Force and DoD. Cutting these jobs doesn’t decrease complexity. It only reduces processing speed.
Most of the federal government is military or homeland security functions . I'm not sure Argentina has to worry about this stuff too much. Also, doesn't your point kind of disprove itself somewhat? Isn't the US doing OK?
Vivek and Musk have experience in the private sector, where you cut jobs first and then the remainder figure out how to work more efficiently because their bosses are driving for increased productivity, with compensation structures aligned accordingly. The proper question is whether that works in the public sector, where as you point out some of the inefficiencies are due to externally imposed requirements and also where compensation structures are aligned with seniority, not efficiency. But it seems to me this is a better framework for understanding DOGE challenges.
Not addressed in the post but relevant: A lot of these processes and regulations were put in place to AVOID having "faceless bureaucrats" making decisions with high impact on the lives of citizens. So if you want to start turning 1000 forms into 100, you're going to have to grant a lot more discretion to faceless bureaucrats.
(I realize that "passing laws" and "deciding things" aren't the same thing, but they're both types of power and authority. If the folks who don't want them passings laws are fine, actually, with giving them more discretion then they currently have, that'd be great, but that's what I assume they want.)
I think this is really important. Almost all formal regulations are put in place to constrain the discretionary power of the organization. If you could trust people to make the right decisions based on a short statement of values, there would be no need for detailed rules explaining how things need to be done. Behind the right wing desire to cut bureaucracy is really the desire to take power away from the organizations that make up the state. But if Congress insists on having regulatory organizations like the FDA, FAA, etc., then there will be a push to write tons of regulations to prescribe exactly how the organization should carry out its mission.
IIRC the Republicans have appointed more judges than Democrats in recent years, so they have an incentive to move decision-making power from the bureaucracy to the courts.
I think that might actually be a good thing. With the Progress Studies folks, we've seen a lot of nostalgia for how quickly stuff could get built before the 1970s - that was back when the regulatory process could consist of "1. Go to bureaucrat in X office to get approval; 2. They grant approval after review; 3. Project goes forward".
The closest thing we've had to that in recent years was when the FCC was given the power to tell obstructionist local governments doing NIMBY fear-mongering about cell phone towers to pound sound - "put up the tower, and we'll give you some leeway in where it can goes as long as it gets put up on schedule".
But on the other hand the bureaucrat in question was often Robert Mose, and the stuff that got built was horrendous city-ruining garbage in many cases.
Or simply have the officials involved in fewer things. If the FDA was a pure research organization, instead of an approval, research, enforcement, and rulemaking organization, you'd need a lot fewer people and it wouldn't matter so much
They actually often have a lot of scope for discretion but don't want to use it because fucking up some law badly and making a big mess is about the only way they could ever lose their job.
Ramaswamy's approach to bureaucracy seems to be based on Parkinson's Law formulated in the 1950-ies. Citing the Wikipedea, it states that "work expands so as to fill the time available for its completion", and "the number of workers within public administration, bureaucracy or officialdom tends to grow, regardless of the amount of work to be done. This was attributed mainly to two factors: that officials want subordinates, not rivals, and that officials make work for each other." The evidence includes, in particular, "increase in the number of employees at the Colonial Office while the British Empire declined (he showed that it had its greatest number of staff when it was folded into the Foreign Office due to a lack of colonies to administer)." So it least in some instances cutting down the number of government employees is justifiable.
Sure, but the US government now has more functions and fewer civil servants per capita than 50 years ago. Seems like removing civil servants won't address the issue right now.
I've always been in the incentives camp. My wife's dad had a horrible experience working for the government: He just simply worked for 8 hours and did all the applications that he managed. But that's wrong, you see! Everyone else did some low fixed number of applications per day, so they got angry at him for making them look bad. In the end, he spend the first 3 hours or so every day doing the work he was supposed to do, and spend the rest doing whatever. I've heard similar reports from multiple people who worked at the government, and to a lesser degree in very large legacy companies. People just simply have no reason to work hard *OR* efficiently.
As I said, especially large companies in a field with significant moat can exhibit some of the same tendencies, since the incentives can be likewise screwed up.
>Even in my contractor days If I worked too efficiently with time and materials, I would get lowballed in future negotiations.
That doesn't even make sense, though; You're not competing with some imaginary version of yourself that always works exactly as hard and as efficiently as yourself, you're competing with other contractors. If you work harder and more efficiently than them, you can still ask for the money other contractors demand and take the difference for yourself. If they lowball you, the correct response is "F U" and go work for someone else.
Which is the reason why contractors generally have the opposite problem; They tend to overwork themselves to keep up. Which incidentally is the experience every single self-employed I know. With one exception: Working for the government or other monopolies as a contractor. But that's again the same kind of market/incentive failure I'm talking about.
Your customers remember what you did, and bargaining isn't logical.
In situations where things are time sensitive getting it done fast might be worth extra; my experience was emergencies get done on time and materials, non emergencies get done on contract and coming in too fast and too efficient is bad, though I will note the worst of it was from working for giant corporations.
They would see that someone was eg. doing PM on a piece of equipment in 3 hours instead off four, note that down, and next time the contract came around they would itemize the hours and fight for a specific flat rate.
Basically, it was never worth it to not quote a fat margin and then spend a couple hours holding the walls up, instead of quoting a thin margin and getting in and out.
This was replicated across all the major companies we worked for in this field, excepting one which is known for not being particularly capitalistic internally and being family owned/operated, even though they are pretty big.
After waiting all day, the clerk said we'd already lost as soon as we got in line at the start of the day because we haven't made an appointment. Hearing the crowd's grumbling since the soonest appointment is three months away, she said "you gotta call the phone number, not use the website."
I get home and call the phone number. They tell us "we can't do anything here you can't do on the website." The in-person clerk didn't care, but she got the crowd out of there thinking they had a solution.
A big part of the issue is being underfunded and we *do* need more workers ie more bureaucrats. Probably at higher pay because at minimum wage you only get people who decide part of their job satisfaction is telling people who've been waiting all day they get nothing.
In my experiences with the DMV in the US, they should be exhibit A for the proposition that work expands to meet the number of people available to do it.
My understanding is that DMVs in the US insist on an in-person meeting to do all sorts of things which could easily be online forms.
dmv commenta are weird to me because in the two states I've had to deal with them, I've never had a bad experience. even at the super busy location I did my first driving test, they had a long line... that was moving quickly. no more inconvenient than shopping at costco.
Seriously. My DMV is incredibly efficient, friendly, and fast. You do most everything online, and if you need to go in you make an appt online, walk in, and you're done in 10 minutes. It works substantially better than many private sector services.
This is the case now in Illinois. Our DMV used to be no less rage-inducing than is depicted in popular media but it is now a miracle of efficiency at least for residents doing routine stuff. I know several people who've had that experience (am married to one of them).
I myself had to go do a not-routine thing late last year because of stupidly allowing my drivers license to expire. Back in the day that would have meant a long line and re-taking some old tests and etc so that's the ordeal I was prepared for.
But in fact it was smooth and simple -- even with some (understandable) proving-who-you-are steps I was in there no more than 30 minutes total. Also no civil servant gave me a fish-eye for being a sloppy knucklehead which years ago they would have....how and when that change happened I have no specific idea. Must have been during the past 15-20 years though just based on my own experiences.
I always hear about Americans queuing for hours at the DMV, but what do you actually do there? Do you have to go in person to register a new car or something? The only time I ever had to go to our equivalent of the DMV was when I imported a foreign car.
I think it's fairly exaggerated - I've been to the DMV maybe a dozen times in the decade and a half I've been driving, and it's never been a terribly long wait - often very fast, maybe 15 minutes - I don't think I've ever waited close to an hour.
Getting a license, *maybe* renewing a license (though I think I can do it online), getting a new license if you move across state boundaries (probably one of the more unique 'American' aspects). Last time I was there was updating the title on my car to add my wife and I was in there for maybe 20 minutes total and half of that was the actual paperwork.
Probably would be worse if I lived like... downtown New York City (which is hugely overrepresented in media portrayals), but I don't know, sometimes city bureaucracy is more efficient, sometimes its not.
I can do a lot on line, but certain key things, like getting someone their first-time driving test or ID, need to be done in person.
You need a state ID to vote, even if not driving, that you get from the DMV, and not having appointments is a way of low-key disenfranchisement and then saying "whaaaaaat, why can't you get an ID???" (And I'm generally in favor of voter ID! But this is some bullshit.)
That varies by state. California, for example, works on the honor system. The more binding requirement is that a lot of the banking system requires ID to open accounts, and that you need to show proof of identity and citizenship (or visa status) for most non-self-employed jobs.
Off the top of my head, I've been to get my learner's permit and first driver's license, when signing up for a new license when moving to a different state, once for registering a private-party car sale, once to sign up for a RealID Act compliant license (*), and once to process a legal name and gender marker change.
(*) A federal law that went into effect recently that set higher uniform federal standards for state-issued IDs to be usable for federal purposes (mainly for air travel and entering secured federal facilities)
Routine renewals and in-state address changes can usually be done online or by mail. And if you buy a car from or sell a car to a dealer, they are almost always the ones responsible for DMV filings.
My experiences with DMVs varies wildly. The offices I've been to in major suburban areas in California (one in an Orange County suburb of Los Angeles and another in a Bay Area suburb) were every bit as awful as reputed, mainly because they were crowded and understaffed. The one I went to in a Seattle suburb an another in a semi-rural exurb south of San Jose were much better, similar to Retsam's experience.
My reading was that Ramaswamy was just appealing to voters who hate bureaucrats and want to see them suffer.
To be precise: To the voters who don't hate bureaucrats because of the red tape, but because they see them as a nobility of a kind, because they are not bureaucrats themselves and their kids are not going to become bureaucrats either.
A couple thoughts to add as someone who a) is in the process of building a residential home and b) spends much of my time working with large organizations, trying to help them "get things done".
On a) - if you have not built a home, it is an informative exercise that is an exception to your comments above. There are numerous regulatory bodies who have a job to "say yes", but the degree of rigor they apply to the many, many regulations is absolutely directly related to the number of staff / time they have. The more people & time they have, the more it slows down the approval process.
I think you're thinking of regulations (the "checklist of 100 things for the FDA") as a firm set of guidelines with clear rules & each one represents a fixed unit of work that must be completed. That is almost never the case - the rules are most often vague & have multiple interpretations & level of rigor. A town with 1 person will check for egregious violations and say "proceed". A town with 10 people can easily fill 4 months checking every nuance, multiple reviews, multiple internal meetings, etc. (See: HOAs for a multiplier effect on this.)
On b) - there are massive coordination costs that happen with large & complex organizations that basically grind them to a halt, even if they have clear objectives & vision. This is why organizations are constantly going through exercises to reduce layers & manage spans of control.
You could remove half of most large corporate organizations and still maintain the same output. Are government bureaucracies more or less efficient than the average corporation?
HOAs, with which I have a lot of personal experience both as a member and as a board member, are actually the opposite problem: they have _zero_ salaried professionals whose job it is to review and approve/disapprove proposed actions.
The rule-enforcing power of HOA boards is effectively absolute so long as they stick to enforcing their own rules. HOA boards do occasionally stray outside their remit and get into real trouble but that is much rarer in fact than in legend. Sticking with the 99.9 percent of cases in which a member simply dislikes how, or how quickly, the board decided on a matter falling within the HOA's responsibilities, the member is without practical recourse.
HOA board members are unpaid volunteers; there is no knowledge or experiential or educational qualification for election; being an HOA board member is a thankless role which quickly feels for a non-retired person like having a second fulltime job. Those realities mean that in most HOAs most of the time elections aren't even contested, in fact some HOAs struggle to find enough warm bodies willing to take a turn on the board.
So: unappealable decision authority lodged with a small group of un-fireable volunteers whose only qualification for that responsibility was being the only people egotistical enough or naive enough to say yes. Hey what could go wrong....
Well lots of stuff, as everyone who's ever served on or gotten crosswise of an HOA board knows too well. It is though a _different_ malign logic tree from that of big bureaucracies.
Vivek's argument might be not that halving the number of federal workers will somehow remove red tape, but more that it will save money and have no downside because a lot of people in the federal bureaucracy aren't doing anything useful at all. Hence his 'nothing will break' assertion. His model isn't that if you lay off half the form-fillers the others will start filling forms twice as fast, it's that ~half of the people working in those offices aren't even filling out the forms and are either useless or actively counterproductive to the mission of the organization. I think he's wrong (I don't imagine the FDA staff is 50% DEI consultants) but I think that's what he's arguing.
Agreed. If you added up all the people who are doing nothing and are counterproductive, what percent of the workforce would that be? Vivek is saying something along the lines of "a clear majority." Such that you could cut a large number of random people and, on average, not reduce efficiency. If those numbers were a minority, cutting half randomly would hit a lot of productive people. If it's literally everybody (and for some organizations this is possible, especially if you fundamentally disagree with that organization's goals) then you could cut at or near 100% and only see upside. Republicans have talked about removing the Department of Education for years. At this point it's an organization that hands out some grant funding and tells local schools what to do. You can cut the grants (and therefore reduce taxes?) or move them to another organization pretty easily. The rest is even easier to remove because Republicans really don't agree with the rest of the DOE's purpose.
To clarify, the grants are a big deal to public school budgets, but they don't necessarily have to be. State taxes could cover the deficit, paid for by a reduction in federal taxes. Or the federal government could just continue the grants with less direction and oversight so that schools have their budgets maintained.
He can't be arguing that half the people in those jobs are doing nothing while the other half do the work, because he wants to fire people at random, not based on performance.
Then a marginally more-qualified DEI manager with an odd SSI# sues you, and Vivek pulls a https://www.giantitp.com/comics/oots1266.html , possibly after publicly flipping a coin, so everyone in your department qualified to run the defense is out.
“I believe very strongly that people on the left are too prone to do things that are emotionally satisfying and not politically useful. I have a rule, and it’s true of Occupy, it’s true of the gay-rights movement: If you care deeply about a cause, and you are engaged in an activity on behalf of that cause that is great fun and makes you feel good and warm and enthusiastic, you’re probably not helping, because you’re out there with your friends, and political work is much tougher and harder." - Barney Frank
I would absolutely love an administration to come in and significantly shrink the bureaucracy in an intelligent way, but because the task is so immense it would take a lot of sober, grueling work that I don't see people in this administration being willing to do. They're focused on publicity, not work, and maybe there's some gain from propagating the idea of large government downsizing, but Ramaswamy's tweet shows they're not serious. I hope they prove me wrong.
Somewhat related: given that immigration seems to be a top priority, I sure hope they don't apply this scorched earth method to bureaucrats handling immigration processing. Wait times are already insane, cruel, and probably illegal.
I don't think Vivek is right about half, but he is right about at least 20%. And we know this is true because the bureaucrats themselves have said so when trying to angle for 4 day work weeks.
Several studies arguing for the 4-day week have shown that productivity is not reduced by reducing the working days by 20% (almost only in regards to government work oddly). So it stands to reason that reducing the workforce by 20% will have the same negligible effect.
Part of the reason people get as much done in less time is because of inefficiencies caused by being tired and burnt out. Most humans work more effectively if they're not overworked. Reducing staff wouldn't have the same effect.
I work in navigating federal bureaucracy (pesticides instead of drugs) and this matches my experience pretty closely. The agency has been steadily losing staff over the last few years, and the result is actually worse than Scott's toy model would predict because approving applications is only, say, 75% of what the program does. The other 25% (or more) comprises statutory obligations like periodically reviewing approved pesticide actives, figuring out how to do (and then doing) Endangered Species Act assessments, etc.
No one is going to sue the agency over not approving pesticides fast enough, but they will absolutely get sued over the Endangered Species Act. So if you cut the number of bureaucrats in half, that entire 50% reduction comes out of approving applications, which now take 3x as long to process, not 2x.
Also: non-clinical/non-animal methods are an issue in pesticides too, and my understanding is that this is largely driven by NGOs/public pressure and, to some extent, companies wanting to be able to say they don't test on animals. Thing is, replacing animal methods is hard! Biological systems are complicated, and you don't want your in vitro test system to miss some adverse effect that would show in a test animal.
I should add that there is absolutely a problem where idle (or not sufficiently overworked) bureaucrats will find new things to worry about that almost certainly are not contributing to the public good. Still, I'm inclined to think this is more a problem of legally mandated priorities and institutional culture than number of bureaucrats per se. We live in a large and complex society, bureaucracy is probably necessary to maintaining that, and keeping the bureaucracy well tuned is, apparently, not trivial.
I do a lot of defense of bureaucracy below, so I will say, this is also true. It can be really hard to tell from the outside, but I'm 80% sure that another office I'm aware of has this exact problem, or a related one that's feeding on itself. The office manager insists on reviewing all final work himself, so there's a giant bottleneck, which he tries to fix by hiring more people in his office to feed stuff into the bottleneck.
But they don't have enough work, so they respond by going looking for problems, or brainstorming up nightmare scenarios, or doing vast amounts of research on piddly questions that should be answered in five minutes.
I said it below and I'll say it again, federal hiring/discipline/firing is a giant mess. I have no confidence in DOGE to fix it, but if they somehow manage it, I'll be glad. My fear is that (to the extent they accomplish anything) they'll simply make it easier to fire/discipline for political reasons, which probably won't be great. Whether it'll be worse than the status quo?
Bureaucrat here (another country): This is such an excellent article that very succinctly expresses what I've been trying to explain to my non-bureaucrat friends. Swap drug approvals for anything else (let's randomly say procurement!) and it's the same issue.
The idea of cutting people is easy, the reality of changing the rules is hard (auto-sunsetting clauses are a great idea though).
This makes me lean towards the idea that the DOGE goal is fundamentally to break things further so as to have it all collapse and I guess start over somehow.
I'll say, if you want auto-sunsetting, you need to combine that with streamlining the process of creating new rules, otherwise you're likely to end up with a situation where you don't know how to issue required approvals--because the autosunsetting I've seen only hits implementation regs, not the statutes they're trying to implement.
Oops, I'll note another issue with auto-sunsetting--it most likely restarts the clock on challenges to regulations. Because you're always reissuing within the statute of limitations. So, to take a streamlining example, NEPA authorizes agencies to identify categories of actions which are excluded from NEPA due to their low potential to cause effects. That's done by regulation. I am unaware of any new CATEX's identified in the last 20+ years (note, doesn't mean it hasn't happened, just that I'm unaware) because such a change would be open to litigation, but the old CATEX's aren't, because the statute of limitations has generally passed (note, actions can still be challenged as not properly falling within, but that's a different issue).
If you have to re-issue every five years, that means folks get another crack at suing every five years.
Yes, think you're right here. Auto-sunsetting is functionally unworkable with the way courts interpret the Administrative Procedure Act. Depending on the sunsetting period, you could probably federal government is unable to issue any rules whatsoever or spends so much time trying to issue and renew rules that it can't do anything else.
The only place I would challenge the reasoning here is the density of actual work per bureaucrat. That is, if government employees are anything like typical Corporate America employees (which they often are), they spend half or even most of their time sitting around doing nothing, or waiting-as-an-activity.
It's not like government bureaucrats are at their desk working on one single form, nonstop, for 45 hours per week, getting cramps in their writing hand. Most of the time it takes to accomplish anything in bureaucracy isn't the work itself - it's when you respond to an email and then wait. It's when you submit a form and then wait. It's when you do your little tiny task to advance the process and then send the item back into the machine and then wait.
So although I agree with your point that bureaucracy isn't "measured in bureaucrats," I don't think it's true that cutting bureaucrats would actually slow all things down. In fact I don't even think it would slow most things down. The time loss of bureaucracy isn't in paperwork being done by hand - it's in paperwork sitting in cyberspace until someone gets around to it. And that can be made more efficient by un-stupidifying the system, and isn't mutually exclusive with cutting people out of the system.
It reminds me of the Kissinger's madman tactic to encourage negotiation.
"Listen, I know you want to negotiate a solution that will work for everyone, but my boss (Nixon), he's a madman! He's got his finger on the nuclear button, and I don't have control over what he does..." thus incentivizing negotiation.
Maybe there's someone at doge who will be saying
"Listen, I know we both want to make the FDA more efficient using some reasonable reforms, but my boss (Musk), he's a madman! He wants to fire half your staff, completely randomly, and then expects you to be twice as fast with approvals..." thus incentivizing cooperation.
The goal isn't to actually use the nuclear button, but to use the threat of it to get a better negotiated settlement. Maybe the over looming threat of a 50% reduction in FDA staff is enough of a threat, and credible enough of a threat, that the FDA will be very keen on working something out with the new administration that just makes them more efficient, rather than something that destroys their whole system. I'm sure there's a lot the FDA, and other bureaucracies for that matter, can do to make reforms ineffective, but if they're eager to work with the new administration, to avoid the nuclear option, then maybe the reforms will be a lot more effective than they otherwise would have been without the threat.
I think there's something to this idea. Whether it's explicitly what Musk/Ramaswamy is thinking or not, I'm sure whatever operatives DOGE end up having will naturally fall into that rythm. Even the softer form, "look, I'm on your side here, but my bosses would kill me if I accepted this deal" is the kind of classical negotiating tactic that can work even if everyone sees what you're doing.
In a way, this mirrors the debate about Trump's tariffs as a negotiating tactic vs actually probable policy outcome. I'm not sure what to think about that one, though.
...But the problem isn't that the FDA isn't efficient, it's that they have employees that cost money to employ. Even if they were more efficient, the department would cost just as much money to run.
Same amount of people, doing twice the tasks either greatly improves the throughput of the FDA, making the bureaucracy run more efficiently and slow commerce down less (which is good) or the higher productivity per employee allows for the system to function, while laying off half the workforce (which saves money).
As I understand it, DOGE is not about saving money by firing a bunch of federal employees, as even if you could literally fire half the bureaucracy, you still wouldn't make a meaningful dent in the deficit as it's a small proportion of overall government spending (the majority being welfare, healthcare and the military). The intent is to improve the efficiency of commerce, so projects and economic activity that would otherwise be unjustifiable (try building a new highrise when you have a multi-year long environmental review), become justifiable again. The increased economic output grows the GDP, decreasing the deficit & debt in nominal terms, grows tax revenue, decreasing the deficit in real terms, and slightly reduces spending the bureaucracy (which isn't relatively high), which decreases spending in real terms.
If the plan is just " Let's cut the number of bureaucrats in half! That will fix the deficit!" it is extremely stupid and there's literally no possible way it will work (like trying to afford your mortgage by going out to eat once instead of twice a month). The more complex argument at least has enough uncertainty and plausibility that we can say it might have a real impacy.
Why do you believe that FDA bureaucrats are going to agree to work twice as hard suddenly for less pay than their private sector peers? If you have some ideas on how this is a sustainable or implementable solution without losing the entire workforce with outside options, I'd love to hear about it.
Because we either cut down the number of tasks they need to do for a given level of throughput (perhaps digitizing or simplifying many forms), or otherwise increase their efficiency.
The key word is "efficiency" not "working hard". I'm not an expert on the topic, but from my general knowledge of government bureaucracy, there are very many tasks that are pointlessly complicated.
For example, I recently ordered a form from a government agency (something extremely simple, basically just confirming that a bit of information is still true in state records), and the only option to receive that form is via mail. They don't do email or even fax. As a result my business is stuck waiting for this form (which is a precondition for applying for another government form), and we're stuck waiting a week for the USPS to deliver a letter. That's one week where we aren't starting this new project, a week of lost profits, which compounds across millions of companies all waiting their individual weeks. (The form also cost $125 + $100 to ensure it takes less than 7 days to arrive).
This could be digitized by a mediocre software engineer in literally a week. Maybe the cybersecurity requirements would make this take longer, but digitizing this form would both save the manual work needed to actually send it, and the customer support I called to confirm that the mail was sent. That would save thousands of man-hours per year, which could either justify an increased throughput, or decreased payroll.
As a long-serving government bureaucrat (but military, so I'm different, I guess?), the problems I see are:
1. Regulations that are sensible individually but ludicrous when combined. This tends to be cyclical - politicians complain about how long process X takes, so we cut some steps, then they complain about the output of process X, so we add some steps...
2. Regulations that have a higher cost than benefit, but the cost is diffuse while the benefit is clear. ("Measure twice, cut once" is fine 99% of the time, but 1% of the time measuring thrice would have been better. Those 1% failures can be spectacular, though, and can inspire calls to Do Something.)
3. Lack of incentive for effective implementation. In business, there is both upside and downside risk for most decisions. In government, there is generally only downside risk, so the default is to follow the safest course possible.
4) Regulations (or policy requirements that aren't formal regulations) that have been put in place instead of punishment (this could arguably be 1 or 2). The big one that jumps out to me every so often is conference policy, where the entire thing exists because one agency did a major fuckup in a quite corrupt way planning an extravagant conference for themselves and so every agency suddenly had a 'conference policy' that was usually so broadly written that you suddenly needed high ranking folks signing off on every training course anyone travelled to. Great use of everyone's time.
I'm hearing comments that Vivek's comment was obviously hyperbolic. I don't discount that as a possibility. But it may also have been a negotiating ploy to ask for more than is wanted. And... ...what fraction of Twitter's staff did Musk lay off?
In Luxembourg they had to enforce some new tax directives to prevent money laundering. and tax evasion. But since the country makes a lot of money out of quietly permitting those things, they did not really want to enforce them. In the end they hired three people to do it, so they were enforcing it on paper. But three people were nowhere near enough to actually enforce the rules, so they mostly went unenforced.
Depends on the law. For example, you can only sell pharmaceutical drugs with FDA approval. If you dissolved the FDA, that doesn't mean that individuals couldn't still sue you for selling them drugs without approval after claiming that it made them sick.
There's also lots of NEPA lawsuits against solar and wind farms that block the project because the environmental review wasn't through enough.
Really, a big part of the problem is that enforcement is pretty privatized in the US. So dissolving the bureaucracy doesn't absolve you from getting permission and you absolutely *will* get smacked down by private actors. Sometimes you'll get smacked down even *with* bureaucratic approval under the arguments that the bureaucracy was too lenient on you.
It used to be that nations just can't have enough tax enforcers to actually tax their people. It's actually the Hallmark of developed nation to be able to effectively tax more, either because of might or pure efficiency. In a way, if the superpower AI can make and enforce the ultimate tax regulation, it'll benefit everyone and we can fire 100% of the IRS.
The number of people needed to enforce taxes is a function of the complexity of the tax system for the most part.
A simple poll tax is simple, you get a bill and you either pay it or you don't. Much as I'd hate to wake up the Georgists here, a land value tax is pretty simple too.
All the complexity and difficulty comes in once you decide that income is the thing you're going to be taxing.
Yeah there are a lot of simple measures that are employed by pre-modern states. The obvious weakness of poll tax is that you have to tax the rich and the poor by the same measure, and have to cater for both. If you use the poor as standard, you miss out a lot on the rich. If you use the rich as standard, you'll bankrupt the poor. If you use the middle, you'll annoy both. This results in less effective taxation. More complexity may mean more tax income, but it can also backfire horribly
> Much as I'd hate to wake up the Georgists here, a land value tax is pretty simple too.
A land value tax is pretty complicated in terms of bureaucrats, since you have to value all the land. All the proposals I've seen by Georgists include "First, hire a ton of people to assess the land, and have them do it again every year".
It's not that red tape expands to consume available resources, it's that having plenty of resources does not incentivise finding more efficient ways to deal with red tape. Halve the number of employees, and suddenly people will stop using fax machines, and start using GPT more.
My understanding is that companies claiming "No Animal Testing" were using compounds that were previously tested on animals--say for example in China or India--and submitted based on the foreign acedemic study as "safe."
That may be an overly cynical view, but I can't imagine Lush's legal staff being comfortable with using their customers as test subjects.
Good points overall, but I would like to push back a bit on the claim that ban-based bureaucracies wouldn't benefit from a cut.
You mention that the crypto industry is demanding regulation, but this is only *after* multiple bureaucracies started opening up expensive lawsuits against crypto companies of all sizes, in a seemingly capricious way. Most companies settle out because they don't want to be bothered, but the cases that actually go to court tend to side with with defendants, and the judges in some of the cases are not too kind to the bureaucracies who are engaging in these capricious lawsuits against people and companies.
If you were to cut 100% of the SEC, crypto would likely rejoice because now most of the lawsuits against crypto companies would go away. If you cut 99%, there would still be some lawsuits, but the number would likely be low enough that the crypto industry would just absorb them for lower cost than dealing with regulatory red tape. If you cut 1% of the SEC, you still have basically the same number of lawsuits and it maybe takes them a little longer to actually create regulations.
Another thing to consider is that when companies "call for regulation", it is worth noting that this is *exactly* how capture happens. The little companies aren't demanding regulations, it is all of the giant companies who can hire a team of people to implement whatever regulatory requirements there are, and who mostly companies just looking to profit, not looking to enact some meaningful change in the world (true for biotech, crypto, social media, journalism, etc.)
Finally, specifically on the crypto side, this particular bureaucracy (SEC) seems to be particularly bad. Even people within the SEC have openly opined about how the SEC should provide clear rules and lay off on the enforcement until that is done, but the SEC leadership continues to not actually fulfill their mandate of protecting investors and instead just harasses the crypto industry. I suspect many enforcement based bureaucracies are formed of well meaning people trying to fulfill their mandate, but there are at least some that are run by people who don't care much about the mandate, and instead answer to some other master.
One could likely make similar arguments about the DEA or the ATF, depending on your political leaning. Both of these spend most of their resources on enforcement rather than rulemaking, and when regulation is "demanded" it is either by big companies, or so people can stop living in fear of the agency suddenly cracking down on them in a capricious way (see cannabis industry for example).
I agree that the vast majority of stuff in crypto is somewhere on the spectrum of scams. The projects the SEC has been targeting though seem to skew unreasonably heavily towards the few legitimate businesses/projects in the space, while they ignore the rampant scams that are happening.
I suspect the SEC's underlying theory is that there is no legitimate purpose to cryptocurrencies in general, so if they can successfully shut down the biggest ones, rest of it will just be a matter of mopping up.
I think it doesn't help the perception of crypto that all the scam blackmail emails require "send this amount of bitcoin to this address, if you don't know how to buy bitcoin look it up online".
That doesn't help crypto to look like it's legit, even though it's not Bitcoin's fault.
How is it not Bitcoin's fault? The exact qualities which make it such a great option for blackmailers, drug smugglers, and the like were a deliberate, integral part of the original design, and that outcome was reasonably foreseeable. Compare https://en.wikipedia.org/wiki/Attractive_nuisance_doctrine
I have the very cynical view that much of the request for "please regulate us" is so that the existing companies can get ahead of that, have the regulations written to favour them, and then they can stifle new competitors who might arise in future by dropping a ton of red tape on them.
Good points on the detail level - but with a birds eye view Elon and Vivek might have a point or two... My own observation in halve a century of life, universe and everything I identified a general law: Mass is attracting more mass - if its in the universe (it all ends in a giant black hole), life itself (cells clump together, build multi-cell organism, which build even larger conglomerates and so forth), bureaucracies, companies, human organisations in general - smaller things clump together and get bigger and bigger over time...
The rare situations a bigger entity is not getting larger are depletion of resources or explosion/war/death.
If you look at each individual example or situation further growth seems inevitable and all the counter arguments for "just cutting the cancer off" sound dull. But practice shows that more often then not someone is just ignoring all the evidence of keeping things as usual and just getting rid of it all. An guess what - "the cycle begins again" to cite my favorite singer/songwriter.
The past years I turned from a huge fan of Elon and his achievements into an outspoken critic. But we must admit - there is a huge track record of awesome stuff that worked besides all the nay-saying. And regarding "Firing halve of the workforce of an overblown org and good things will happen" I look at former and current Twitter (7.5k vs 2.8k, 1.5k in 2023)
But the big elephant in the room for me is this insane military budget. But when we try to learn from the Romans we clearly can see that the Imperium can easily grow some further years till it will implode finally.
Shared my view with ChatGPT and it has some interesting points and views to add:
So, I'll say, there are massive problems of government efficiency which need to be addressed, most obviously in procurement and hiring on the bureaucratic side, actually passing bills on the legislative side and remembering that no one actually fucking elected them to impose their preferred policy preferences on the judicial side. Eating Policy (https://www.eatingpolicy.com/) is good on this. I wish DOGE would indeed make government more efficient, but I am...not optimistic on their chances of success given their fairly clear lack of interest in how any of it actually works or the legislative action which would generally be needed to fix it.
Not that I'm a fan of trump nor musk no other oligarchs pushing their private interests down our public throats.
But - in such cases like public or private company internal bureaucracy growing like cancer getting everything to an halt I more and more get the impression that creative destruction might sometimes be the only way out. And either its a strong man not giving a shit on all the "concerns" - or its something even worse like a natural disaster, a war like all the ones we've seen the past decade or two or a civil war like in Ruanda were one group tried to erase the other one after long times of suppression and corruption. We've heard a lot of positive things about Ruanda since woman are now the majority in Politics, Government and Decision making. Similar to Germany after its had lost WWII and most man and infrastructure and woman had to take over. Very depressing as a man getting the impression this might be a actual pattern on earth.
Eh, I'm unconvinced by that argument, but I think the key point your missing is that most (though not all) of the red tape isn't being created by bureaucrats, it's being created by congress/executive action/court action.
Now, it's not the case that none is created by bureaucrats, there's quite a lot that is, and it drives me up the wall, my worst moments are when I'm looking at some regulation or policy document that imposes a bunch of bullshit on top of the bullshit that congress gave us, but, like, I'm guessing what DOGE will actually focus on is:
1) Federal contracting rules, especially as related to Musk companies. I don't know enough about this to comment on source of problems.
2) NEPA, especially as it relates to approving various things Musk wants to do. This is some congress, some executive action, a lot the courts and a tiny bit agencies (indeed, the main thing agencies have done by regulation is identify what is categorically excluded from NEPA...which oh boy do I not want sunsetted).
3) Federal hiring and firing--again, don't know enough, but I believe this is mostly legislative, a result of attempts to professionalize.
> (indeed, the main thing agencies have done by regulation is identify what is categorically excluded from NEPA...which oh boy do I not want sunsetted).
If the sunset clause just says things need to be periodically re-examined, stuff that's obviously worth keeping can get rubber-stamped every few years. If it's a mandate to "remove two for every one you add," might be able to squeeze out a few by consolidating them into broader categories, which then save future trouble by including things that never quite came up before but are, by reasonable standards, sufficiently similar.
So, the problem is that if you rubber stamp a renewal, you're going to get sued under the APA. There's a reason that agencies haven't added to their CATEX's (at least the ones I'm familiar with) because the moment you reopen those to litigation and they're no longer protected by their age, you're going to be stuck in litigation over them forever.
Even from an Irish perspective, if they could cut away a lot of the underbrush from contracting rules, it would really help. You do need to keep the balance between "this is a cosy way for me to funnel taxpayer money to my brother-in-law by hiring his company to do all our contracts" and "this imposes way too much burden on us since the situation here is different so these rules don't really apply".
Oh, it's terrible and then you get contract challenges on almost any big enough contract. To be fair, contract management is always going to be a major problem.
My favorite one, though it's not contract specific is the conversation I always have with new people where they say 'X is categorically exempted from NEPA, so we don't have to do anything, right?'
And I have to respond with hearty laughter and 'No, don't be silly, we still have to do a bunch of work to demonstrate that it's categorically exempted from NEPA.' It's not enough to be right, you need to have been provably, with existing documentation, right at the time you made the decision. Which is fine for 'we're spending 5 million dollars rebuilding a road' and insane for, we're spending 50K doing tree trimming on our parks.
"It's not enough to be right, you need to have been provably, with existing documentation, right at the time you made the decision."
Tell me about it. Regulations for the sake of regulations, in order to CYA from any possible lawsuits.
There's to-and-fro here in the comments about "is it beneficial to have a litigious society or not?" and I do think people are forgetting that "They could build this stuff in the 1950s, why can't we?" is also because back in the 1950s, there were way fewer activist groups and ordinary people didn't immediately reach for the phone to contact a lawyer about everything. In 1950 if little Johnny falls off the swings in the public park and hurts himself, even up to breaking a bone, well that's all part of life and growing up. Today, you bet the parents will sue the park, the manufacturer of the swings, and anyone else they can think of because it shouldn't have happened.
Thing is, you say the military budget is insane and that is what should be trimmed. (I broadly agree, because of course there is bloat and featherbedding there). But another commenter wants to trim the bureaucracy so that the one important element of government can be preserved - which is the military.
So even there, arguing over "we should cut defence spending in half/no, we should increase it to twice the level!" is going to be the problem over "what gets trimmed where from whom?" for reducing the bureaucracy.
> Maybe he is working off a thesis where red tape expands to consume the resources available to it (as measured in bureaucrats)
One argument for this is that countries with a fiftieth the population of the US (like, say, Paraguay or Singapore), which presumably have approximately a fiftieth the bureaucrats, still end up issuing approvals for things on, more or less, a similar timeframe (usually faster, the US is unusually slow at things). There are some economies - smaller countries presumably have fewer individual projects to approve for things like new factories, and mooch off the FDA for things like drug approval (e.g. some governments might just have one guy whose job it is to read the FDA approval notice and go "yeah, looks solid"). But these are still far from linear, and yet we don't see approvals for things in Paraguay or New Zealand take fifty or even ten ties as long as they do in the US.
So, the issue is 'what are the requirements for approval' and though bureaucrats (hi!) can and do create some of those requirements, most are made by congress. We're a rich country, so we choose to place more requirements on approvals than most other countries. The obvious flaw with your size argument is that increasing population increases the number of approvals requested, not just the number of bureaucrats. For the FDA this is only somewhat true, but for, say the Clean Air Act, or Clean Water Act, as discussed below, the number of approvals grows pretty much linearly with population/land/land ownership.
The easiest fix would be to increase bureaucratic discretion across the board, including judicial discretion, that allows every bureaucrat to unilaterally bypass forms and red tape, first in exceptional cases, then in more cases as bureaucratic efficiency increases. The problem is that legislators don’t like to give executive and judicial discretion because it cedes power. Legislators are the problem. https://wwnorton.com/books/fewer-rules-better-people
The primary driver of inefficiency is decades of lawmaking requiring the civil service to abide by a variety of rules (The Federal Acquisitions Regulation, National Environmental Protection Act, Administrative Procedure Act) and the removal of bureaucratic discretion within agencies from "street-level" bureaucrats. Trump has minimal control over this, and if he replaces the bureaucrats with party loyalists, it'll create an even worse problem. First, partly loyalists are just bureaucrats who are taking a different top-down mandate, now from a single executive, and the same problems arise; they have to check a decision against a mandate and will be unable to break from the mandate when a situation requires it. But secondly, the best bureaucrats are the ones who do not blatantly violate the laws, but ones who know how to bend them in particular places so as to get a job done while avoiding a lawsuit. The Trump-style loyalists do the opposite; they openly and brazenly announce their violations, and that makes their decisions subject to litigation that ties things up for months or years. That's not efficiency. Unfortunately, this is mostly a legislative problem which require legislative solutions.
I still don't see why they can't just... ignore the litigation. There's no reason they have to cooperate with leftist courts, just like the US doesn't need to let the Hague try them for warcrimes.
> If you don’t, you end up like Ramaswamy, who seems to think that halving the number of bureaucrats will halve the number of forms that need to be filled out. I think in his worldview, the FDA will think “Now that we have fewer bureaucrats, it would take forever to complete our current process, so let’s simplify the process.”
I read it a little bit differently. Much of the DOGE effort is focused on budgetary issues, and trying to find ways to cut the deficit. If you fire half the bureaucracy on day 1, you're saving a whole lot of money from the very start.
Meanwhile, promulgating new rules to repeal old rules (the formal process by which "the number of forms" gets reduced) is a long, laborious process that can take months. (Fixing *that* is up to Congress.) This is likely to also happen, but it's more of a medium-term goal. Cutting costs is a short-term thing that can happen on day 1.
And if the rumors on remote workers are even partially true, they don’t even need to fire anyone. Just require 90% of workers to come to the office 5 days a week, and a significant number will quit.
TBH I think that's a bad idea for the same reason I think private-sector RTO mandates with the same goal are a bad idea: it's blatantly illegal. Creating unpleasant working conditions to push someone into quitting so you won't have to fire them is known as Constructive Dismissal, and it's a good way to get sued.
It'll be interesting to see whether you could get a five-day workweek interpreted as "unpleasant working conditions." I would be surprised, but I guess it depends on what kind of jury you get.
Remote workers work a 5-day week too. I'm one of them. My entire team is. But forcing the extra expense and stress of a commute on people who are accustomed to working without one — especially in areas that get winter! — is clearly unpleasant. Especially when, as Swami suggests above and some businesses have been distressingly clear about, the RTO mandate is being done with the by-design intention of making things unpleasant and causing workers to quit. Attempting to willfully commit Constructive Dismissal with a clear mens rea is likely to be pretty persuasive to a jury, even if they don't personally find the means by which it's being committed to be all that unpleasant.
Yeah, yeah, sure they do. I'm familiar enough with what "work from home" means that you're not gonna take me in. But you might be right that a typical jury would be convinced you're working just as hard, if not even harder, as if you were in your office.
If you're trying to insinuate that people working from home are not actually working, you're not familiar enough with it at all. The data is pretty clear on this point: productivity is typically *higher* for workers who work from home. Exactly why is open to interpretation; my personal take is that it's a result of higher morale. But the facts are really not in dispute.
I was working from home, now I'm back in the office. Could I prove to a jury, if necessary, that I worked as hard from home as in the office?
Yes, easily. Provide evidence that payroll ran on time, accounts were processed, emails answered in a timely fashion, returns submitted, etc. as per my normal work.
Did I sit at the desk every single minute of that time? No, but I was *more* productive precisely because in the slack periods, I got my domestic work done during the day instead of having to cram it all in after returning home from work in the evenings. That meant if I was working on something that ran over normal working hours, I could work on it until it was finished, instead of having to rush out the door to make sure I could get home and not miss my transport.
Now I'm back in the office, do I sit at my desk every single minute? Yeah, more or less. Do I *work* every single minute? No, because there are slack times. So I surf the Internet, do online shopping, etc.
If people don't have the 'work ethic', then it matters little if you force them to return to the office. You will then have to have a supervisor or manager standing over them to make sure they are working, which is a waste of the manager's time and will either slow things down (as the manager is not getting *their* work completed) or will necessitate hiring on 'standing over the workers' managers, which is not going to reduce headcount and expenses.
Is there a way to, collectively, bring about a culture where fewer people get sued?
I had an odd personal experience where I went to a fitness activity (think squid game but pretty banal) and they had a long completely unenforceable contract that basically said 'if you die for any reason you can't sue us', for some reason I got grumpy about this and didn't sign, and the rest of my friends had a great fun completely risk-free time.
Unenforceable non-competes I think similarly have a lot of value, where you can create a non-enforceable culture of people being willing to take risks and collaborate with a much lower chance that you will lose key assets to competitors or employees leaving to make a startup with your intellectual capital.
I feel like I would genuinely be better off if I could sign a contract that allows my doctor to kill me without risking liability, because I think in general the doctor wants to help people, and is laboring under terrible incentives on that goal. Even if this contract wasn't 'real' even if it was unenforceable, I think the doctor would give me better treatment if they thought I was the type of person to sign this contract.
Imagine doing "whatever it takes" to cut $2T in federal expenses, or doing something as silly as laying off 50% of all government employees. If that costs you more than $2T in output per year, growing at the pace of real GDP, then the initial cost cutting won't have been worth it. Scott asks the right questions - WHO is worth laying off? WHAT regulations are worth removing from the books?
I'm sorry but if you're patting yourself on the back for removing a regulation designed for an unaired show, you might as well start putting on the clown paint on your face. Isn't it more likely that some form of the Pareto distribution is at play here, with a minority of laws causing a majority of the latency and unproductivity of the government? Similarly on the human capital side, isn't it more likely that a select few people at every layer of the federal machine are responsible for a majority of the stalling and enacting of red tape?
Also, if we're saying that the federal government is a burden on taxpayers, it's not so much because people feel like they're paying an ethereally high amount of taxes, it's because people feel they're paying too much and getting too little in return. Whether that comes from the state of infrastructure around where they live, the curtailed speed of innovation they're seeing in their industry or whatever, people ultimately see their taxes as an investment and they're unhappy with their return. It seems to me that blanket budget cuts and layoffs would increase chaos and do little to reduce the faults of the bureaucracy.
If you want a better return on your investment, you typically have to accept more risk per unit of return. If you want a meaningful change in how people perceive government spending, you're unlikely to get there with something as "simple" as laying off 50% of the workforce - you need brand new ideas that won't please everyone, and thus are much harder to actually be put in place. You might want to start introducing more competition for government jobs, by increasing salaries meaningfully and reducing job safety, like making it depend on regular performance reviews. You might want to start having clear goals for people instead of letting hundreds of thousands of people coast their way to retirement, or create fake jobs for their spouses (stuff I read in comments on this post). But this would involve seeing the government slightly more as an enterprise rather than a safety net. The majority of its spending could still be done to support lower income people and at-risk populations, but by making it harder to get involved and retain a job once you have it, you'd naturally generate competitive dynamics and, hopefully, better outcomes for the people's investment in their country.
Note this isn't a perfect solution, in fact I could already see many flaws as I wrote this comment, but the brunt of my point is that you're not gonna get to a visibly better situation by slashing useless regulation (because you need to slash "currently in use but problematic" regulation, which is going to anger people on the being slashed side), or by massively reducing the workforce (for reasons explored in Scott's article), or by doing any "easy" stuff. You could of course optimize the process by singling out actually problematic regulations and erasing them, picking the low hanging fruits, but once you've cut the 50 most annoying regulations, you'll still have a whole lot of people screaming at the size of the government.
By what metrics are you judging efficiency in this case? Yes, removing pensions and entire departments would be efficient if your mission is simply to cut costs. But when you cut cost in your own life (say deleting a bunch of subscriptions, or changing your energy provider for a cheaper one), are you not being very intentional with the costs you've chosen to eliminate? If I took a kitchen knife to your subscriptions, maybe I'd end up removing one that, upon further inspection, provides you with a lot more output than it costs you. Maybe if I force you to destroy your current home energy infrastructure, I cut costs immediately but then you have to make a large investment in Tesla electricity generation that has a 10-year payback period.
There's part of me that wishes they actually would try some insanity like firing half the civil servants. 98% chance everything breaks, Ramaswamy looks like an idiot, and everybody appreciates the important work of civil servants a bit more (with the downside that lots of things are now broken, but hey, I don't live in the US and I appreciate their sacrifice to teach themselves and the rest of us a lesson). 2% chance everything's fine, Ramaswamy looks like a genius, lots of public money is saved, the rest of the world gets a remarkable example to follow, and I get to have my opinions changed.
I worry that "repeal two old regulations per one new one" would have a perverse incentive, unless you can 'bank' repeals. Otherwise, you'd want to delay removing regulations until you need to pass a new one, rather than the ideal case of removing regulations as soon as they're determined to be unnecessary.
Presumably part the goal is just to get folks combing through the existing regulations, running a sorting algorithm weighted by willingness to remove.
Lot of what's found will be things that are mostly useless, but also not urgent, and a list of those can be accumulated as easy 'permission slips' for adding future regulations... but that list is also a clearly non-renewable resource, so it'll force some degree of cost-benefit consideration when any new additions are being considered, meaning a higher floor for the incoming material's quality.
When such a search turns up some genuinely terrible regulation which deserves to be removed ASAP, it can simply go on the top of the stack, get promptly paired with whatever was already latest under consideration to be added. Even if the new addition turns out to be useless, and the excised blight only counted as one of the two required, that's at worst break-even on the stock of permission slips.
Ongoing process of cost-benefit analysis hopefully means there'll be beneficial but, again, non-urgent proposals sitting around in an informal queue, getting debugged and fine-tuned while they wait for such an excuse, rather than half-baked concept sketches pushed straight to production and forgotten about on the basis of "something must be done => this is something => something has been done."
Former lawyer at U.S. Environmental Protection Agency (8 years in the New England Regional Office) with three specific (and distinct) examples showing the naivete of Vivek’s view.
Example #1, an example of the folly of the supposed distinction between agencies whose job is to “ban” vs. “approve” things: The Clean Water Act, which has the stated goal of ultimately eliminating all discharge of pollutants into waters of the United States, prohibits discharge of pollutants into those waters except with a permit that applies (I am greatly simplifying) a combination of technology-based and water quality-based standards to that discharge. A newly built facility that wants to discharge pollutants needs to apply for & obtain a permit, and then the Clean Water Act requires that it must be renewed every 5 years, but an expired permit is valid until the permitting agency issues an updated permit, and most permitting agencies have a backlog. The renewal process involves (again simplifying) looking at technology-based and water quality-based standards. (Also, both the original and all renewal permits are subject to public notice and comment.) Sometimes a renewed permit is just reissued because there’s no reason to change anything; sometimes it gets more stringent (e.g., because the water quality of the receiving water has degraded). A company that wants a new discharge wants the agency to move quickly and approve a permit; a company with an existing facility, existing discharge that it doesn’t seek to change, and an existing permit would be happy if the agency never gets around to it.
The people who write Clean Water Act permits (engineers, supported by biologists and lawyers) are the same people whether it’s a new permit (which companies want approved quickly) or a renewed permit (which companies, simplifying, want to never happen); writing new permits vs. renewed permits aren’t separate skills. In fact, it often makes sense for the same permit writer to be attached to a particular permit over the course of their career, for institutional knowledge.
So, simplifying, if you cut 50% of permit writers, you’ll make it harder for new construction and facilities (whose discharge might not even be that bad if they’re using modern methods) to get approved, and easier for older, out-of-date facilities to continue very bad discharges.
Example #2, on why banning & approving aren’t separate agencies: I worked on both the Clean Water Act and the Clean Air Act. (At EPA's DC headquarters, people focus on just one statute or even part of one statute; at the regional offices, where the rubber meets the road, people are sometimes less specialized.)
Clean Water Act permits are weighted more towards renewals (because of that 5-year renewal requirement, and the statutory goal of eventually eliminating all discharges to water), whereas Clean Air Act permits are weighted more towards new facilities (simplifying, new big facilities with an air pollutant discharge need a preconstruction permit requiring them to meet technology and air quality-based standards; older facilities face far fewer requirements). If we want to simplify, we could say that most of my Clean Water Act work was more about cracking down on older facilities (“banning,” though not really – a discharge is almost never banned, it’s just subjected to more stringent requirements than the facility owner would prefer) whereas most of my Clean Air Act work was about approving new facilities (“approving”). If Vivek had fired me based on my SSN, he’d have eliminated both "banning" and "approving."
Example #3, on why sunset requirements for regulations can actually increase bureaucracy for no good reason: You cited Idaho’s automatic sunset requirement for state regulations. Here’s a counterexample to consider. The Clean Air Act requires states to develop State Implementation Plans (a package of regulations) to attain air quality standards that they’re not presently attaining, and EPA approves those regulations. New Hampshire developed some regulations to control e.g. volatile organic compounds (I don’t remember the details – stuff like requiring gas stations to have certain technologies to limit VOC release), and EPA approved them – they met Clean Air Act requirements. The problem is, New Hampshire (I guess like Idaho) required all state regulations to sunset after I think 5 years. But if New Hampshire had let its volatile organic compound regulations expire, EPA would have been required to step in and taken federal control over New Hampshire’s gas station regulations. Nobody wanted this. So, like clockwork, every 5 years New Hampshire would need to re-enact its volatile organic compound regulations (time spent by the New Hampshire Department of Environmental Services staff), and EPA would need to approve them (time spent by EPA). In most cases, this was pretty routine, because NH wouldn’t change the regulation much or at all, but we did need to check – we couldn’t rubber stamp its proposal without actually reading it. Now consider that NHDES needed to do this for many air pollutants, not just VOCs. Thus, as a result of NH’s requirement for regulations to sunset, a considerable amount of time by both state and federal regulators was wasted in re-enacting and re-approving regulations that everyone agreed were perfectly adequate the first time around. This actually ended up consuming a fair amount of my time.
By the way, if Vivek had eliminated my job, EPA would still be under a legal obligation to approve the (new, but not really) NH regulations, and if it didn’t, environmental groups could have sued EPA, in which case a different EPA lawyer from my office would have been required to stop whatever they were doing – perhaps approving an air permit for a new facility that would bring jobs and electricity! – to work on the lawsuit.
I may think of more examples and if so I’ll add them.
This was fascinating, thank you for telling us about your personal experience.
Re example #2: what is your personal view on the best use of your time, between the more "banning" side of it with regards to the Clean Water Act and the more "approving" side with the Clean Air Act? Is the complexity regarding both the same, or is one much more straightforward than the other? If that's the case and let's imagine approving is both the more personally gratifying and straightforward process, couldn't we optimize future Acts by having a more "approving" template in mind for them, instead of spending a lot of time penalizing older facilities? Similarly, do you think a more "approving" slanted procedure would lead to more suing and unrest, since the older facilities with the most problematic pollutants would see their life majorly extended (even more in the case where the new Act is seen as very stringent and costly)?
Re example #3: can you see the benefits of a more specific 5-year sunset requirement, based on the (probably many) cases like the one you're detailing here? It would have a less dramatic impact on the number of pages of regulation in a State's code (fewer brownie points with voters for the Littles of the world?) but it would allow the sunsetting of specific regulations that are not anticipated to generate more work than they're eliminating. Or is your view that sunsetting as a tool is fundamentally misguided?
Thanks for the kind comment. Again oversimplifying: as a general matter, new things pollute less per-unit-of-whatever-generated than old things, because technology improves. And the statutes are designed to impose more easily impose more stringent technology requirements on new facilities (where it can be priced in from planning) than on old facilities (which may be harder to retrofit).
I'm not sure approving is more personally gratifying or straightforward (these are, after all, facilities that will emit pollutants, and neighbors will yell at you at a public hearing for a permit to build a new one that's going to pollute in their neighborhood) but -- here air and water can be very different -- often approving a new facility that pollutes a moderate amount will help drive an older facility that pollutes much more out of business. An example would be in New England, the construction of new natural gas power plants (which pollute! and this pollution needs to be controlled, often more than the company wants to do! which requires bureaucrats) led, from the 1990s forward (through economic forces that are sort of predictable, though out of EPA's control) to a decrease in usage of older, less efficient oil and coal power plants (which pollute much more, and can be harder to retrofit/control), thus resulting in air quality improvements.
On #3, obviously not all state regulations are subject to a federal approval requirement - most aren't! But my guess is that people overestimate the # of regulations that are genuinely totally obsolete and can be completely repealed (whether automatically or by deliberate action) and not need to be replaced with something similar that addresses an updated version of the same topic. But that's the type of ordinary regulatory updates that all agencies should do anyway - sunsetting the old one on an automatic schedule doesn't really help, since it takes time away from considering which regulations need to be substantially modernized vs. which don't, and forces regulators to waste time re-enacting regulations that were perfectly fine.
People may also over-index on things like "number of pages in regulatory code" as an important measure. But nobody prints out the entire regulatory code (except as a stunt for a TV ad) and has its size as a burden on them. What matters to a regulated entity is not the number of pages, but the actual burdens of the regulations involved. You could write a one-sentence regulation that says "All burning of coal, oil, and natural gas is hereby and forever banned in this state without any exceptions" and it's very short and crisp but it's obviously way more burdensome to all industries than a 30-page regulation specifying control technology requirements for pollution control.
It's certainly true that some regulations are written in wordier fashion than others. It's also true that some regulations are written in more difficult-to-understand language than others. Notably, simple language isn't necessarily shorter -- sometimes making a regulation "easier to read" to a layman makes it longer. This is a value tradeoff.
As for truly obsolete regulations that serve no valid purpose anymore whatsoever, in some platonic sense it would be nice to get rid of them because they're taking up some type of space and presumably imposing some type of burden on somebody, even if rare. But it's usually a very low priority. By analogy, virtually every state has weird laws that everyone agrees are obsolete, like banning spitting on Sundays or playing music involving more than two guitars. They take up pages in the statute books, but legislatures rarely repeal them because it's not worth the time and effort. An automatic sunset provision doesn't really help with this because the time wasted re-enacting regulations that were fine to begin with (harder if you've fired 1/2 your bureaucrats) almost certainly exceeds the time it would take to sort through and propose & finalize repeal of outdated regs (which is also harder if you've fired 1/2 your bureaucrats - with a skeleton staff, you'll focus on the absolutely most essential things, and repealing obsolete regs that aren't bothering anybody will never be that).
There are other ways to reduce regulatory burdens on industries while obtaining environmental quality improvements -- for example, cap-and-trade is a proven method for reducing certain types of air pollution faster, and at lower cost, than "traditional" regulation. However, it, too, requires bureaucrats - someone has to design, operate, and enforce the system.
So much comes down to the insanity of the American legal system. Whether it's constructing a building, implementing a new regulation, or approving a drug we let anyone with deep enough pockets slow action to a crawl and cost whoever is trying to take action ridiculous amounts of money to push it through.
For your amusement, an effort post I did over at Slow Boring on the approval process (though I will note, if you reduce resources enough, people may just be able to do stuff without getting caught/stopped, whether that's good or bad in any individual case, I think it's pretty bad overall for general order reasons):
Okay, effort post. Let’s walk through a small federal project/approval (apologies for the length of this). This is fairly generic and isn’t a copy of any individual project I’ve dealt with, but rather is a composite which doesn’t reveal any confidential information. I will occasionally add complicating factors I’ve seen arise at various points in brackets, but those are fairly uncommon.
Let’s choose a project most agencies will have to deal with, assuming they manage land. A power company needs to run a power line across their property (note, this simplifies things, if we’re talking a federally pushed project, then there’s a whole bunch of planning that is happening at early stages to figure out where/how to do something, getting this to line up with NEPA is a horrible mess as you can’t evaluate environmental effects until you know what you’re doing, but can’t know what you’re doing until you’ve evaluated environmental effects, so you do this stupid little dance of identifying alternatives, then coming up with reasons to screen them down to a reasonable number to evaluate, then evaluating those, at least for larger projects).
This will begin with a real estate application, generally, requesting a permanent easement to use federal land. [[Or it will begin when they got caught just doing construction prep work on federal property and then act very confused when its pointed out they don’t have any rights to do that yet, after all they got permission from the city/state/maybe even federal grant money or some other permission like a CWA permit, see below, so surely they can just build on federal property!]]
Having applied, the first thing that happens is that the realty specialist (or equivalent, terms vary significantly by agency) will need to figure out how to pay for this work. In some agencies there’s a pot of money for issuing such permissions, in others it's funded by the local office, in others it's funded by the applicant. If there’s just a pot of money, the question is ‘is there money for this action now, or does it go on the pile for next year, assuming we get money then.’ (remember you’re usually dealing with one year money and your not getting any more until next year, hopefully, assuming you’re not furloughed, or under a CRA where you’ll get fractional funding).
If it’s funded by the local office, you need to reach out to them and make sure they’re willing to fund this one, not some other local action. If it’s funded by the applicant, you need to figure out how much it’s going to cost. Sometimes there will be set schedules of costs, sometimes there won’t. A standard choice is ‘we can give it a general estimate and get to work with that, or you can pay us for a more specific estimate, which will cost X thousand dollars.’
Given this is a private action with no benefit to the federal government, we’ll say they’re paying the costs themselves. Let’s say this is a reasonably successful company, they want this powerline, they say ‘just give us a general estimate and we’ll give you the cash, give us anything left over at the end, and if you need more, tell us.’ The realty specialist says okay, send us a check, or make an EFT like this for ~10K. Then they wait ~1 week while the company does that. [[This is quite efficient, many companies are extremely uncomfortable making payments with no guarantee that this is all it will cost, or that they will get their easement in the end. And so there’s often a lengthy discussion about how this works, both educating the applicant and repeatedly confirming that no, you can’t promise anything at this stage.]]
Okay, so, we’ve got their money...except we don’t actually have their money, to be able to use their money, it needs to get from the finance folks, back to the realty folks, so it can be divided up and used to fund the people who will actually do the work. This is horribly inefficient and can take extremely variable times depending on a bunch of factors I’ve got no visibility on. Expect at least a month. No one can do any actual work on this action until they’ve got funding.
Now, they’ve got money to review the proposal. Step one, talk to the field folks to see if this location is available for power lines. Has it been designated in such a way that it’s closed to development? Is it needed for other purposes that the power line would interfere with? Is this smack dab in the middle of a park used extensively by recreators using drones? Is there anything on the ground that means we should tell them no, or we should place limits on it (if it’s in a park, can we route it around green areas? Can we schedule construction work to not interfere with rec season, etc.) In fact, this frequently happens earlier, as this is the most likely way for the answer to just come back ‘no, this shouldn’t be allowed’ in which case you don’t want to bother collecting funds and usually field folks can give you a general thumbs up/thumbs down fairly quickly as they’re familiar with the areas they work.
[[Note: Thumbs down is not necessarily final. Almost all of the things I’ve mentioned are field preferences, or policy concerns, both of which can be overridden if you’ve got a good argument and are willing/able to elevate your concerns. Almost all policies have the potential to have exceptions granted to them, by their issuing authority, if no one else, though even attempting that will blow timelines all to crap. I don’t think most applicants realize this, but firm nos are rare anyway.]]
But this time let’s say it’s a field, owned as a storage area near a facility, so they’re not worried about having a power line there, so long as it avoids the actual storage lots. So, having cleared that hurdle, the proposal is sent up for what’s frequently referred to as ‘EC [Environmental/Cultural] Review’ or ‘Compliance.’ These are where all the laws that Slow Boring talks about are going to be considered. The big 3-4 are: NEPA, ESA, CWA, NHPA. NEPA is your umbrella statute. An action like this is almost certainly going to be categorically exempted from NEPA so you aren’t working with an EIS, or even an EA, you just need to demonstrate it fits within the scope of the categorical exemption, and that no special circumstances kick you out of that. On its own, this won’t take too long, but you generally don’t complete this process until you’ve finished everything else so you can be sure that none of the others kick you over to an EA.
ESA is concerned about endangered species, if there are none affected, and a power line like this is not likely to have such effects, you’re fine. If not, you’ll need to consult with the relevant service (either U.S. Fish and Wildlife Service, NOAA Fisheries, or both, depending on the species/location). There are theoretical timelines on this, but unlike the NHPA, discussed below, failure for the services to respond does not free you to proceed and there are no legal consequences to dragging things out on their end. They are, at least the offices I’m familiar with, chronically understaffed and overworked, so the 135 days for formal consultation is more a joke than a reality. But, this one we’re able to get to ‘No Effect’ so we don’t have to consult at all. We note that in our NEPA file and move forward. If we did have to consult, we’d likely get to ‘not likely to adversely effect’ and do informal consultation which theoretically has a 60 day timeline and they’re okay about meeting that timeline, as far as I know [[Note: no timelines start running until they determine they have a ‘complete consultation package,’ this starts with a biological assessment assessing the effects to the species, but if they have any questions, they’re likely to view the package as incomplete and assert their clock hasn’t started to run, this is one way they’ll defend themselves on meeting their timelines, sure the actual consultation took two years, but we only had all the information we requested in the last two months, so really, we were more efficient than you could expect! This can either be true, or total bullshit, depending on circumstance.]]
CWA is the clean water act, it’s concerned with water/wetlands. Here we’re going across a dry field! Yay, no concerns. If we were, say, on the side of a river, they might need a clean water act permit if they needed to put any fill into the water, (and maybe a section 10 permit if they were interfering with navigation, say by stringing a low power line across a navigable river but that’s a different issue) similarly, if they were going through a wetland and needed to partially fill it to stabilize a pole base they’d need a permit, from the Corps of Engineers. This would probably qualify for a nationwide permit, but that’s not our problem. [[Unless it is because we’re viewing this as one action and the Corps wants to rely on our compliance work, rather than do their own, or our compliance people want to make sure we’ve got the Corps documentation so we can prove our action complies with the CWA, or both--we may well need a meeting, or even to formally appoint one party to be lead federal agency, or just agree to work together and sync up our permissions. Setting up this meeting and making sure we don’t get crosswise with each other can take time.]]
NHPA is concerned with historical preservation. This is a field, so we’re not concerned with historic structures (probably, but note things like levees/irrigation ditches/etc can count if they’re old enough we’ll have to evaluate them), but we are concerned about archaeological properties, as they’re going to be digging holes and doing construction work. So, our archaeologists check our files, do we have enough survey information to determine if there are any archaeological sites present. Shoot, we don’t, at least not in this area. It’s a storage yard, but the fence and handful of pads went in prior to the NHPA, or someone just installed them without talking to anyone and so there hasn’t been a survey done. We could go do a survey ourselves, on their dime, and will if this is high priority/urgent, but this is a small project and our archaeologists time is better spent reviewing the results of five surveys rather than going out and conducting one, so instead, our realty specialist is told, ‘nope, we don’t know enough, have them file an ARPA permit application to do an archaeological survey of the area they want to install the power lines.’
The applicant gets that answer, hires an archaeologist, [[or more likely, comes back to complain about having to apply for a permit to apply for an easement, then asks who they should hire and is extremely annoyed when we cannot make a recommendation, and then eventually hires a contractor]] who applies for an archaeological resources protection act permit from us. That comes in and is reviewed by the archaeologist who checks to make sure that the work they’ll do is going to be sufficient to meet our needs and will generally consult with the local tribes, as we’re going to be consulting with them on the overall project, so making sure that this is going to produce what we and they need, just makes sense. Any time we do that, we give them a minimum 30 days. Plus our own review, plus getting to the top of everyone's pile of crap, ~45 days, if you’re lucky.
The contracted archaeologist goes out to do the survey, but this is just standard work for them too, and they’ll need to get out to this random field, do shovel tests, some borings, then come back and write up a professional report. ~60 days if you’re lucky. [note: the applicant has some control of this, but they’re likely contracting this out and most contractors aren’t just sitting around with big holes in their schedule.] [[Note: if they actually find important/interesting artifacts, or even worse, human remains, then timelines are likely gone to shit.]]
While that’s happening, folks aren’t sitting idle, the NHPA stuff can’t proceed, but you can work on drafting the categorical exemption, just leaving the NHPA bit empty, to be filled in at the end. Okay, ESA isn’t an issue, nor is CWA, we fit the categorical exclusion--shit, I forgot the Migratory Bird Treaty Act. Okay, but this is an open field, but, there are a couple of trees on the edge of the property, we can’t disturb nesting birds. Okay, easy solution, either you do it outside nesting bird season, or you have a bird nesting survey done before you begin work, to make sure you didn’t disturb the birds. We can put that requirement in the easement (or this may be broken into a bigger construction license and a smaller, permanent easement, as the footprints for those two actions will be different, but let’s simplify it and say we’re putting it in the easement, after all, then we can say the same thing for ongoing O&M, make sure they don’t disturb birds).
Okay, archaeological report is back, our archaeologists review it good news, nothing was found, and it’s in good enough shape that we can use it, without having to go back to the archaeologist who did the work and tell them they fucked up their report (or worse, their survey). We’re able to say no historic properties effected by this action. We write up that report (or stick our cover page explaining our determination on the ARPA report), and because basically all these actions are streamlined to minimize everything, we do all the steps that the NHPA imagines you’re doing sequentially at once here, we propose the area of potential effect to be the footprint of the easement, we describe our research to find the potential historic properties and we propose our determination of effect, ‘no historic properties effected’ and we send it to the state historic preservation officer and all the regional tribes (that is, tribes which occupied this area historically, not just the ones which happen to be in the region now). When communicating with the regional tribes, we also make sure to offer government-to-government (now being called nation-to-nation, in order to annoy me and make me learn new acronyms) consultation. If they say yes, all timelines are shot, because we’ve got to schedule an in person meeting between our leadership and the tribe’s leadership to address their concerns. This can take days to months depending on the priority placed upon it and schedules.
The SHPO never gets back to us, because he doesn’t care about this action. But, 30 days pass, so under the NHPA, we can proceed, since he didn’t respond. Most tribes didn’t respond, one did to say they agreed with our determination. We follow up with all the other tribes to make sure we’ve made good faith efforts, they don’t respond again. [[Note, if they do respond and disagree, we need to try to resolve that disagreement, before finalizing. How long that takes depends on how much everyone prioritizes it. We don’t have to agree with them, but we do have to evaluate their input, which means getting it, which can take time, and responding to it, which will take time, especially since they always have the option of going, ‘I don’t like that response, our council requests government-to-government consultation on this matter, which freezes everything until it can occur.]] NHPA process is finished, it’s plugged into the categorical exclusion, which is finalized and signed. The whole package goes back to the realty specialist who pulls out the one requirement being placed on it limiting construction/maintenance activity around nesting birds, adds that to the template easement document and then pulls out the stuff he was doing while compliance was working. [[Given this is a standard one, the realty specialist most likely already knows that EO 14026 (minimum wage for federal contractors) and EO 13706 (sick leave for federal contractors) do not apply to this ‘federal contract’ and so we don’t need to circle back and discuss the way we do for weirder situations.]]
So, this is a permanent easement on government property, they need to pay consideration for this. [[unless you’re the Corps of Engineers and they’re a rural electrical utility, in which case section 1172 of WRDA 2016 means they don’t have to pay consideration, they have to provide documentation proving they meet the standards of the exception to consideration]] How much? We need an appraisal. This is a small action, so we can probably use informal procedures, but that still takes time and money, but we’re running at the same time as compliance and they’re going to take longer, unless the appraiser is really slammed. Okay, so we’ve got our appraisal, we’ve got our compliance terms, oh, right, we need an actual legal description of the easement area, don’t we? Okay, we go back to the applicant, and ask for that. They get us a survey, but we can’t just accept that, they could have given us anything, so it needs to be reviewed by our survey folks and it needs to be turned into an exhibit we can attach to our easement. Again, this is not going to take longer than the compliance, unless someone’s really slammed, or really slow.
This is a low risk easement, not like an oil pipeline, or lease or anything, so we aren’t concerned about RCRA/CERCLA...probably, and so aren’t going to do, or have them do an environmental site assessment to make sure there’s no issues in the long run, so the realty specialist, besides the above, is mainly acting as a project manager and communicating any questions/needs back to the applicant, who is getting increasingly frustrated throughout this at how long its taking and that everything they’re doing is costing them money, on both sides of the transaction.
But, we’re finally done with compliance and we’ve got our draft easement! Now it needs to go through peer technical review to make sure you haven’t made any mistakes/missed anything, then my supervisor’s technical review, then it goes to the attorneys which is where those fuckers [[me!]] look at everything I’ve done and ask annoying questions like ‘why don’t these numbers match up?’ and ‘hang on, let’s not just copy and paste their project description into the easement, I know they’re asking permission for a two-strand power line, but let’s just issue them a power line easement, we don’t care how many lines they hang from it, right? Let’s not set ourselves up to have to amend if they hang an extra line. Okay, go confirm with compliance that that doesn’t change anything, then change the wording and keep all these communications in the file!’ And they answer all my questions and then it goes to the person who can actually sign it. But we have to sign it second, by policy, so you send it out to the applicant.
Then you answer their questions about the certificate of authority and the need for them to both notarize the signature and have a corporate secretary confirm that the person signing has authority to sign on behalf of the company. Then they send that back, along with their consideration check and the person authorized to sign on behalf of the federal government signs and that goes back out. Then you file everything and discover that you’ve got $37 left in the account that they provided and it will cost ~$200 to refund their money and so you’re going to be stuck with this accounting issue forever and have to explain it to every new accountant for the rest of time. Oh, and the company is having a problem, because the county is asserting that they won’t record the easement, because the exhibit map is in color. Can we please re-do it with a black and white exhibit map?
And you’re done! Call it 45 days to get funding, 45 days to issue ARPA permit, 60 days to work on ARPA permit, 45 days to turn that report into a determination and get it complete and 14 days to take that finalized report and get to final document which goes out to the applicant. We’re sitting at ~7 months. Now, you can cut a big chunk out of the middle if ARPA isn’t needed, that’ll get you down to ~3.5 months. But you can also extend out massively if GTG is sought, or anything archaeological/ESA is present/effected.
This system is obviously real messy, because it’s trying to address everyone’s potential concerns on every action and the statutory expectations are built for larger projects, but apply to everything. There are ways to try to streamline, but I’ll point out, simplifying NEPA doesn’t actually solve the problem, because NEPA’s your umbrella statute, even if something gets entirely exempted from NEPA such that you have to do nothing, you’ve still got to comply with everything else and a lot of work is being hidden behind the lag time that is NHPA consultation.
Note, that the NHPA consultation looks like your big time suck, and the smaller the project, the more likely that is to be true, because you’ve got this statutory 30 day consultation clock. But if your dealing with a big project, the NHPA timelines don’t get THAT much longer (worst case is negotiating an MOA to resolve adverse effects to historic properties, which sucks, but is not going to be anything like as bad as the timeline on an EIS) while your NEPA timelines can extend out to basically infinity with a large project.
Also, I know there are a million acronyms there. I defined most of them, but if any are undefined, then that’s because I decided leaving some unexplained would be more realistic to the applicant experience. That’s definitely it, not a mistake on my part.
So, where is there room for improvement? Well, lots of places. An obvious one is that the government’s fiscal understructure is extremely messed up as is the actual software/architecture for moving money around and distributing it. This is extremely complicated though and outside my area, so I don’t have a lot of great suggestions.
NEPA on these small projects is very rarely an issue, the big delays are going to come mainly from NHPA and ESA with their requirements to bring in other parties (the services and the SHPO/tribes, respectively). Any time you’re doing that, timelines stretch, painfully. The NHPA does include a ‘if they don’t respond within 30 days you can move out’ while ESA doesn’t, but that just incentivizes responding within 30 days to raise some issue and prevent the clock from running, so I don’t think it solves the problem, generally (also, putting that on ESA timelines would incentivize shrinkage of the services so they just couldn’t respond). Instead, I’d focus on reducing the circumstances under which consultation is required.
Removing the existence of informal consultation under ESA and moving all determinations under ‘no adverse effect’ (‘no historic properties effected’) to the pile of ‘don’t have to be consulted on’ would be helpful.
If congress and the president could resist the urge to create a bunch of rules and a bunch of individual exceptions to the rules, things would also be more streamlined, but that’s pretty clearly a pipe dream. Then there’s other stuff, like having a single government-wide land use fee/consideration schedule might simplify things quite a bit (though the BLM charge rates being set incredibly low, then stuck there is not an optimistic example, though of course, these funds generally don’t go back to the agency, so this doesn’t have significant operational effects).
For tribal consultation, the threat of GTG consultation requests hangs over every project and timeline, you could simply cancel those rights, but I tend to think the better option would be some sort of standing meeting, as its the ambiguity which messes things up. The tribes have asked for GTG on this and it will be discussed at the next meeting in two months, isn’t a message anyone’s going to love, but it’s better than, it will be added to the next GTG which we’re still trying to schedule. The problem is the number of agencies and number of tribes and the tribal position that GTG only occurs on an individual tribe-agency setting, in person, preferably in their headquarters. I don’t have a great solution for this. This one comes straight out of an EO though, so is probably the easiest for another administration to mess with, but fortunately, tribal issues seem to actually be depolarizing these days in an interesting way.
The other solution is to remove the threat of litigation on this stuff by amending the APA...but for little projects there’s not THAT much gold plating, because the whole thing is so small, there’s just not that much room for it. But even the smallest actions have a lot of individual moving parts, which have to be run through, even if no one can sue us over them, as congress, or the regulatory agencies have required them.
"If congress and the president could resist the urge to create a bunch of rules and a bunch of individual exceptions to the rules, things would also be more streamlined, but that’s pretty clearly a pipe dream."
Oh man, if we could just get the politicians to sit down and shut the fudge up, wouldn't it be lovely? Alas, it will never happen!
I still like to talk about the time we got a (long-promised and perfectly usual) funding from central government for housing provision, and *three* local politicians from three *different political parties* all issued press releases to the local papers as to why it was they, them, personally, who got that money released and thank you voters of the county, remember this for the next election as to who gets things done round here.
Need I add *none* of them had anything to do with it, as it was routine disbursement of funds due to the council? (They were the same when I worked in student grants; not these particular politicians, but in general yeah - local councillor to local representative would ring up, ask "Did Joey Murphy get the grant?" and if told 'yes', would then tootle off to Joey's mam to tell her "By the way, your boy is getting his grant to go to college, now no need to thank me (nudge-nudge,wink-wink)".
Instead of halving the number of bureaucrats, I propose we halve each and every bureaucrat. This will increase time spent working since they won’t be getting up from their desks and walking around all the time. Guaranteed reduction in number of steps.
In the long run, this will also disincentivize becoming a bureaucrat, and thereby force a more practical and considerate reduction of red tape.
The DOGE mission is mostly theatre and can only achieve incremental or even non-material progress. The main benefit of DOGE will simply be awareness in the public's mind on the reality of the size and waste of the administrative state, whether or not we have the ability to meaningfully solve the issue in the relative short term (5-10 years). The reversal of Chevron, DOGE, and the well documented failures and drawbacks of the growth the federal government are all steps in the right direction. I really struggle to understand how the massive growth of government in the last 100 years has been a net positive to American life. The same is true at the state level. Top down awareness could create a level of outrage and information that fuels a bottom up change in behaviors around how we legislate and govern.
I actually worked at the worst bureaucracy imaginable for 5 years. We did professional licensing for non-medical/non-legal services (barbers, realtors, contractors). My particular job was doing two separate hearings for every applicant who had a criminal conviction, making a recommendation to the board members actually making the decision, and then all the applications were approved. In the case of barbers, every single application was (and should have been) approved. There was no reason at all for this job to exist.
Whenever I hear people talk about bureaucracy, I understand where they're coming from, but they always seem to assume that either the bureaucrats don't know what they're doing is stupid or don't care. Scott Aaronson's conception of "blankfaces" comes to mind.
The truth is that every person involved is acting rationally. We constantly attempted to lower the burden of going through this absurd process. We even skirted the "everyone has to be interviewed" law for minor offenses on the theory that nobody would care/have standing to sue us if we did.
The conservative legislature constantly did things to make that "get around this stupid law of having to interview everyone" process harder. They'd insist we hire non-lawyer staff to run the hearings and oversee the process so they could use non-lawyer pay scales. Those folks were more likely to blindly follow the process. They'd demand we "get rid of half our regulations." They'd do anything other than *DO AWAY WITH THE LAW THAT REQUIRED US TO DO A RIDICULOUS THING* or *STOP LICENSING PEOPLE WHO DON'T NEED TO BE LICENSED!*
Why wouldn't they just repeal this law? Because there's a huge difference between not caring if rapists are barbers and standing up in front of a room full of people and saying "I think rapists should be allowed to be barbers." This communicates a bunch of stuff other than the practical point - similar to communicating on several simulacra levels described by Zvi Mowshowitz and others.
So conservative lawmakers with dual incentives to appear tough on crime and tough on regulation pass nonsense laws that look good but do nothing. The Board for each profession is a citizen board made up of licensed people - they have incentive to make the process hard and even if they're individually conscientious of this bias and careful (which many, but not all, were!) they don't have the legal training to make it easy. The director has incentive to appease his legislative and board member masters by appearing to work hard to cut red tape while making no fundamental change to how quickly licenses get approved. And I, poor sod that I am, have incentive to keep my job and not do or say anything that could be grounds for a lawsuit, *including* explaining to people/acknowledging how absurd the process is.
We ultimately end up in a mirror universe where 200 people perpetuating a crazy system that makes no sense is the only way to preserve their livelihood, their social position, and their integrity.
Once you see this in one place you see it everywhere. As an example, when my child was born my parents told me about a free service the fire department in my area ran that would ensure we had correctly installed her carseat. The service had been discontinued because of liability concerns (if they got it wrong, they could be sued). They directed me to the hospital who told me the same thing. Ultimately nobody in the state (other than friends/family) was willing to give me a quick opinion on whether I was keeping my kid safe because of concerns that they might be held accountable for it. In this case that seems fine (I don't really think I needed a second opinion), but it's emblematic of the bigger issue. Things don't get done out of a general desire to avoid having responsibility fall on any one person. Even the responsibility for avoiding responsibility is diffuse.
The only way to solve this would be to inculcate a sense of courage, not occasionally, in individuals, but society-wide, in institutions. And I have no idea how you'd do that.
I like those "sunset" provisions, so that dumb rules and regulations can be allowed to just quietly expire and go away. Most bureaucracies tend to accumulate a pile of procedures and rules over time to deal with edge cases and odd situations, and then some of them stick around mostly being useless and ignored until someone sues.
One bummer is that DOGE seems to be moving on from rhetoric around mass firings of head-count to immiseration to get people to quit - IE, move the department across the country, get rid of workplace flexibility, etc. Which is a very effective way to get rid of the most capable people in the department with plenty of other options, while keeping both the folks who stay in the lines but suck at their jobs, and the folks who are so mission-driven about it that they won't quit (note that their idea of the department's mission and Way of Doing Things probably does not coincide with DOGE's).
I appreciate the article, but I hope the model that's wrong here is the model that says Trump, Ramaswamy, and Musk are serious about what they say. I hope they're trying to scare people about X when they really want Y. Along the lines of nominating Matt Gaetz for AG and Kennedy to head HHS in order to get their second nominee accepted. Ramaswamy is too smart to imagine this is a real way to cut bureaucracy.
Bureaucracy tends to grow, regardless of the amount of work to be done. Some people in industry had good experiences with some bureaucrats, but that doesn't necessarily mean that all bureaucrats are doing useful things.
From the link you shared: "the number of workers within public administration, bureaucracy or officialdom tends to grow, regardless of the amount of work to be done. This was attributed mainly to two factors: that officials want subordinates, not rivals, and that officials make work for each other."
I would argue it's possible to fight back against, or rein in the two main factors of this law.
The fact that officials want subordinates instead of rivals could be alleviated by creating incentives (monetary or prestige-based) around rivalry, and disincentives around furthering subordination.
The fact that officials make work for each other is even more pernicious, but maybe there's something to be done with a mix of performance reviews, reducing job security and having tasks ("work") be reviewed through an explicit lens of goals that have the highest impact on the people/enterprises/infrastructure served by any given department, thus making a lot of "work for the goal of work" less likely to be distributed.
I think auto approval after a cutoff date is fully in the Overton window now. Either the regulatory agencies will figure out how to radically expedite the process, and even lobby for those changes, or the process will be expedited by the cutoffs
Is it just me or are Scott's and Ramaswamy's arguments rather tangential to each other?
I read this tweet as a call to reduce the federal budget by reducing payroll expenditures (this is the most common use of the term "downsizing" in my experience). His other tweets suggest he also wants to reduce red tape, and theoretically it's possible that he believes there's a direct causal relationship between those two things. But on the surface, looking at this article and the tweet contained therein, it seems that the two are thinking about different variables in a large and complicated equation.
I think Scott’s article would remain relevant even if that was the case, in showing how those two separate goals of Ramaswamy often are opposed to each other, and pursuing one might come at the expense of the other.
This is true, but I think that relevance is diminished by the fact that (unless I missed something by reading this while waiting for morning coffee to kick in) he doesn't even acknowledge the main thrust of Ramaswamy's argument, much less mention the trade-off.
It looks to me like Scott is just missing the point entirely; this is incongruous because he's smarter than I am and not prone to bad-faith argumentaton, but I can't reconcile it otherwise.
Yeah I keep reading because Scott is usually thorough enough to at least also discuss Parkinson's law, but here it's not mentioned at all. In his anger it seems he forget to steelman.
Similar to Peltzman's observation, I think there is a design optimization principle where benefits on one frontier, when used optimally, tend to be converted into benefits on multiple frontiers.
For example, if you suddenly gain the ability to make the same sized engine in a car go even faster and be even more powerful for the same price, you might instead convert those capability gains into several different benefits: you might make a smaller engine for cost savings and have more cargo space with the same performance characteristics in the same vehicle.
This same optimization principle works in reverse. If the cost of an engine doubles, or you have to fire half your staff, if you're being intelligent you'll probably absorb the cost across many frontiers.
So I would predict something like an FDA that approves somewhat fewer things in a worse way (where "worse" may open them up to more liability) and that also cuts back quality, patient safety, (A double edged sword as an employee at a medical device manufacturer - we are more careful for the sake of compliance but at great expense), etc.
However, in order to predict the distribution of the capability losses, and how much fat the FDA has to cut, we need to have some level of understanding of what the FDA is doing internally. I honestly don't respect anyone who tries to answer what will happen without trying to understand what is happening internally at the FDA. For example, I have seen metrics saying there was a 70% increase in FDA employees since 2005. Okay, but at the end of the day do they need them or not? That's less than healthcare GDP growth since 2005 IIRC.
I think it would be far more reasonable for the government to pay an independent consultant to map out what is happening at the FDA and proceed from there. Of course this is impossible because of partisanship, but one can dream I guess. But yes, ultimately if you want to cut the red tape, figure out where it is, do that directly (change incentives?), and then fire proportionally to how much you actually overhired, if you did. It's impossible to fix complex systems by treating them like black boxes and then smacking them with a hammer over and over.
I wonder if the crux of the issue is the courts, rather than the bureaucrats. Under common law regulations grow unboundedly, because new precedents impose requirements on regulators as they get sued.
Inquisitorial systems seem intuitively better at adjudicating complex regulatory regimes. They only consider the letter of the law, so they can process cases faster. And because they can’t make binding precedents, they avoid the vicious cycle of regulation -> litigation -> precedent -> new regulation.
It would be interesting to see a comparison between, eg, Louisiana and Mississippi to see if this actually holds.
When I was a government contractor, the contracting process was extremely wasteful because there were too /few/ bureaucrats (program managers). Small contracts are much more-efficient than large contracts, at least if measured by number of publications produced (a bad measure, but it's hard to find a good measure that can be applied to all contracts in the same way). Government agencies award mostly big contracts because big contracts are much more-efficient at spending money. If you have to give out $500,000,000, it's a lot easier to spend it on one project than on 10,000 projects. Same reason venture capitalists would rather invest $10 million than $1 million.
I once was the PI on a $100,000 project, reporting to a DARPA program manager who was also managing a $100,000,000 project. He notified me at the start of the project that he would not answer phone calls or emails from me and would not read my project reports; and near the end of the project his office told me not to bother writing a final report, as nobody would read it.
But re this: "Vivek . . . seem[s] more qualified to wield chainsaws than to understand legal minutiae."
I am a Vivek stan so I just needed to chime in here and disagree.
Vivek Ramaswamy went to Yale Law School. That's usually ranked as the #1 law school although sometimes I see people calling it #2. I applied and didn't get in. Everyone I know who DID go to YLS seems to me to be a hard-working genius.
To conclude, even if Vivek hasn't practiced law since law school, he is probably qualified to understand legal minutiae.
I’m curious about how that “eliminate two regulations for every new regulation made” has worked out for Idaho. Republicans have been trying to push this sort of catchy policy at the federal level and it seems insane to me.
Maybe at the beginning you’ve got some low-hanging fruit - but you don’t want to touch more of that than you have to, because you want to save it to delete when there are new rules you want at some future point. And then there’s a further question about what counts as “one regulation” - can you delete two old regulations and write a single new regulation that includes both of those as well as the new one you are writing?
Really? I’m not aware of any country that manages food safety, energy security, financial stability, or construction safety with a few stable rules and no regulations that take account of changing technologies.
Maybe they need a Lord to make the career civil servants and holdover political appointees bend the knee. Or a Musk Ox to trample the ones who don't? I'm thinking the nominative determinism is the easy part.
As a bureaucrat, all I can say is, Oh, You sweet summer child.
I think any organization balloons beyond necessary size. Mid-level managers are often rewarded for growing their teams, and rarely for working more efficiently with fewer people (I managed 20 staff! Hire me for more money!). Most management staff are not incentivized toward efficiency. You would think that money inflow problems would enforce efficiency, but I've never seen this realized in 10 years of government; its always threatened and never happens. There is always more money for whatever program we want to continue. There are many programs and initiatives and committees that aren't particularly evidenced based and never really produce anything. Calls to make bureaucracy more efficient result in longer forms (more social determinants of health questions, more demographic questions for representation, more evaluation questions for program feedback, all justified and in the name of doing better work, but none of it shorter or easier for the client). A sense of safetyism or racial justice can be used to justify just about anything, and no one tolerates objections to those values.
Seriously, I used to work in a unit that delivered free health services to low income folks, like 70% of my work was forms. Making sure that people didn't get too many services, and qualified, and justifying spending on every single person to the federal government with a double receipt trail. I have a lot of questions of how much more good we could do by just dispensing care with fewer qualifiers, so that we remove the huge administrative costs of government programs. Yes, there would be some cheaters, and also a larger chunk would actually go to the public.
Just like the Paperwork Reduction Act should have been named the Bureaucratic Job Preservation Act. And generally increased the length of each form, because now they had to add a page saying "we complied with the PRA".
Yeah, in my current job I am anticipating with much joy *another* layer of bureaucracy laid on, this time allegedly to reduce the form-filling for the service agreement we engage with. I sure hope it reduces the form-filling, but the fact that they went for a whole new set of acronyms and a pilot programme testing out the new form before scaling it up nationally and webinars to inform us about the upcoming training for the implementation etc. doesn't fill me with confidence on that.
Mostly I think the new year will be same as the old year: providing the same information to three different bodies, all tasked with implementing chunks of policy in order to cut costs and bureaucracy (but of course really in practice increasing both). And on top of that I have a whole new set of five letter acronyms to learn off and try and remember which refers to what.
"I have a lot of questions of how much more good we could do by just dispensing care with fewer qualifiers, so that we remove the huge administrative costs of government programs."
The irony there is that all the extra box-ticking is imposed in the name of economy and efficiency and cost-cutting. Some politician gets up in the Dáil (or Congress or Parliament) to question the government of the day on reducing the tax-payers' burden by cutting costs and ensuring public funds are being spent as intended. The Minister (or Secretary or whatever title the relevant administration official holds) replies that naturally under *their* administration, the public finances have never been better and moreover they have increased the number of services available to the hard-working people of the nation.
Then they go off and tell the civil service to draw up a report on public expenditure, which means the civil service has to bother every entity providing public services to dig out and present the information to them on top of their existing work. And then the report shows money flowing out, so now the minister tells them to set up something to reduce expenditure.
And that means you, me, and the other public-facing service providers now have to present double and triple receipt checked individual forms filled out to demonstrate that Mrs Murphy is getting €100 a week as she is entitled to receive and not €101, and if she is getting €101 you are to seek repayment of that extra €1 over the next five years.
"Also, doesn’t Congress pass like one bill per year now?"
Now google "Chevron deference" and see what an incredible mess the Supreme Court has created. In the interest of destroying the "administrative state" regulators are disempowered to use expertise and evidence to decisions. So the paralysis continues and expands, new causes of action are offered to claimants and the regulated, and the judiciary is empowered to make these calls instead. Congress could take action to direct and clarify intent, but Congress is no longer a functioning body.
>I was surprised to see someone with such experience in the pharmaceutical industry say this, because it goes against how I understood the FDA to work.
The tweet is from November 2023, when Ramaswamy was running to be the Republican nominee for President.
He was campaigning as a Trump-like populist, so (like Trump) anything he said had nothing to do with any attempt at presenting policy. Instead, the strategy is to make a statement that emotes something the desired constituency agrees with ("bureaucracy bad >:("), while also dialing up the insanity of the statement to 11, ("let's fire 50% of bureaucrats at random xD") so it goes viral when opponents argue with it, rightly calling it insane.
This is what propelled Trump to massive popularity in 2016 and got him his Republican nomination, but didn't work for Ramaswamy because all the populist voters were already all-in on Trump.
That origin of the Tweet makes any steelman of the position ultimately pointless, even as an intellectual exercise. The Tweet fundamentally doesn't come from an intellectual position of "how to make the government more efficient," it's just emoting and virality-seeking.
This article is interesting in intuiting the principles of governmental bureaucracy, but is also the equivalent of wondering why someone would throw a perfectly good milkshake at Nigel Farage by asking: "was it not a good milkshake?"
I don't think this is actually his position, but maybe the strongest steelman of Vivek's tweet is the accelerationist argument, basically:
'The best way to fix this problem is through careful consideration and dedicated effort to reform the system, but that's been true for more than a century and no one has ever done it. It's never going to happen. My plan is idiotic and will ruin everything and cause a gigantic crisis, but people will be forced to *actually* get off their ass and respond to that crisis, and the reforms they pass to fix the crisis with half as many bureaucrats at their disposal will end up being better than the current system in the end.'
I don't actually think that would work, but it's probably the version to argue against.
Why won't they respond to the crisis by just hiring twice the number of bureaucrats?
Or, if you can block them from doing that, do nothing and blame you for causing and exacerbating the crisis with your refusal to hire the people they say are needed to fix the problem? Their primary incentive is not to *fix* problems, but to avoid being *blamed* for them. And here you are, volunteering to take all the blame.
I mean I think in practical terms the answer is 'because the people I have undue influence over control all three branches of government for the next 2-4 years and I think I can push them in useful directions.'
Like I said, I personally don't think this would actually work, but that's the steelman version I think.
They'll let the crisis play out for two years so they can take control of the legislature in the midterms, and then allow fixes only so long as they can take most of the credit and so win the Presidency in four. So, 2-4 years of abureaucratc paralysis with all the unpleasantness that entails, and then unchecked Democratic control until the GOP somehow rehabilitates its image.
Granting that firing half the bureaucrats would paralyse the red tape, could that paralysis be then used to cut the amount of red tape? Presumably the system would find a new equilibrium that's not "nothing is ever done"
I’m a bureaucrat, and think this is a good analysis. One thing to note is that a lot of lawsuits against Agencies go to other Agencies. There are about 2,000 Administrative Law Judges in federal Agencies that are responsible for a lot of these roadblocks. The Merit System Protection Board, Federal Labor Relations Authority, and Office of Federal Operations in the EEOC, are specifically the reasons it’s nearly impossible to fire bad performers in the Federal Government, and the President gets to appoint members to them. Overhauling those 3 would greatly improve the government’s efficiency in my biased opinion.
How often are the bureaucrats themselves responsible for creating the red tape? Obviously Congress doesn't pass every single regulation themselves, that's the whole point of the bureaucracies. Firing bureaucrats doesn't reduce the number of existing regulations, but it could be necessary to slow the rate at which new ones are added in order to achieve any net reduction.
> Two years ago, after consulting with industry, the FDA finally banned 23 phthalates but said that the other five were okay, releasing a 58 page decision explaining its decision. Two days ago, the environmental groups sued, saying the remaining 5 phthalates are still bad.
If they're going to get sued anyway, it seems like the report is kind of pointless? Or is just supposed to be more efficient to do all this pre-emptively rather than waiting for the lawsuit? Seems like now they're just doing all the legwork twice. It sounds like the real problem is the fact that anyone can sue anyone else for anything (which, as far as I'm aware, is not something most other countries allow). So rather than just having a top-down regulatory process, or a dispute-resolution process worked out in courts, we have both, stacked on top of each other (and the legal system can take forever to get anything done too... see e.g. https://en.wikipedia.org/wiki/Google_LLC_v._Oracle_America,_Inc.).
Presumably they will keep being sued until they produce a decision (and supporting materials) so ironclad that the lawsuit fails. At least some parties will be negatively affected by the decision and will have standing to sue, so only a decision capable of surviving challenge can actually be final and implemented.
> Halving the number of FDA bureaucrats wouldn’t have literally zero effect on this balance. It would mean that approving new drugs would be delayed twice as long. This would be a little more outrageous than the current delay, and might shift an outrage-minimizing FDA director slightly in the direction of cutting rules. But solve for the equilibrium: there would still be more delay than there is now.
Doesn't this imply that the best way to decrease approval times is to alternate between sharply cutting --> increasing headcount? The underlying model for time-to-approval is something like:
1. Decreasing the number of total work hours required to gain approval
2. Increasing the number of bureaucrats doing the work
You're saying by sharpy decreasing (2) it will force an increase in (1). But you can go back and increase (2) to where it was, and you'll be in a better place than where you started. Rinse, and repeat.
A particularly well documented example of the paperwork processing issue is wait times for the ATF to process NFA forms. Up until last year a 9 month wait for approval was common but after additional clerks were hired approval times dropped to a matter of days.
This of course begs the implicit question that if the vast majority of applications are approved does the manual form need to exist at all? Implementing a self service registry could probably be done without legislative action, but the ATF would need an actual incentive and institutional will to do so.
Aren't you assuming that all bureaucrats are equally competent and hard-working? If a significant number of them are 10-3 long lunch break types, firing half of them would just require the rest to, like, work a full day.
I sometimes think we should have an automatic sunset period for all laws (or maybe just regulations), combined with a rule that you can't approve or reapprove a law or regulation with a simple majority vote unless the person sponsoring it reads it out loud to the people who want to vote for it.
If it worked, then I imagine the written rules would become a lot shorter, which is probably a win. But you'd need to prohibit rules which include the text of non-rules by reference.
This would presumably shift power to executives and judges, since shorter rules tend to be more open to interpretation.
I know something about the Idaho example. They reduced regulations by restricting state regulations and instead relying on federal regulations. Most states add regulations beyond what the feds require, but Idaho no longer does. And it’s not a disaster for their citizens, because the federal regulations protect them from most bad outcomes.
This is part of it and another example of why using Idaho (or any other state) is not particularly applicable to the federal government. To continue with the FDA theme of this post:
"The U. S. Food and Drug Administration (FDA) publishes the Food Code, a model that assists food control jurisdictions at all levels of government by providing them with a scientifically sound technical and legal basis for regulating the retail and food service segment of the industry (restaurants and grocery stores and institutions such as nursing homes). Local, state, tribal, and federal regulators use the FDA Food Code as a model to develop or update their own food safety rules and to be consistent with national food regulatory policy."
Idaho, like many states, just adopts a version of the Food Code for their jurisdiction. If FDA didn't produce the Food Code, then Idaho would have nothing to copy.
> How much will they get yelled at if they take too long to approve drugs, vs. if they mistakenly approve a bad drug?
An alternative way of framing this would be, "How many people will die if they take too long to approve drugs, vs. if they mistakenly approve a bad drug ?". This is a complex question; for example, in the extreme case where we abolish the FDA and cease all drug regulation, maybe pharma companies will remain conscientious enough to do at least some minimal drug testing, so on average more lives could be saved; however, upon seeing a mass uptick in drug-related deaths, consumers will conclude that drugs are too dangerous now and stop taking them anyway, thus driving up the number of deaths. And maybe we'd get a situation like in early 1990s..2020s Russia, where snake oil vendors dominated the market.
I mean I think it's important here that those *aren't* the same question, and that this can shift the balance from where it ought to be (many people would say it has).
The whole post can be boiled down to: bureaucracy size might not be the only metric for you to look at if you want to improve the overall efficiency or throughput of the system.
There are tons of other levers for you to consider here. Complexity of requirements. Number of hand-offs. Outdated processes taking days instead of minutes due to lack of automation. Staff motivation levels. Staff incentive structures. Bullshit jobs that should have never been created to begin with. Org structures that have not scaled with the needs of the organization. Etc etc.
The size of the bureaucracy is certainly *a* metric to look at, and organizational size is generally correlated with sluggishness, but it would be silly to reduce the whole audit to just this one metric. I'm confident Vivek knows that well, but perhaps the nuance of improving organizational efficiency doesn't fit nicely into a bumper sticker unlike "fire half of everybody".
This was what I was thinking about. If bureaucracy isn't measured in bureaucrats, then what unit is it measured in? That sounds like a question that is actually worth talking about. Obviously it's a political question, but I wonder if there are approaches that could attract reasonably broad acceptance?
Like, The Economist has something like an ease of doing business survey that it conducts annually, I think. I wonder if that methodology could be applied in some way? In industries with large companies that have actual compliance departments, you could look at the budget for compliance as a fraction of turnover, I guess? Another possibility would be to measure corruption in some way. Excessive bureaucracy preventing lucrative business is a recipe for corruption, so if you could quantify that, it might give you a snapshot of where you should be focusing your streamlining efforts.
As someone with experience in defense contracting, I think your overall thesis here is definitely sound. If anything, our government customers are understaffed, because they are overloaded with low-value added but legally required bureaucratic cruft work. Firing half of them will make the problem worse.
For example, we have projects running late and over budget not because we can’t execute them, but because (in what should be an entirely expected occurrence) development is hard and the requirements and circumstances change, but the bureaucracy can’t respond in any kind of agile way. The changes are not a big deal from a technical perspective, except that these contracts are firm fixed price (an outcome of the last round of government efficiency do-gooding) so any tiny change that impacts cost at all requires a contract update. Which incurs a massive amount of bureaucratic overhead and schedule hit to get all the paperwork filed, and our government contracting officers don’t have the bandwidth to keep up.
Literally, on a nominally 24 month contract, an update that changes the bottom line price by 1% has taken 4 months to actually get on contract. And in the mean time the contractor either has to work at risk of never getting paid (which we are increasingly less willing to do because the federal bureaucrats have been told to stop offering us any accommodations in return) or sit on their hands while the delivery slips to the right.
Now, there are definitely some bureaucrats that are worse than others - an annoying stickler in the wrong spot can bring hundreds of people’s productivity to a screeching halt - but you won’t get rid of that roadblock by randomly giving the agency a haircut.
This post seems to be missing the fact that many, perhaps most, bureaucrats that work in bureaucracies do not do the work of processing forms or dealing with the public. Many are middle managers, HR functionaries, and all the hundreds of different make work type positions that don’t really need to exist. You couldn’t just randomly cut half and get just these positions, but cutting half wouldn’t slow things down by 50% either. It might even speed it up, depending on how many stupid meetings could be avoided.
> But the way that the SEC actually makes rules is: [ ... long complicated process ... ]
> The incentives, for the SEC, are bad. If it makes new rules, that is a lot of work, and those rules will be attacked from every angle. Probably they will be struck down, in ways that limit the SEC’s power and create greater confusion about what is allowed.
> Meanwhile if the SEC just sees some crypto project that it doesn’t like, it can sue that project — probably in a court of its choosing — say “this violates longstanding securities law,” and have a decent chance (not a certainty!) of winning. Its chances are better not only because it can pick a more sympathetic court, but also because it can pick a less sympathetic antagonist: It can argue “we need the power to regulate crypto” in a case where investors lost everything and the value of regulation is clear, rather than writing general rules and getting sued by a nice upstanding firm that doesn’t like them.
This sounds similar to what you're saying but a big difference between your understanding of bureaucracy and Matt Levine's (at least with regards to crypto) is you think the bureaucrat's goal is to make clear bright-line rules so everyone knows the game, and Levine's view is that bureaucrats are playing political games to support unofficial goals.
Oh, and on your section II, it’s important to note that not only are all agencies not the same on the “our job is to approve things” vs. “our job is to deny things” spectrum, often multiple agencies with overlapping domains are in conflict with each other!
So, for SpaceX: on the one hand, they’ve got some environmental bureaucrats to deal with in California that are absolutely BANANAs (build absolutely nothing anywhere near anything). On the other, they’ve got customers in the Space Force that would really like them to get this expensive critical National security asset into the sky as soon as possible, thank you very much. Who are also pissed off that the CA BANANAs are stepping on their turf (literally, in the case of their Space Force launch sites).
So you need the Space Force / DoD bureaucrats to work on your behalf and make the BANANAs shut up. If you fire half the DoD and half the environmental bureaucrats, the balance of power doesn’t change and nothing speeds up. If you fire half the DoD and the BANANAs stay because they are state of CA employees, well now you’ve played yourself and things will take even longer.
Bureaucracies and law share an important quality in common: They calcify knowledge. This comment isn't intended to be a precise description of something, so much as to convey a concept.
When this works the way we want it to, after a fertilizer plant explodes a rule is created which prevents future fertilizer plants from exploding for the reason that this fertilizer plant exploded. Say, all fertilizer plants have to have a dedicated ammonia processing unit that follows these standards and is constructed out of stainless steel and is maintained and checked every month.
When this works the way we don't want it to, the rule prevents a new process from being implemented which is more efficient and/or better and/or has an overall lower risk profile because the rule requires the old process to be in place. So the new fertilizer plant which doesn't use ammonia still has to have the dedicated ammonia processing unit, even though it doesn't use it, so nobody builds it because that extra expense makes it unprofitable.
And in the third situation, the rule doesn't matter - the new fertilizer plants don't produce fertilizer, but rather a precursor to fertilizer using pure nitrogen that weighs less and the new plants don't quite meet the definition of a "fertilizer plant" as defined in the original rule.
Anti-regulatory and anti-bureaucracy efforts tend to focus on the second situation. Pro-regulatory and pro-bureaucracy efforts tend to focus on the first situation. Meanwhile the third situation keeps being the reality of libertarian progress.
The older a rule is, the more likely it is to be in group 2, until at some point it joins group 3. The newer a rule is, the more likely it is to be in group 1.
The older a bureaucracy is, the more it will tend towards group 2 and group 3 rules, both because the rules for creating rules slow down the creation of new rules, and also because older bureaucracies will have more rules in the first place. The faster technology changes, the less time this takes.
At some point in the rate of change in technology and the average age of bureaucracies, it becomes nearly impossible to create group 1 rules; by the time the rule has actually been calcified, it is obsolete. And at some other point, it becomes nearly impossible to create group 2 rules. I think we're currently stuck in the pit of "It's very difficult to create new group 1 rules".
In the ideal situation, we only have group 1 rules; that's basically what we've been trying to do for the past century but it hasn't worked out great for us. In the second-best scenario, maybe firing half the bureaucrats might slow things down enough that we're left only with group 3 rules going forward.
Lol. First of all, Vivek is a very smart guy but he spent about 7 years running his company so saying he's an experienced pharma guy is a bit of a stretch. Seven years is just about enough time in pharma to realize you don't know very much.
Second, I don't think he's talking about the FDA specifically.
My experience with the FDA is that they're very very busy. Most reviewers are swamped, and they don't have a lot of extra time on their hands. On the contrary. I remember showing up at an FDA Advisory Comm Meeting with a phalanx of 30+ support personnel (several people to run the powerpoint slides, several statisticians, people to pull up data real time, etc.) after practicing 8 hours a days for 6 months for our presentation, and the FDA clinical reviewer showing up with no support people and a thumb drive - we had to load her presentation on our laptop for her because she didn't have a laptop. In general they do a good job.
Besides, FDA is funded 50% or sometimes more by User Fees industry pays, so I'm not sure they can save money by firing the reviewers. Where the fees go is negotiated between the FDA and industry.
Now, one area where there might be savings is the scope of FDA's remit. They have extended jurisdiction into a lot of areas that may be beyond what Congress intended, and if half of their headcount disappears, those would be the areas they might cut back on. These are things like regulation of compounding pharmacies, regulation of off-label promotions (they lost a huge case,Amarin Pharma, Inc. v. FDA, on Freedom of Speech grounds, and they didn't even appeal it, probably because they didn't want to lose and make it binding across the whole country), non-interstate commerce, importation of drugs (if you go to another country and you get treated, then FDA can claim that you are importing the drug… because the drug is in your body, see?), lab-developed tests, right-to-try (experimental therapies) etc.
One addendum about the compounding pharmacies. It used to be all drug was compounded by pharmacists (and physicians). FDA was created partly because when manufacturers started selling drugs, the pharmacists complained, "hey, these guys who aren't even always pharmacists are making drug and just selling them."
in other words FDA's original job was partly to protect pharmacists from inferior and unfair competition. If you weren't a pharmacist, you had to show that your product was not terrible.
Now, ironically, the FDA is trying to shut down compounding pharmacies.
The user fees are an interesting example because FDA only started using them in the early 1990s. My understanding is the pharmaceutical industry and others actually worked with Congress to authorize FDA to collect user fees. They wanted to help fund FDA because they wanted FDA to hire more staff and speed up the review process.
Scott- you’re misunderstanding these two nitwits by engaging with them in good faith. They don’t want “efficiency,” they want to cripple the administrative state. Is it any wonder that Elon Musk, who is under very active and very public prosecution for all manner of securities fraud, thinks it would be great if the SEC had half as many employees?
This whole project is “make the drug applications take two years instead of one, so we can then justify getting rid of the FDA,” not “we’re just these nice well-meaning bros who want to help make the trains run on time.”
This may ultimately be true but doesn't undercut the need for good faith engagement, which isn't directed to these two people personally (are they SSC readers?) but towards public opinion.
Part of proving they're operating in bad faith is proving their solutions, which much of the public takes seriously, couldn't possibly work
Respectfully, this is the failure mode of Scott-style “rationalist” thinking and I see it happen all the time. They’ll keep using arguments as soldiers and throwing them at you until your forces are depleted. While you’re busy trying to calmly and rationally rebut points 7,8, and 9, they’ve already made up points 10, 11, and 12.
I don't see a connection between this claim and what you're responding to.
I think also "assuming people one doesn't like don't actually believe the things that they advocate for" is a take that's been heavily influenced by your political position... but undoubtedly, my having the exact opposite impression re: Musk (don't know much about Ramawhatever) is influenced by my politics too.
I'd be fine if the outcome was "completely get rid of the FDA", though. If that's the ultimate plan, more power to 'em!
People dramatically underestimate the extent to which Congress and the courts are responsible for the state of our government institutions. It's of course understandable why some elected officials want to deflect blame to government institutions. Sometimes, they do mess up. But it's pretty lazy to ascribe "laziness" as a monocausal explanation for how we ended up here.
Let's take the current wildfires in LA as an example. Everyone who works in wildland fire management in government knew something like this would happen somewhere due to continual encroachments in the wildland-urban interface. Wildland management officials also understand that clearing out brush (either mechanically or with prescribed fire is helpful). To emphasize this point, people in the field literally refer to prescribed fire as "beneficial fire." They know how to mitigate some of these severe outcomes. It's worth spending some time pondering why there is a gap between what bureaucrats "know" they need to do versus what actually gets done.
As one example, to continue the theme of this post, paperwork exists. Burns on federal lands have to comply with environmental permitting laws or other things like the Clean Air Act (go read about the interaction between the exceptional events rule and prescribed fire). The bureaucracy is not magic; it cannot ignore laws just because they're insane or result in long timelines or a lot of paperwork. The US Forest Service spends a massive chunk of its national budget (~40%) just on the paperwork to do stuff.
This is stupid, but again, it is also not the fault of bureaucrats. We need to either remove the threat of lawsuit or for Congress to pass new laws and fix bugs.
The bureaucrats also know all this stuff is insane! On wildfires specifically, they've asked Congress for comprehensive fixes to improve the way we manage wildfires. Has Congress done anything? No, of course not. This means we get to continue waiting 4+ years for wildland fire managers to complete environmental impact statements.
Here is a comprehensive report the government spent a lot of time on with some reasonable fixes to address all of the things that people are suddenly now armchair experts in:
I feel like Scott's model is not capturing an important element of the FDA (or other) bureaucratic process in step 1 of his model. Specifically, when regulators are not totally constrained by statute, and can ask for process or application changes based on their discretion, more regulators or more time per regulator inspecting means more work for compliance.
Said another way, I think Scott's model of drug approval should be reworked from “drug approval is processing 1,000 forms” to "the number of 'forms' required for approving/getting a drug to market is modeled as 1000 + f(bureaucrats)."
This model reflects the fact that there are many statutory requirements the FDA must fulfill at each step of the process (the 1000 forms). It also recognizes that additional bureaucrats add 'forms' or requirements at their own discretion.
I would offer the following as an example of where discretion in enforcement is both necessary and genuinely increases the amount of compliance work/difficulty in getting a pharmaceutical made in proportion to the number of additional bureaucrats.
The FDA inspects pharmaceutical production facilities for compliance with good manufacturing practice (GMP). Strict GMP enforcement ensures only high quality pharmaceutical products are produced, there are no adulterated products/byproducts/etc. During this process, the FDA inspectors must have some discretion to flag process deficiencies that are not exactly anticipated by the guidelines.
However, because the inspectors have discretion, more eyes means more flagged items. I want to share a good example of this from my work. Pharmaceutical manufacturing almost always includes a chromatography step -- the purification of the API/molecule of interest on a column packed with a specific resin. Over time, the column resin discolors due to impurities that are being removed from the pure API (this is the purpose of the process after all). In research settings, the columns are clear glass or plastic because you want to monitor the fouling/color of the column.
The manufacturer of the columns we used (glass covered) told us that they offered stainless steel housed columns so that when we get to GMP production, we can switch over. We asked them why we'd want that? We were perfectly happy with the glass system, it met pressure specs etc. They responded that FDA inspectors would always flag a slightly discolored resin as a hazard (e.g. a contaminant or potential for bacterial/fungal growth). In reality, the resin being somewhat discolored is totally normal and doesn't effect the quality of the chromatography or separation (you have checks for this). But, because FDA inspectors would see that, not like it, and require a process change or an expensive swap, companies responded by enclosing the columns in stainless steel and just hiding it from inspector eyes.
Now, imagine that this is happening across an extremely complex pharma production pipeline with many inspected processes, machines, inputs, etc. This is a situation where 10 inspectors in the building are going to find 100 "problems", and 50 inspectors are going to find 500 "problems" with your process. Note that those 50 inspectors are also more likely to find real deficiencies or catch problems you don't see.
The IND and drug approval processes share many similar features to what I've described here. Because they are not totally statutorily constrained, and they have essentially no political cost to finding fault/deficiency, the more bureaucrats you have, the more compliance items you will have to resolve.
Ultimately, I think Scott's model needs to adjust for this fact. Clearly cutting bureaucrats will reduce the throughput on the statutorily required and generally unavoidable forms. However, it may also substantially reduce the burden the companies face from discretionary enforcement.
Many people that function as federal bureaucrats aren't even employed by the federal government. Many roles in HR and Compliance departments, most especially in extremely regulated industries like banking, are effectively monitoring, enforcing, and recommending new federal regulations as their full time job.
The problem is the scope of the federal government, not strictly the size. Fix the scope and everything else falls into place.
Admittedly dumb, I smell barbarians at the gates again. They won't have much time to enjoy their gains this time, though, if they gain anything at all. The nuclear power that's most efficiently organized will probably call the shots after some trouble.
Russia and US combined have 90% of the world's nukes, though. If they joined forces (which seems increasingly likely), they would easily be able to reduce the entire rest of the world to ash if they wanted to.
They're not selling. This is just basic game theory. If World War III was a complete free for all, there's a much lower chance that either the US or Russia comes out on top. But if they work together to deal with the rest of the countries first, now both of them have a much higher chance of winning than before, since now there's only two possible victors. And they wouldn't have to worry about nukes until later, since they basically have a monopoly on them.
I understand that many people such as yourself have absolutely no inkling of the difference between a United States 'green card' and U.S. citizenship but, you know, you really should try harder. FACT is that both 'quick' and belated google searches show no such thing as you claim.
It doesn't pay to be ignorant, really it does not. Learning is Good.
So the difference is largely a difference of a few years, since the green card gives people the right to petition for citizenship after 5 years, and so it’s a matter of waiting. Seems like this is a debate ok semantics.
As a green-carded naturalised U.S. citizen, I can personally assure you that the wait, the paperwork and the (meanwhile) duty of serving the USA IN a war isn't semantical at all. It's a teensy weentsy bit more dangerous than your classroom semantics.
ALSO I don't think this is the place for actual debate or flaming, but if you feel any urge to continue, please take it to my actual real email: dawson55@gmail.com.
I highly recommend Michael Lewis’s book *The Premonition* to get a heartening glimpse of how quickly it is possible to address a knotty problem outside the bureaucracy. It’s about a few scattered professionals who took it upon themselves to figure out the covid situation & smart, prompt ways to address it. Most tried but failed to implement their ideas & inventions. Some of what they came up with turned out to be flawed, but you get the strong impression that these people would not have pushed back at all against good evidence of flaws, but instead thrown themselves into developing an improved plan.
For example, one guy was an expert on flow structure within fractures and soil. He helped his high school daughter do a science project on contagion, using approaches he had developed to model collisions of particles. They ended up concluding that schools were the setting with by far the most close & prolonged contact between people, and based on that thought that closing the schools would be the best step to slow the spread of the virus. It later turned out that children were not powerful vectors, though. I forget the details of the reason. It had something to do with ways kids respiratory tract is different from adults. So what we really needed to identify was the settings in which there was the most contact between adults.
I don’t know what we would have ended up doing if you could have gathered these 10 or so people, funded them and let them run the show, but I’m pretty sure we would have navigated covid far better than we did. Oh, and we should have added a good, honest communicator to the batch to be spokesman (maybe Scott, if he could be a bit folksier. Though he might demand to be spared TV appearances).
It wouldn’t surprise me to hear that some of the people Lewis’s book features are not as admirable as they sound in the book, but even if you shrink all of these people’s smarts and integrity by 50% I think they still would have made a decent team. Or if one is in fact a genuinely bad apple, they could be replaced. You don’t run across people like the ones Lewis talks about every day, but they aren’t as rare as hen’s teeth. I know a couple. You probably do too.
And of course I don’t know how to set up the government so that groups like that are the ones who take care of business.
Regarding crypto regulation, my sense is that actually the rules are quite clear, it's just that crypto companies don't like those rules and would prefer to be subject to different, laxer rules. But just saying that would sound bad, so instead they falsely claim that there is no regulation or that the regulations are ambiguous.
> They did it through the power of nominative determinism. In that year, they elected a governor named Brad Little. His administration is called the Little Administration.
For those who showed interest in English prosody on an earlier open thread, I'll note that the government-diminishing phrase "little administration" has primary stress on the penultimate syllable of "administration", whereas the government-assigning phrase "Little administration" has primary stress on the penultimate syllable of "Little".
Can you confirm or deny the claim that stress placement in English is nearly random? I read this in some book (that there's no handy rules that work nearly all the time, as in e.g. Spanish), but I'm too lazy to think about it.
There are two levels of stress you'd want to consider.
Every word has its own stress which is just part of the dictionary entry. (Or, in other words, word-level stress is an arbitrary fact and the only approach you can take to it is memorization.) Your comparison to Spanish makes me feel you're probably thinking about this kind of stress. At this level, your claim is correct.
Stress (and other prosodic phenomena) are also applied at a higher level, to highlight the grammatical structure of a clause or to focus emphasis on a certain point. This is what I'm talking about in the comment above. This kind of stress is governed by rules. It might be hard to find good literature on those rules, though.
At this point, I'm fine if they just close down the fda entirely.
Close them utterly and then let them go before congress with a rationale for why they do what they do better than the private sector could do it. Let them beg to be allowed to continue their "work of protecting American citizens". Let them explain why so many of their ex employees work for big pharmaceutical companies. Let them justify why they have surpassed information.
And no more lies about being objective and not under the pay/thumb of pharma. We know it isn't true.
And some shady pharma operator, probably not one of the big ones but a marginal one trying to make it big, will start selling a drug that, oops, causes ten thousand horribly malformed flipper-babies to be born, and the public will overwhelmingly vote for whichever political party most convincingly promises to stand up the UberFDA, just like the old one but a thousand percent moreso.
You say that speeding things up requires either preventing lawsuits or changing laws, but of course, one can also just *change the regulations* rather than go through Congress. This probably requires a change of leadership at the regulatory agency, to someone who actually wants to make those changes, but an incoming administration obviously has the power to do that.
Of course, it remains true, as you say, that you don't do this by randomly firing people! If you want to change the regulations, you need spare people to make that change; if you want to automate things, as Charlotte Dune discusses in her comment, you need to hire people to perform that automation (and then afterward you can lay off a bunch of people, but the people doing the automation will probably need to be new hires).
By the way, you actually *can* just reduce what you can sue (say) the EPA for. The US government has sovereign immunity; you can only sue it when it's made a law saying you can sue it. I don't really approve of this principle, but given that it exists, you absolutely can just find what law lets people sue the EPA and cut down what it allows suing over. (Although as other commenters have pointed out, the recent overturning of Chevron deference sure makes a real mess of things. Not that those are directly related, but they're on the same general point...)
See also Matt A's comment about how much bureaucratic process was put in place to remove bureaucrats' decision-making power; people will have to be OK with that! Also see the comments people made about the bloat coming from agencies' *contractors*, rather than the agencies themselves -- in some cases, reducing bloat and improving efficiency may require *growing* the agency so the *contractors* can be fired (Alon Levy talks about this all the time wrt the MTA and other American train agencies).
In the 70s, the government did a traffic analysis to figure out why traffic was coming to a standstill at roughly 16:30 every day. What they found was that 1 guy would get on the highway, merge into the passing lane, set his cruise control to 55mph, and jam up traffic for the next 4 hours. This made local news.
This man was John Nestor. We know this, because, in a massive display of entitlement, he wrote a letter to the editor in the Washington Post identifying himself as the man in question, and justifying his actions by, in brief, asking why he should bother speeding or dealing with merges and the traffic on the right for the benefit of other people?
Now, why does he matter? Because he worked for the FDA. In drug approvals. He applied that same ethic to the drug approvals he oversaw - why should he put himself at risk for benefit of a billion-dollar multinational drug company? The number of approvals he granted over his career was a nice round number. In fact, it was the *roundest* number, 0. If your drug ended up on Nestor's desk, you've lost your entire investment and there's nothing you can do about it.
Despite actively sabotaging the approval process, Nestor couldn't be fired. Eventually, he was shipped out to Iowa and removed from the approval process, so at least there was some improvement to the process and the traffic.
Did I get you? Of course this story doesn't end on a positive note. Ralph Nader launched a campaign to restore Nestor to his position and eventually won, because success must be punished as often and as hard as possible, and failure must be rewarded.
Cutting the bureaucracy in half would remove a lot of Nestors and other bureaucrats on that scale - maybe not everyone will kill everything, but only allowing no side-effect, perfectly safe, no-risk products can also do immense harm to the public. Getting rid of half of the bureaucrats will eliminate half the School of Nestor.
It will also expose the rest. Since fewer people are getting more work, the harmful people will be more obvious, and addressed directly. Eventually, there will not be 1000 bureaucrats processing 10 forms a year, but 50 bureaucrats processing 200.
Or, you know, reducing the number of forms needed for anything to begin with. 1000 bureaucrats with 10 forms apiece for the year can drag out each form and crush the private sector by design or indifferent malice. Musk complains that designing and building a rocket takes less time than getting the approval to launch from 1 desk to another. That is possible because there are too many bureaucrats, so they have time to stall out on processes and approvals.
Most salient example: Bureaucrats demand environmental impact studies to clear out deadwood and underbrush for fire prevention. This process adds several *years* to the process, making the process longer than the fire cycle. So many agencies don't bother trying, because it's just a waste of money that will never actually get to clear out the deadwood. And now LA is burning down, and if they had just let a couple guys with heavy equipment go in and work continually, the fire wouldn't've happened.
The fires in LA are not a Climate Change tragedy or the result of extensive arson. The only factor that mattered was the deadwood, and it was not cleared because the bureaucracy said "Wait. (And by wait we mean no)"
> A woman had written a letter to this newspaper complaining of being tailgated on the Beltway by a large truck whose speed-hungry driver wanted her out of the way. Truck lovers fired back, accusing her of wanton lane-hoggery, and a prolonged epistolary squabble ensued. Still, the whole affair might have expired of inanity. But on Oct. 1, in a quiet corner of the letters section, appeared Nestor's modest manifesto:
"On divided highways I drive in the left lane with my cruise control set at the speed limit of 55 miles per hour because it is usually the smoothest lane. I avoid slower traffic coming in and out from the right, and I avoid resetting the cruise control with every lane change.
Why should I inconvenience myself for someone who wants to speed?"
The public exploded in a partisan frenzy. Nestor's name ("It's as Irish as Paddy's pig -- comes from County Clare") entered the local lexicon: His supporters called themselves "Nestorians," and an editorial minted the term "Nestoring." But for the reluctant eponym, the celebrity rankled. "After all," he protests, "I have really accomplished some important things." Chiefly during his 21 years as a medical officer in the Food and Drug Administration, where he developed a reputation for whistleblowing that makes the Mormon Tabernacle organ sound like a kazoo. Dr. Sidney Wolfe of the Naderite Health Research Group calls him "sort of the ideal public servant."
(Quotation marks around Nestor's manifesto added for clarity.)
I also don't think it's the case that he didn't approve any drugs in his career - just that he didn't approve any between the years of 1968 to 1972.
I'm certainly no partisan of the FDA or of John Nestor.
As for the takeaway of the parent comment, I think it hinges on this claim: "Cutting the bureaucracy in half would remove a lot of Nestors and other bureaucrats on that scale - maybe not everyone will kill everything, but only allowing no side-effect, perfectly safe, no-risk products can also do immense harm to the public. Getting rid of half of the bureaucrats will eliminate half the School of Nestor. It will also expose the rest. Since fewer people are getting more work, the harmful people will be more obvious, and addressed directly."
Getting rid of half of the Nestors obviously does no good if you're also rid of your competent employees - the ratio hasn't changed at all - so you have to believe that firing half of everyone will make the system more transparent to management and make management more inclined and able to remove Nestors. Since I don't see any reason to expect this, I'm going to side with Scott Alexander.
That's a good point. I misremembered the takeaway as "we hate this sort of thing" rather than "yeah let's fire half the people on a random basis", heh.
"so you have to believe that firing half of everyone will make the system more transparent to management and make management more inclined and able to remove Nestors."
From my OP:
"It will also expose the rest. Since fewer people are getting more work, the harmful people will be more obvious, and addressed directly. Eventually, there will not be 1000 bureaucrats processing 10 forms a year, but 50 bureaucrats processing 200."
One person blowing up the drug-approval process can get Congressional hearings and might actually be an election issue, like Fauci. 1000 people blowing up some of the drug-approval will not.
- Red tape amount is related to the number of bureaucrats: If there are far less, this incentivizes moving away from 800 page documents. There are certainly various cases where these are required by law/regulation, but I would not be surprised if there are many that are not actually needed.
- Existing bureaucracies are inefficient. Just because someone said they had a good time with those people doesn't mean it wasn't inefficient! They're just decently sociable. (You can enjoy, for example, talking to a college advisor even though the college is gouging you and the advisor isn't spending more than twenty minutes on you)
- This is an empirical question of how much work is actually done in various parts of government. How much slacking off occurs?
- This will pressure other parts of the government and companies to downsize on redtape. They can focus on the departments that don't keep up, and hopefully actually talk about reducing X,Y,Z requirements.
- It would be better to select who stays based on competency, but I think part of the idea is to avoid goodharting of metrics (and just straight up fraud/exaggeration that ingroups which have existed for decades can have)
- Internal delays. If you have to talk with five different people about five different things, that is going to add delays in how things are handled. If, instead, those five different things are handled by two people? Less delay. Presuming enough slack, of course.
(One issue with some examples of good bureaucrats is they're often top 1%, but this is like saying that because there's top programmers at twitter then cutting half of all programmers won't help any. It does increase workload, but it also avoids a massive amount of internal inertia. Government has more issues due to required inertia via regulations/compliance/forms, however.)
I think the idea behind firing the bureaucrats is to put pressure on Congress to reduce the FDA's mandates. There's a reason why the FDA has to many things to do -- because of pressure from special interests. Well, if the special interests can't get anything done anymore, then they're going to have to give up on some of those things.
Agree with the article in general, but to me raises the question: What would it take to reduce the risk of lawsuits more generally? How to make the country less litigious?
I don't think it's likely to happen. In addition to individual motivations - the 'right to one's day in court' which, though originally from Britain, is stronger in the US than anywhere else - there are collective efforts - activism; green tape; 'lawfare'; lobbying; ever-expanding rights...all building into a massive law/academia/media-admin landscape. And huge numbers of people, at different times, and often with good motives, contribute to the general gridlock.
I think being litigous is actually good. In non-litigous country, seems like problems are still there but solved in extralegal ways. It may be good if the problems are better solved that way, but most of the time it's because the legal officers can't be trusted. A litigous country means that people at least trust the legal system, and you don't really want to reduce that trust.
Loser Pays would be a very good start. Yes, the poor would still be able to sue the rich, presuming the suits had merit; there's several ways that could be arranged depending on how exactly you structure the system.
It would change the incentives that inform settlement negotiations. It's fairly common for defendants to settle cases they're very likely to win at trial because it's cheaper to settle than to go to trial. Plaintiff attorneys know this and it factors into their decisions on whether or not to advise their clients to file cases.
Former (Canadian) bureaucrat here, and I think you're making a prior assumption that may not be true (and was certainly not true in the depart of the Canadian government I worked at back in the day). The number of bureaucrats may not be optimized to the amount of work they have to do. Some of them may not be working very hard, or really very much at all. At least in Canada it is almost impossible to get fired as a unionized government employee. If you don't commit a criminal offence or cause your department to get onto the national news in an embarrassing scandal, you are probably fine even if you don't do any actual work. My experience was that about 5% of the people there did all the work and the other 95% averaged about 2 to 3 hours a week of actual work. The primary reason I left was that the temptation to coast like that was ever-present and I found it to be somewhat soul-destroying.
Whelp, looks like we should get rid of all of them then.
Actually...what if they just shutdown the government and just...didn't open it back up? Veto every budget bill, force congress to veto override or impeach/remove?
Has Musk thought of this yet? Military still gets paid, essential services would still need to be provided. I'm sure 90%+ of the people affected would be democrats...could someone steelman the counter-position here?
The counter-position is that a bunch of good things don't happen that need government approval - for example, no new power plants are approved and no oil wells are approved. If people build the power plants and oil wells anyway, they'll get sued, and then the courts have to deal with it (and they'll shut it down, because it's blatantly illegal and the courts aren't that biased).
I suppose the next version is that you first repeal all the laws banning anything, but that's not happening. Even if you magically did though, you would then get a ton of people mad about people doing stuff they don't like (like dumping chemicals in rivers, building coal power plants in their backyards), and then you would be voted out and everything would be put back in place.
The final version is that you just abolish the government entirely and get rid of the constitution, but again that's not happening, and then you lose the military and immigration control, and presumably you like those. I mean, maybe you're a hardcore libertarian and you at least have a consistent policy, but even Trump and Musk don't want that.
The core issue is that Americans (and people in general) *want* many things to be illegal. The FDA exists because there was a huge public outcry about drugs that were causing harm, and they elected politicians that promised to do something about it, and they created the FDA.
Also, my interaction with the federal government is in the energy industry. The main thing that the bureaucrats are doing that's relevant for us is making everything simpler and faster. Currently every electric company and state has processes that are different from each other in meaningless ways, but building anything across state lines or between different companies is extremely difficult because all these processes differ. The bureaucrats at the Department of Energy are coming up with simpler processes that unify all of the different ones. All of these companies and states can't agree on a single process, and someone at the federal level needs to do it.
So if we get rid of those bureaucrats, the result is that fewer powerlines and power plants are built - not because the federal government needs to approve them, but because the federal government is forcing industry to get on the same table and work together.
The federal government used to train bureaucrats to reduce red tape. They came up with a (forgotten) training course during World War II, to teach bureaucrats to reduce paperwork delays and thereby aid the war effort. I rediscovered this initiative and wrote about it here:
While I can see your point for this particular FDA case, I think it's bad to dismiss the "too much bureaucrat" problem entirely. I still believe that "the bureaucracy is expanding to meet the needs of the expanding bureaucracy" because it still occurs even in pure capitalistic corporations who have to red tape one of their department talking to another, even when no one would really sue them if they don't do it.
It's just a fact that more people is harder to coordinate than less people, and some companies have tried to be leaner to account for it (either by mass layoff, or spinning a new taskforce solely for specific task so they can run faster). It's harder for government to do it because civil servant can't really be fired or demoted, hence the nonsensical tweet. But the problem it tries to solve is a real one.
Can anyone point to good research on how to effectively create and align incentives to get an effective bureaucracy? It's been an issue for...~3000 years? Surely there's been some systematic work by now.
Honestly if we can figure out the bureaucratic alignment problem, the AI alignment problem ought to be easy :P
But optimistically it may have been solved! For all I know Carthage had a perfect bureaucracy but because they couldn't mobilize like the Romans their solution was wiped from the earth forever. Wait, is that optimistic or pessimistic?
> Can anyone point to good research on how to effectively create and align incentives to get an effective bureaucracy? It's been an issue for...~3000 years? Surely there's been some systematic work by now.
Singapore has a famously effective bureaucracy, and LKY talks about some of the bureacratic practices and measures he instilled in his book Third World to First.
For that matter, China has a much better state capacity than the US, as everyone will say SIngapore isn't a good example because of its size. But anyone who's traveled to China or done business there over any appreciable time is amazed at the sheer pace they're able to build roads, power plants, skyscrapers, and everything else.
This seems to be mostly achieved by giving below-head-guy levels more autonomy and pretty clear KPI's to execute towards.
I'm not sure what the level of "legal recourse the average citiizen has against the government" is, but I'd certainly bet on it being "significantly less than in the US or Europe."
So the solution is probably more autonomy, more autocracy, and less legal remedy.
We had that, and then Old Penn Station was torn down, and Robert Moses happened, and then Overton Park v. Volpe happened, and now we don't trust autocratic public planners.
It turned out that there are costs to giving planners autocratic power and the backlash was significant enough that government in the US at national state and local level added emergency brakes all over the process of getting stuff built.
It also turns out that if you add in brakes everywhere, people will use them, whether that's for an intended purpose or not. So the pendulum will swing back and we'll probably spend the next decade or two tearing out the brakes until something suitably popular is destroyed or a vogon constructor fleet shows up.
This is probably right, but I'd like to cut the USA some slack on this problem. The thing is, when you're not the world leader, you have a model that you can follow. When your economy is already at the frontier and working better than anyone else's, the vast majority of things to be done will just make the country worse. And those few things that will make things better are, by definition, untested (in the conditions that exist at the frontier).
There's certainly something to be learned from China, but they mess things up all the time.
— How likely is it that their decision will get challenged in court?
And if it gets challenged in court, what amount of paperwork do they have to show the judge to prove that they made the decision on a “reasonable basis”?
"
But is there any reason to believe the bureaucrats are well calibrated in this regards? Is a 5,000 page 6 year Environmental or FDA review any less likely to trigger a lawsuit then a 50 page 6 month one? Does it make it any more likely for the agency to win the lawsuit? I could easily believe it doesn't matter or maybe even the reverse (I'm sure there's a lot more to nitpick in the 5,000 page version)! Has anyone rigorously looked at this? Are the agencies constantly experimenting (ha! no!) or do all the incentives run towards being overly cautious, to 'doing more' even if that's actively detrimental, or even Molochian standstill (something like CA HSR where they'd get more money and more bureaucrats if they could actually get moving but the whole process is so regulatory borked that the whole thing is likely to get shelved).
I have some experience adjacent to the adversarial end of the bureaucracy (general contractor stuff on military bases but no clearance past the gate), and was always struck that the actual factual federal bureaucratese were all vastly more competent than their private sector and contracted equivalents.
If there were two or three times more government paper pushers at camp pen lets say, they could easily make back their salaries multiple times in efficiencies and cutting out stupid privatization.
Here's a meta-factor that I think you're not taking into account: If you yell at people loudly enough, make it abundantly clear that you can and will swiftly fire them, and tell them to figure out a way to achieve goal X or else, they often change their mindset and figure out a new way to achieve goal X. Because you have significantly changed their incentive structure.
This is not hypothetical. Musk does in fact use this technique. He even got rockets to land on their tails this way, you know?
I'm not making a strong prediction about whether he or Vivek will succeed at DOGE. But I do think your account of the present incentive structure at FDA is not telling the entire story.
"they often change their mindset and figure out a new way to achieve goal X."
If they have the power to do so. There's plenty of examples of people lower down the chain of command advising "this is a bad idea, it will cost too much, it's against regulations" and getting over-ridden by the higher-ups.
It's all very well for me, Acting Grade III, to figure out a new way to achieve goal X. But if the person who can sign off on that is the Grade VIII who doesn't want to do so, then whether or not Vivek fires me won't get that done.
And that is the problem we foresee here: it's not the Grade VIIIs and above who are in the firing line, it's the rank-and-file who are the ones dealing with the public. You can't fire the Principal Officer because they have an iron-clad contract and will take you to court and the case could run for years, but you can fire twelve front line staff because they're disposable. Now you can issue a press release about "number cutting achieved!" but you still won't have improved the services.
I think an overlooked aspect of the excess burden of bureaucracy is an excess number of bureaucracies. Many agencies have authority/responsibilities that overlap with another's; for instance, both the DoJ & the FTC can bring antitrust actions, so private actors have to jump through two similar-but-not-identical sets of hoops.
The crypto industry's complaints are at least partially driven by this: nobody's the exclusive regulator for them so everybody's pestering them (that the SEC and its we'll-tell-you-what's-prohibited-by-suing-you-for-doing-it approach is in the mix definitely exacerbates the situation).
There are places where refactoring the agencies such that at most one had authority over a certain thing could have salutary effect, e.g., carve up the FDA and combine the "food" mission with the USDA and the "drug" mission with the DEA.
I work in private. A oversight group was making various paper work requests of our group that got so burdensome we hired a team with the sole task of doing the paper work. Overtime the team managed to increase the required paper work. The requirements didnt come from them but they would always propose solutions to the oversight groups concerns that were maximumly bureaucratic. Now our group wants to interact with oversight group more directly. So there are some cases where the paperwork does in fact adjust to the number of bureaucrats.
There’s deregulation for its own sake, and then there’s maiming the system & documentation to make space for sycophantic new hires. Who will be “the only ones who can fix it” and will add myriad well hidden regulations that benefit themselves, or whoever they’re hired to benefit. Firing half the workers is probably an effective prelude to this. The code will end up longer, processing times will be longer, but whatever benefits are available will flow reliably toward the leader’s allies. And the old guard is no longer numerous enough to stop them.
I do think they're genuine about their desire to reduce bureaucracy. The party doesn't care what you do as long as you stay loyal to them. Dissenters can be dealt with on a case-by-case basis; no need to establish concrete rules for that.
I agree to a point - I think some of them are genuine in their desire to reduce the size of the bureaucracy. I think of “reduce the size of the bureaucracy” as an abstraction that can obviously be carried out in multiple ways. People who really like that abstraction clearly may or may not have thought about how to actually make it happen in a certain instance. Also if it were an easy problem it would already be done. So I think the people with the genuine desires for smaller government are the tip of the spear, and their earnestness will open up a whole bunch of gaps which will be filled by people with slightly different goals who want certain outcomes more than they want small government. The Idaho example is interesting because like many other commenters I think it doesn’t scale.
The harder we pull the string, the tighter the knot gets!
But seriously. Something Idaho did right was having each department cut their own code. They’re more likely to know which parts actually matter. But over a certain threshold - whatever it is, the feds probably passed it long ago - the code is so large you’d have to hire more people to revise it (removing that advantage of familiarity). Even county land use codes - just one little area of code - this is the part I’ve seen - a hundred pages of regulations and one wants to run screaming from the building “it’s alive!” The two people who wrote most of it last time know how it hangs together and for everyone else it’s …. Time consuming.
Also complexity is part of community life and some rule systems are legitimately complex.
If Ramaswamy was serious about smallifying bureaucracy he would set up a commission (I know) to find all Those Two Guys who wrote the code, bring them all back from whatever they’re doing now, sit them down, pay them handsomely and task them with making it elegant. That’s a plan I could get behind. The people who know what’s actually in it are the only ones whose edits will not make everything worse.
Cutting chunks off the document and firing half the staff doesn’t actually help. It’s not the first, second or third step in any process that will actually smallify effectively. It’s not an exact parallel to computer programming but one useful part of the analogy is that these regulation writers don’t comment their code and it’s not always self-explanatory even though it’s in the common language. (Yay lawsuits for clarification purposes).
For social media purposes, though, saying “bureaucrats first against the wall when the revolution comes” sounds compelling to some audiences. I appreciate Scott forever for his willingness to gently and thoroughly explore the actual results of this type of thing. I’d put him on the editing commission (he wouldn’t do it though!) because he has the ability to consider how to move language around to create structures in society. Not everyone has that. The incoming folks will try to streamline the government and they will prove things about systems and unintended consequences … probably often.
That all sounds pretty plausible to me---I just have a knee-jerk reaction to things that sound like "well, we can't ever go in the 'make it smaller' direction" (because I often hear that from people who profess to /agree/ that that would be better, and it's like "okay, so we're doomed forever: if any attempt to fix it makes it worse, might as well just let it balloon insanely for eternity, I guess?!", heh).
This is why I don't favour libertarianism, it replaces rule by elected representatives with rule by lawsuit—a clearly less efficient system. Reduce the scope of civil courts and increase the scope* of elected representatives (e.g. Governor Little).
*In practice they already have the scope they just need to *want* to do it.
I think this presupposes a degree of benevolence from the bureaucrats. I work for a Department of Energy contractor doing environmental remediation. The bureaucrats I deal with speak in terms of benevolence (“save tax payer money,” “achieve site cleanup”) but act in ways detrimental to those stated goals. More and more resources get tied up in talking about something than actually accomplishing anything. And when you act to achieve an objective, the bureaucrats second and third guess your decisions. Prolonging the endeavor is the bureaucrat’s objective, regardless of their stated intent. This is further exacerbated by decision by committee to avoid any responsibility or accountability.
Chiming in with first hand experience that generally aligns with this analysis.
I am a former bureaucrat who was part of approving international aid funding. I sat between the bureaucrats doing programming and the Ministers office (political staff) signing off on the funding.
Red tape is part of the democratic process. If something stupid gets funded, the Minister or Prime Minister wears it and could take a beating at the ballot box. What if the FDA approves a bad drug or former Taliban was allowed to immigrate? The court of public opinion and the news media create incentives for more red tape to protect against backlash. It’s easier to accept backlash for slow speed vs bad outcome.
We had a specific file that the Minister rejected for political reasons and tried to push onto the bureaucrats. When the political decision finally came to light, all those forms and double checks ensured the person who made the decision was held to account.
Secondly, I was tasked with simplifying the overall approval process from the top. Cutting red tape requires someone to deeply understand why something was added in the first place then efficiently condensing things. Chesterton’s Fence is a useful framework here. People like Vik appear to think everyone but them is a dummy. To me, he is just another form of elitist arrogance.
Absolutely agree. My experience with people doing programming/grants/etc was a desire to move as quickly as possible. The political level are the ones who often get squeezed for a mistake then run it downhill. Ironically, leading to a new piece of red tape somewhere.
I think the core conflict in the US is actually between Yale Law School grads and engineers from Cal. One group has been endlessly creative interpreting laws and procedures in novel ways to advance their goals while the other tries to create new things.
Probably that he doesn't actually want to solve the theoretical problem, he's just throwing red meat to the hogs.
Dude made his money by hyping drugs that had already failed in trials then dipping when it came time to carry the bag, why expect him to change his ways now?
The feeling most people have about civil service workers is that they don’t work very hard, because they don’t have to — they are almost impossible to fire if they just make sure they cover their asses, so that’s what they spend their energy on.
I honestly have no idea how true that perception is, but I know it’s pretty widespread. If you start with that, it’s easy to believe that laying off half the staff would encourage those remaining to do their job like it mattered to their own personal future, rather than act like time-servers and paper-pushers.
I think there’s also a widespread sense that government agencies spend an unreasonable fraction of their time doing things that are not actually in their mandate (DEI, for instance, or the CDC deciding that racism is a disease) and even things that are arguably not properly authorized by the Constitution in the first place. It *might* be that cutting departments in half would force them to scale back what they do to include only their legitimate duties, but I fear there is a step missing in that argument — more likely what you have to do is explicitly and thoughtfully cut back on their missions and then set staffing to match.
I have a vague intuition that the biggest problem with bureaucracy is about who will eventually have to take the blame if a decision is incorrect? Often, when you deal with the “bureaucrati in front of you, their hands are tied by a higher authority, and they’re probably grateful for that. But the same idea persists up the layers; a perfect paper trail, a diligent assigning of responsibilities, and the whole process completely infected with CYA, which I think most of us can understand to some extent. Ultimately the buck stops at some big shots’ desk, and big shots come and go; they resign, they reappear…
> who will eventually have to take the blame if a decision is incorrect?
That's what scapegoats are for. There's always someone that everyone hates that you can blame everything on. Or you can plan ahead and hire dedicated scapegoats that can be sacrificed when things go wrong!
Why would it encourage them to work harder, when the criteria for being fired is not "least hardworking" but "wrong social security number"?
Firing civil servants for not doing their job, is not actually impossible but it is very very hard and it will almost certainly remain so no matter what Trump et al do - that's baked deep into the system. Laying off civil servants because you've decided their organizations are oversized, or changing the rules re mandatory retirement, or firing them for some weird new reason nobody has bothered to set rules and precedents against, will be much easier.
So that's what will be done, and what is done will not serve as an incentive for the remaining civil servants to work harder. Except perhaps on polishing their resumes and networks.
But I generally agree. We might hope that a growing sense that their jobs are not nearly as safe as they thought they were might spur them to work harder, on the grounds that if mere bad luck can strike them where they thought they were untouchable, maybe slacking would also be a bad idea. But I agree the model is simplistic.
I suspect you might get better results by firing everybody in a target organization and then staff it back up to 50% levels; layoffs who could point to solid accomplishments would be more likely to get rehired. Maybe?
But I'm afraid there are really only two ways to reduce government to half its current levels. One is to have the consensus will to spend a century doing it, bit by bot, as we have spent a century getting into our current fix. The other involves M. Guillotine. Neither is very likely.
I think it's definitely a stupid rule/worried about getting in trouble problem and not an intrinsically obstructionist bureaucrat problem. I have worked for county and municipal governments in Texas for the last 10 years (as a sysadmin and now cybersecurity) and the descriptions of how the federal government works or how the government of San Francisco works have always baffled me. My experience has been that when we needed to do something we sort of just did it. It didn't take 5 years. We were not worried some strange activist group would sue us. At most we did have certain state or federal rules we had to follow (CJIS - criminal justice information systems standards, say) and an audit every X years. Or if something was really expensive, you would have to wait until October to put in on the next years budget. But it did not take us 10 years of review to build a facility or 1 million dollars to build a public toilet in a park.
I think population density and the density of existing infrastructure would have an impact on that, although to what extent I would not know. Building a tunnel under the east river of Manhattan is a different proposition than doing something of that scale where you are working, I would imagine.
I think so but Europe builds subways way faster and for way cheaper than Manhattan, and I don't know why it should cost 1 million dollars to build one toilet even in San Francisco.
One of the cities I worked for had an on paper population of 13000, but a day-time population of over 100,000 because it was in a very popular shopping area and also had several major employers (a couple of hospitals). There was a big stretch of city road that was having a lot of wrecks from people making risky left turns. They decided to build a median down the whole length. The whole process of doing that from planing to completion was like a couple years at most.
Another place was an exurb with a population of about 50,000. I worked there for about 6 years. During that time, the city built 1. a brand new fire training facility, 2. a huge new fire station and police sub station on the West side of town, 3. an incubator facility for local small businesses to rent office space. They also remodeled the public library, bought and remodeled a building for the growing planning and engineering department, and then remodeled the old planning building and gave it to us (IT) because we had run out of space for storage where we were formerly squished in the server room at the old PD.
This is all within the Houston area. Not as dense as New York or LA but hardly the country either.
I agree with you about the insolence of office and the laws delay. The reasons for it are often perverse and difficult to understand. I can almost guarantee that it’s working for someone somewhere.
I sometimes wonder about the amount of damage Western society today faces because of its own wealth and generations of capital accumulation at the family level. When your family is rich enough for you to be genuinely comfortable your entire life without doing anything productive, but you aren't the brightest bulb, so you end up joining these 'advocacy' organizations. And because lots of rich people know people like this in their circles, this is going to continue to go on, despite how parasitic it is.
It is easier for staff to repeal regulations than it is to implement them. It is also easier to ignore them than it is to enforce them.
The problem with work exceeding the capacity of bureaucrats is both imaginary (it doesnt actually exist anywhere) and false. The real problem is they are sitting around trying to think of things to pile a regulation on. Thinking of a person to file a lawsuit against.
Javier Milei fired a bunch of federal bureaucrats and slashed regulations and size of government. Now Argentina's inflation has dropped from 300% annually to 20% while the economy expanded 4.5%. This seems like proof that getting rid of unnecessary bureaucracy can be a good thing, no?
First off, the Argentinian system was burdened by an outrageously bloated state bureaucracy and an irresponsible fiscal policy that resulted in several national bankruptcies and caused the absurdly high inflation rates. So in that case, yes, dramatic measures are probably needed. AFAICT, the US is nowhere near that situation. Yes, there was some inflation, but that happened in Europe as well, probably due to similar issues - Covid, the Ukraine war and other factors disrupting supply cains and economic activity. No, a national bankruptcy is not in the cards unless Congress is extremely foolish. An unemployment rate of 4% is also not particularly worrisome. So, to take an analogy, while there's good evidence that chemotherapy can save lives, you probably shouldn't prescribe it if the patient doesn't actually have cancer.
Second, where do you get the number of +4.5% from? What I read is that they have barely managed to stop the bleeding, and are hoping for a recovery this year. Also, unemployment and poverty rates did go up, as expected. This may be necessary hardships on the way to recovery, but that recovery hasn't fully manifested yet.
All I know is that I bought into some Argentinian companies and my buddy got into Argentina-heavy ETFs on my advice when Milei was elected, and they've made us fantastic returns—better than nearly any other positions in the portfolios.
Does this indicate Argentinian economic strength? I dunno; that was my impression, and why I bought in, but if not I can live with that too (the profit helps).
Milei has a lot more authority to get rid of rules and regulations than the US President. Most things the federal government does, it is mandated to do by Congress (and that includes the writing of regulations).
In practice, I doubt that the number of forms provided has any impact on the outcome of a case, and if it does it will surely be at least logarithmic to the number of forms. But there's no incentive NOT to have one more form, so up it goes.
Scott, I'd love you to do an article looking at this on the UK side. We have the Lower Thames Crossing (linked below) - a tunnel whose application document alone has broken records, at over 350,000 pages and the application process costing over £300m.
Are some bureaucrats needed? Of course. But any system that creates this outcome is utterly broken beyond doubt.
Plainly you can save money by employing fewer bureaucrats, but if you want to make government more efficient you have to reduce the number of forms and regulations in parallel with those layoffs.
There is a new problem in all this: no one has ever had to prune government employed bureaucrats to a significant degree. Let me give an example: when I moved to Canada in 1985, hospitals managed their own budget and were run by volunteer community boards. Physicians gave their time to sit upon committees, unpaid (and had a vested interest in keeping them to a minimum). My province's government decided to save money by employing administrators to 'professionalize' the running of hospitals, and have several levels of management to co-ordinate services so that there was no duplication. What could possibly go wrong?
My community hospital went from 29 beds, 24/7 ER,X-ray,and lab run by 2-3 physicians and one administrator to 8 beds, ER 8am-4pm on weekdays, same for x-ray and no lab. But has five physicians who are salaried and see very few patients, and 14 administrators.
The regional hospital went from 100 beds and four administrators to 50 beds and 74 administrators.
Attempts to streamline this result in new vice-presidents being appointed (along with support staff) to "oversee change." And patient care spaces get taken over for offices for them. Administrators never reduce their empires, but always expand. It is like watching an old oak tree being smothered with ivy, parasites sucking the life out of the body. We haven't faced this before, and have no good idea of how to reverse the process. But we have to figure it out and do it, or else we shall see the old joke come to life that "this hospital would run better if there were no patients in it."
I'm sure it's much the same everywhere, and the same kind of thing is happening in the NHS in the UK. Bureaucracy grows like a cancer, with much the same effects on the host.
Sorry if this has already been said in the comments, but as a lawyer, I found the "what if they get sued" section to be disappointingly beneath your usual standards of rigor, Scott. The case you cited for the "reasonableness" standard is not an employment case, but a claim under the Administrative Procedure Act, regarding the standards a federal agency must meet when issuing regulations and responding to citizen petitions about issuing or amending regulations. (Google "CRS petitions for rulemaking"). Instead, reductions in force of federal employees are governed by an extremely complex set of rules that probably only a government lawyer or union rep could explain (Google "RIF rules"). Except to the extent a such a DOGE-esque layoff relied on an agency interpretation of the rules governing reductions in force, the APA would be irrelevant. If anything, Vivek's tweet is another example of how he and Elon seem blithely unaware of how much law they'll need to tangle with in order to achieve their goals.
"Fire half of the bureaucrats" is obviously a simple solution to a complex problem. That proclivity is not a conservative trait nor a liberal trait. It's a human trait. It was incredibly effective for cavemen.
The problem is that our subconscious mind evolved for the animal to survive and thrive and its environment. It took millions of years for it to evolve to the point where caveman was well served. It hasn't evolved much past. However, our rational capacities have skyrocketed in the last 100,000 years. What we have today is it chimpanzee attempting to operate a supercomputer.
The brain essentially operates just like a computer - garbage in, garbage out. Every computer programmer thinks that what he's written is solid until he tries to run it. The programmer attempts to blame someone or anyone other than themselves for the multiple crashes. Then comes the arduous process of trying to locate and fix all of the bugs. But the computer is built to be an incredibly rational entity. The human brain is built for every dog to snatch his big a piece of meat as he can and bite anybody who tries to interfere. So the problem is that our cognition makes decisions based on far too little information and grossly lacks the predilection to do sufficient research to do any better.
In reality, the problem isn't due to Congress, the federal bureaucracies, the massive number of regulations, or the selfish idiots voting these selfish idiots into office. The problem is human nature. Very few people possess the cognitive capacity and skills necessary to do any better than what they're doing. The result is a shit show that no one cares to take responsibility for.
I challenge every reader to present a solution to that problem. I guarantee it's deeply errant if the proposed answer is less than 57 pages.
Einstein found the problem that resulted in general and special relativity sitting in a chair imagining what light produced in a moving train looked like to both someone in the train and someone on the platform. That probably took 10 minutes at most. But it required thousands of pages of complicated mathematics to prove it.
Great article, but my only gripe is this: Everyone wants a simplified process for getting drugs approved, but criticizing the number of “processes” and “forms” without specifying exactly which processes should be eliminated or simplified isn’t that helpful
People are very good at identifying problems with the status quo but bad at thinking about trade offs. The FDA that Scott seems to want would be more opaque, arbitrary, have more lenient quality standards. There would be some upsides to that, but I’m very skeptical that the public actually wants that.
It’s also much easier said than done from a regulatory and statutory perspective.
I don't think "more opaque and arbitrary" is something Scott wants for the FDA.
More generally, I think it's meaningful to say "we err on the side of too many regulations and not enough approvals" without reviewing every extant regulation and process; in past posts, Scott's made a case for why the FDA ought be more lenient, which involved things like calculating cost vs. benefit as it stands now vs. hypothetically approving more drugs (e.g., what has the cost been from delaying good drugs vs. approving bad ones, and what is this likely to look like with X% more lenient standards).
In other words, this is the sort of big-picture question which could go either way, and which posts of Scott's were spurred—IIRC—by a few people writing articles saying it does go the other way ("FDA needs to be stricter")—so it's not a meaningless question, even without saying exactly which pages of the 20,000 page* FDA Guidelines for Drug Approval need to go. Building consensus that yeah, this probably needs to be fixed in a certain direction /is/ helpful.
>but I’m very skeptical that the public actually wants that. [...] It’s also much easier said than done from a regulatory and statutory perspective
Yeah, Scott's aware—cf. the bit about people "yelling at" the FDA, which is treated in greater detail in the other posts on this question.
I think it's important to remember why the FDA (and similar bodies internationally) were established; it's because in the halcyon pre-regulation days, private enterprise was putting literal poison into products sold to the general public as food, drink and medicine. And literally killing their customers:
"The swill milk scandal was a major adulterated food scandal in the state of New York in the 1850s. The New York Times reported an estimate that in one year, 8,000 infants died from swill milk.
Swill milk referred to milk from cows fed swill which was residual mash from nearby distilleries. The milk was whitened with plaster of Paris, thickened with starch and eggs, and hued with molasses.
After the extraction of alcohol from the macerated grain, the residual mash still contains nutrients. Therefore, keeping cows stabled near distilleries and feeding them with swill was an economic advantage.
...In May 1858, Frank Leslie's Illustrated Newspaper did a landmark exposé of the distillery-dairies of Manhattan and Brooklyn that marketed so-called swill milk that came from cows fed on distillery waste and then adulterated with water, eggs, flour, and other ingredients that increased the volume and masked the adulteration. Swill milk dairies were noted for their filthy conditions and overpowering stench both caused by the close confinement of hundreds (sometimes thousands) of cows in narrow stalls where, once farmers tied them, they would stay for the rest of their lives, often standing in their own manure, covered with flies and sores, and suffering from a range of virulent diseases. These cows were fed boiling distillery waste, often leaving the cows with rotting teeth and other maladies. The milk drawn from the cows was routinely adulterated with water, rotten eggs, flour, burnt sugar, and other adulterants with the finished product then marketed falsely as "pure country milk" or "Orange County Milk"."
That background, and high profile cases like thalidomide, selects for a high degree of cautiousness in any government entity - see the quote from the Department of Labor about "out of an abundance of caution and to avoid any doubt as to its validity" above. So you triple-check everything and make sure all possible 'i's are dotted and 't's are crossed, even if that takes longer.
It seems like the problem here was one of "information liquity": apparently, no one made the connection between the milk and infant deaths (one wonders where the "8,000" number came from---controlled for general conditions among poor New Yorkers?) because "germ theory was in its infancy" and (presumably) because people were generally ill-formed at the time.
I bet such a strategy as "selling diseased, rotten milk that kills thousands of infants" wouldn't be very viable in today's America... although---actually, conditions in factory farms even these days make me question that, come to think of it---
Lord Rama *is* associated with cutting a lot of people in half, though. Probably some of them were bureaucrats!
Less crucially, but maybe it'll be relevant on some object level or whatever, I think you're modelling Ramaswamy's motivations wrong. I think he just wants government to not have to pay half of those people's salaries. Possibly he overestimates what proportion of the federal government consists of bureaucrats, or maybe he's just like everyone else on this point and hates bureaucrats more than he hates soldiers or medicine.
As for the issue itself, I think the root problem is that America chose litigiousness as its solution for the tension between liberty and responsibility. This probably made a lot of sense at the time! However, as time passes and the monetary stakes go through the roof, the value of optimizing for gaming these systems also shoots up, sometimes far past the value of e.g. creating new products. Then the same ends up true of the government, developing new bullshit counters becomes legitimately more productive than solving people's remaining, often intractable problems – and now we have huge bureaucracies, cost disease and medical insurance so good it makes people homicidal.
Other countries don't seem to have this problem or not to the same degree. Admittedly part of it is probably because they just don't have the wealth and industry – Belgium isn't making all the new drugs, so their FDA only needs to regulate the importation of drugs somebody else already proved safe at huge bureaucratic expense – but a lot of the rest seems to consist of just being perfectly content banning things that Americans would chafe at. No bureaucracy, the politicians just pass a law saying you can't own scissors under the age of 30 or some stupid shit like that, and people just *accept that*, and in exchange, if you decapitate yourself with your scissors you can't sue the company because you were either of legal scissor-owning age and should have known better, or you had those scissors illegally and it's not their problem.
I work in the electrical industry, working with a bunch of different groups including electrical utilities, transmission operators (the people who actually run the cross-state electrical grids), state regulatory bodies, and the Department of Energy. My experience is that a lot of what the Department of Energy has been doing, particularly in the last four years, has been making it easier to build.
Each member of these groups has their own idiosyncrasies and different rules and procedures for developers who want to build things like power plants or power lines or data centers. If I want to build a power line from California to New Mexico, through Arizona, I have to worry about the processes of three states' regulatory bodies, plus the processes of the people I'm connecting to on both ends. The bureaucrats at the Department of Energy have been doing a lot of really important work in simplifying these processes and forcing the different groups to all work in the same way.
Getting rid of the federal bureaucrats doesn't mean that there's less red tape, it means that there's *more* red tape, it's just from private entities and state bureaucrats instead. It doesn't make any sense to get rid of a federal bureaucrat if it means that every state in the nation has to hire one to replace it, and then every business has to learn fifty different processes instead of just one.
"If I want to build a power line from California to New Mexico"
then God help you, given that California seems to want to slap warning labels about "this is a carcinogen" on everything including air and water? 😁
I think a lot of us who have worked in/adjacent to the public service have similar war stories about how it's much different on the inside than the commonplace image associated with it.
"red tape expands to consume the resources available to it"
There is a HUGE amount of truth to this.
And in the opposite direction, the absence of resources does indeed drive process simplification (and shortcuts). This shouldn't be dismissed out of hand.
I would argue that well considered process streamlining has been historically rare in government; especially when compared to changes driven by resource limitations.
Could another thing be labeling each regulation with an individual person responsible, like a git repository? If the specific element is challenged successfully, then the person who put it there is blamed, with suitably severe punishment to further disincentivize people to expand regulation?
I more meant the person who actually wrote the regulation. Not the “make a regulation” guy (which is congress), but the guy who hammered out the details.
That may not be as simple as you think; just looking at some new rules on the Federal Register, one is from the IRS, one is from the fisheries service to do with tuna catch allocations, and I'm not terribly sure who gets the blame or credit for signing off on that one.
"Karen H. Abrams, Acting Director, Office of Sustainable Fisheries, National Marine Fisheries Service."
But the allocations are the result of a series of agreements:
"Atlantic BFT fisheries are managed under the 2006 Consolidated HMS Fishery Management Plan (FMP) and its amendments, pursuant to the Magnuson-Stevens Fishery Conservation and Management Act (Magnuson-Stevens Act; 16 U.S.C. 1801 et seq.) and consistent with the Atlantic Tunas Convention Act (ATCA; 16 U.S.C. 971 et seq.). ATCA is the implementing statute for binding recommendations of the International Commission for the Conservation of Atlantic Tunas. HMS implementing regulations are at 50 CFR part 635. Section 635.27(a) divides the U.S. BFT quota, established by the International Commission for the Conservation of Atlantic Tunas (ICCAT) and as implemented by the United States among the various domestic fishing categories, per the allocations established in the 2006 Consolidated HMS FMP and its amendments. NMFS is required under the Magnuson-Stevens Act at 16 U.S.C. 1854(g)(1)(D) to provide U.S. fishing vessels with a reasonable opportunity to harvest quotas under relevant international fishery agreements such as the ICCAT Convention, which is implemented domestically pursuant to ATCA."
So if I disagree with transferring the tuna quota from December to January, who do I go after as responsible in the end?
Sometimes there are named individuals - as here. But as here also, those people seem to have moved on or been transferred:
"...(W)e are affirming and ratifying a prior action by Jane Oates, Assistant Secretary for Employment and Training, and Nancy Leppink, Deputy Administrator of the Wage and Hour Division.
...The Final Rule was signed by Assistant Secretary Oates and Deputy Administrator Leppink. We have full and complete knowledge of the Final Rule action taken by former Assistant Secretary Oates and former Deputy Administrator Leppink. Subsequent to the Secretary of Homeland Security's documented approval of the Final Rule dated November 25, 2024, in consultation with the Secretary of Labor and Secretary of Agriculture, and out of an abundance of caution and to avoid any doubt as to its validity, we have independently evaluated the Final Rule and the basis for adopting it."
So even if you're angry about letting migrant workers in to work on the harvest, Oates and Leppink are now former officials, so how do you punish them?
Probably whoever “we” is. Current person who ratified it. So there always needs to be a person standing behind it, who is willing to stake their personal reputation on the rule being good. It would require an ID based commit system, like git. Whoever committed the bad code is the person responsible.
I think that would come naturally! Ie. whatever group they work under would eventually stop assigning work to them because they'd be afraid of institutional shaming.
>possibly illegal (does the administrative branch even control how judicial procedure works?)
Not the Executive, but Congress could strip lower court jurisdiction over cases where someone sues the Federal government via the Exceptions Clause (Article III, Section 2), allowing only the Supreme Court to take such cases. The Supreme Court hears <100 cases/year, so this will reduce the burden on the government by ~99%. There's also tricks involving International treaties (Missouri v. Holland), but again this requires Congress to cooperate.
That being said I'm sure the Executive could still do lots of good if they applied every trick in the book to overcome the bureaucracy, even if it won't last forever.
On the one hand the argument is strong, on the other hand having incentive system of adding bureaucrats until you get the results you want has some obvious failure modes and cutting the staff by half with the message that you expect them to streamline things creates credible incentives to actually streamline.
"the message that you expect them to streamline things creates credible incentives to actually streamline."
Yeah, we had a guy who streamlined like heck in the old job in the social housing department. Turns out he just promised everybody who applied for a house in the projected new development "yes". Naturally, this ended up with 100 people all expecting to get their promised houses in the 25 accommodation development. Not his problem, because he had quit to go work elsewhere before the chickens came home to roost.
That sure streamlined things, but it didn't in fact solve any problems. What it ended up doing was *cause* problems, because he slipped up and made the mistake of sending a letter to one of these applicants in effect telling them "yes" (generally it had been verbal assurances) and so the applicant was able to go to court with this as evidence and force the council to give them the house.
Very popular decision with the rest of the people on the waiting list, as you might imagine.
I could have streamlined application processing for the grants by telling everyone "yes". Problem is, there wasn't the money for that. Streamlining is great *IF* the *entire* process gets streamlined (e.g. yes we will be building 200 new houses and apartments to come onstream by this time next year) but not so much if it doesn't, and the public-facing level of "I process your application" civil/public servant does not have the final say on "sure guys, send the diggers in to start on the foundations!" Not even the head of the department has, that decision rests finally with the holder of the purse strings, which is the national government in the capital.
It also creates credible incentives to look for a job in the private industry, since they pay better and may be less arbitrary in who they fire. Guess who'll be the first ones to leave?
Also, the number of public service employees who have any competencies to streamline anything is probably pretty small. "Incentivising" (or actually, threatening) THEM might make a bit of a difference - for the rest, all that remains is the demoralization.
As an urban planner who has worked on trying to streamline approvals, I really appreciate this post. People too often argue about small or large government, and there’s not nearly enough discussion on how to set up regulations so they’re less inefficient.
For three years I’ve been working on getting a housing development approved that has the support of politicians, the public, and city staff. It has even been approved by council! But still individual staff members who get the file have found reasons to delay the project by years. There is just no incentive for them to do otherwise.
One day I will run for city council and will attempt to set up better incentives.
Reminds me of David Graeber's book Utopia of Rules. In it he says something like: Once a bureaucracy is established, it's almost impossible to get rid of it. Looking at hundreds of case studies didn't reveal a single instance of red tape getting removed; it only ever gets added. To paraphrase Max Weber, the only effective way to get rid of a bureaucracy is, really, to kill them all, which is what Genghis Khan did.
Vivek's proposal reminds me of the joke ascribed to Pope John XXIII; when asked "How many people work in the Vatican?" he allegedly replied "About half".
The problem, of course, is to identify *which* half. Cost-cutting and headcount reduction proposals are nothing new! But the catch is, it's never the Assistant Deputy Vice Junior Secretary levels of management which get cut, it's the low-ranking grades that are public facing.
And then people complain about "I can't get through to a live human on the phone; my application has been in limbo for weeks; why does it take so long to process a simple request?"
Well, because the staff were cut in half and not replaced, but the volume of work remained the same and may even have increased. I saw this with my own two lying eyes in the public service in the wake of the Croke Park Agreement:
"According to the implementation body, as of March 2012, headcount had been reduced by 28,000 and the annual pay bill had been reduced by €3.1 billion. Over 7,000 public servants were reportedly redeployed."
Yay hurrah, efficiency increased and bloat reduced, right? But what does that look like on the ground? On the ground, in the office were I was, it was a colleague reduced to the point of nearly having a breakdown and needing to go out on long-term medical leave due to stress. You see, a new scheme for rent assistance had been spun up from the pilot programme and was now live nationwide, and she was doing it for our area. On top of all her other work. With no assistance, dropped into "this is starting now and you are handling all the applications". And getting all the hassle about "why aren't people getting their payments?" when it wasn't her fault that the information was not being processed, because it was all piling up on someone *else's* desk in our city office and not being put through the system.
And had she gone out on (deserved) sick leave, the files would have just piled up on her desk and nobody else would have been able to deal with them, since there was no slack or redundancy in the new, lean and mean, cut costs and cut headcount, no new recruitment when people leave or retire, regime. A conscientious, hard-working woman literally reduced to tears because there was too much to do, not enough time in the day to do it, and all the shit was rolling downhill to her, who had little to no control over what was going on.
And then the public continue with the idea of 'lazy civil servants who don't care, are just clock-watching, and are inefficient' because their application still hasn't been dealt with after six months. It's not because the person responsible is lazy and inefficient, it's that they are over-burdened and driven to the end of their rope.
That's the problem with all these top-down schemes that come from Big Bright Ideas - nobody ever asks the people on the ground, the customer and public-facing staff who are the interface between the bureaucracy and the clients, about what they think or how it will work out. What are the obvious points of failure, what won't work as intended (I still fondly remember "new software package to implement online housing applications that won't accept names with apostrophes - in a country where people are called O'Brien and O'Mahony and other such names").
I'm not denying there is waste, inefficiency, and featherbedding in the public service as in any other industry. I am asking "Sure, fire the half who don't do the job - but make sure you know which half that is, by a better metric than 'SSN ends in odd number'".
When there's a cutback, the things cut back are always the public-facing stuff. The managers never get cut, just the people who actually do stuff the public cares about.
So your application takes longer to be approved, the park is closed, the welfare office has reduced hours. But the number of managers is still the same; they've moved them to a different office, or they're presiding over a reduced number of reports.
Will the Cabinet get reduced in number? Will the number of agencies get reduced, or their overlap in function? Probably not.
"Will the Cabinet get reduced in number? Will the number of agencies get reduced, or their overlap in function? Probably not."
Oh yeah, we all know that. The irony here being particularly thick, since forming a new government agency (DOGE) means carving out its own territory and hiring new staff to run it 😁 Well, I suppose they could transfer some of the "your SSN ends in an odd number" laid off staff to work there!
"But the article suggests that ordinary people’s lives were made easier, and that the move has brought businesses to Idaho. It gives the example of repealing a regulation about what kind of doors pharmacies can have."
Out of curiosity I looked that up, and the old regulations were about security when dispensing filled prescriptions to be picked up outside opening hours - so basically "make sure junkies or thieves can't break in to steal the drugs":
"27.01.01 State Board of Pharmacy Rules of the Idaho State Board of Pharmacy Section 605 Page 72
03. Storage for Delivery.
Filled prescriptions may be picked up for delivery from a pharmacy when the pharmacy is closed for business if:
b. The secured delivery area has walls that extend to the roof and solid core or metal doors, and all doors and other access points must be equipped with locking devices and be constructed in a manner so that the hinge hardware is tamper-proof when closed;"
The new regulations are much more stripped-down, but I imagine Idaho pharmacies still adhere to "make sure doors and locks are secure" practices from the old regulations, because imagine if the pharmacy is broken into, and the insurer refuses to pay on the basis of "a five year old could have picked that lock"?
The regulation might have good intentions, but there are externalities. Where I live, it’s okay for a public facing business to have an entrance that is wheelchair accessible with a little help (eg, a small step). Same with bathrooms. But if that same business asks for ANY permitting to make even a minor change, they are required to build a ramp and re-do their bathrooms. So businesses either don’t make improvements, or do it without permitting, or have to spend a lot more money than they might be able to afford. This happened a particular business which I frequent, which is an old school gym. I’ve spent time there and I’ve never seen someone in a wheelchair. Also I would imagine there are better alternatives for someone in that situation than an Olympic weightlifting gym. And I’m sure if there was a particular person with physical limitations who really wanted to pursue their dream of Olympic weightlifting, the staff would be welcoming and assist as necessary.
I think the ADA does many great things, and I’m totally on board with new building construction rules. Our cities are becoming more ADA compatible every year, and that’s good. And because we are on that trend line, maybe we need fewer rules.
Likewise, there’s a difference between having a practical recommendation for pharmacies which they might be eager to adopt, and a regulation, which leads to profiteering by the one steel door maker in town.
You can also incentivize more directly, for example by punishing pharmacies that get broken into due to poor security, but not if they found a way to manage. Similarly to how the worst penalties for drunk driving actually occur if there’s an incident.
To understand bureaucracy, everyone needs to read https://www.eatingpolicy.com/ (and Recoding America) by Jennifer Pahlka. It is absolutely brilliant and, saying this as a civil servant, perfectly accurate.
From all I hear about the federal bureaucracy, it is practically impossible to fire people, so actually doing work is voluntary. About half the employees choose to do so, at varying levels of effort.
So firing the half that doesn't work would be a big improvement. I think that is where the DOGE guys are coming from.
Of course, before you do that, you need to somehow make working mandatory.
They also plan to remove tons of regulations, which in either case would require far fewer bureaucrats.
A lack of conceptual clarity in Vivek's view is the ultimate goal.
Is it to make things which the government wants to happen, happen more quickly? Then in some cases, you might need to hire /more/ civil servants. (This is Scott's point about # of drugs approved per FDA employee.) This applies to whatever it is you might want to do. E.g., nobody seriously thinks that if Trump wants to deport large numbers of immigrants, the best way to do that would be to first fire half of the ICE agents.
Or is it for the government to spend less money? At the outset, it's worth noting that all civilian federal employee salaries & benefits together are a very small portion of the federal budget. If you want real cost savings, you'd have to make major cuts in entitlement programs (Social Security & Medicare, mainly) and the Department of Defense procurement budget. Also worth noting is that reducing federal employees often /increases/ the amount of money the government spends, because their work is still legally required to be done, and so they get replaced by outside contractors (Booz Allen, McKinsey, etc.), who (1) charge more, and (2) whose spending should be reviewed by federal employees to make sure they're not overcharging, except if you cut the federal employees whose job is to make sure that the contractors aren't overcharging, then you'll /definitely/ get more contractors overcharging, because unlike federal employees who receive a fixed salary, outside contractors have a profit incentive to overbill..
For almost any other goal you might think of, "fire half the employees, at random" is VERY unlikely to lead you there. (Unless you define the goal tautologically as reducing the number of employees, no matter what the downsides and without reference to any purported benefits.)
This is not to say that there aren't problems with many (perhaps most federal, state, and local government agencies. But these critiques are already well-known and understood within those who actually look at the problem. (Cf. Scott's old post on "Yes, We Have Noticed the Skulls.") Serious people who have examined government operations (Don Moynihan, Jennifer Pahlka, the Partnership for Public Service, etc.) have suggested important reforms to improve both effectiveness and efficiency. (For example, modernizing procurement regulations for software to reflect how software is developed nowadays.) But Vivek's tweets are more like the ramblings of a drunk at a bar.
Many people are not aware of this fact, but comparing to Soviet Union, number of government bureaucrats in modern Russia increased, and it's still less bureaucracy per capita than in EU. Because governance is hard and requires human resources.
I am occasionally in the position where I need to claim expenses from finance offices in various public bodies. Most of the time this works reasonably well: there are electronic forms to fill with my bank details, I submit receipts and eventually I'm reimbursed. That's how the system should work.
There's one body that I deal with which:
- uses a paper based system, is spread over multiple locations and allows some of the workers to process paperwork from home offices
- hasn't updated their processes to allow for expenses to non-employees (e.g. one person asks for my employee number every time, and insists I need permission from their CEO to travel anywhere)
- has allowed a low-level employee to email my employer asking blankly whether I'm committing fraud
- regularly re-litigates expenses claims paid 6-12 months ago
- takes up to a year to pay out claims, never pays an amount the equals my claims, doesn't notify me when I'm paid, and seems to have no record of what is and isn't paid
Most people have experienced the second type of process. To the people on the inside it's probably largely indistinguishable from the first kind. I think the post overlooks this type of poorly performing process. Fixing bottlenecks like this absolutely could speed up all sorts of things. (I don't think Elon and Vivek have the discernment to identify these types of issues and reform them without damaging all sorts of well-functioning processes, however.)
You make interesting points from first principle thinking, but I think you’re missing the tacit knowledge of having managed a large organization.
People in organizations want to feel important. You might think that delivering results makes individuals feel important but this is the exception for a typical organization. That’s because delivering results feels like everyone collaborating together towards a single goal and that requires an outstanding leader that squashes distractions. This involves both being careful with planning but also firing a lot of people who naturally impede collaborative progress.
What happens most often at most organizations, is that employees who don’t have status try to feel important by creating new layers of complexity. The canonical example these days is people trying to turn woke viewpoints into company policy, like going through codebases and removing the term “black list”. This makes them feel important. Because it’s much easier to come up with reasons to tell people not to do something, this sort of effort is almost always in the form of blocking others from acting.
The DEI example is salient today, but large organizations are full of these distracting or blocking efforts. People trying to unionize, employees preventing Google from working with pentagon, organized outrage of some form or another, and so on.
Having fewer people, and importantly, having the possibility that there will be even fewer people in the near term, makes those that remain focus more on results and keeping their jobs. And importantly, it also makes the kind of people who seek status in unproductive ways to leave, because they can tell their antics are no good there.
Regardless of Rama’s explicit argument and his experience to justify it, I’m almost certain fewer people will within a medium term speed up for government.
Every $1 of spending on IRA enforcement activities results in $5 - $9 of return. There exist situations where MORE government is more efficient. General rules are hard to infer.
You're defining "efficiency" in terms of maximizing the government's net income. That is not the goal of any reasonable person who is not running a government. The goal is to maximize the prosperity of the nation.
If that $1 if spending results in $9 of "return" but also $20 of reduced private-sector productivity, e.g. because people who should be doing useful stuff are instead chasing down the paperwork for the latest audit, then the nation is less prosperous. If because of that lost productivity, people are laid off and the government is on the hook for $9 more in unemployment benefits or whatnot, then even the government is losing on the deal.
Or maybe those aren't the actual numbers. But you're leaving some of the most important terms out of the equation altogether, and in so do revealing something ugly about your true intentions.
You can disagree about what taxes should even exist, sure, but this is just mitigating tax evasion. There is no extra work to be done by accountants than just normally filing tax returns and paying the bills. If you want to make private companies more efficient by reducing the paperwork you should counter-lobhy turbo tax so that the government is allowed to tell you how much you owe in taxes (which they know but are not allowed to tell you!!)
Maybe so re the FDA, but as to the corporate sector and vast number of years involved in corporate restructuring elimination of people generally meant elimination of layers of unnecessary bureaucracy.
The bonfire of "red tape" (British-speak for the 1000-forms) here in the UK during the Cameron government of 2010- was a significant contributory cause of a fatal fire for 72 people in West London in 2017
Here's a summary from one of the UK news magazines:
The point is that improvements in fire safety regulation was delayed precisely because of a "Little Administration-esque" attempt to prevent new regulation from being drawn-up without other regulations being 'burned'.
Lord Rama should careful what he wishes for - he may accidentally invoke Lord Shiva.
I like Scott's articles, including this one, but he is being quite naive here. Ask yourself who gains if the FDA is cut in half? Politics is about the distribution of power--there is no constituency for across the board objective efficiency.
Of course, you can ask the same question but going the other way. Who gains if the FDA is doubled in size?
I think you are maybe missing the point of why ramaswamey said something idiotic (but i can't tell): he's posting a tweet and getting responses. The republican party has a major ideological tenent against "bureaucracy", and he is only appealing to this.
As far as "D.O.G.E" is concerned, it appears that it probably won't exist, as elan musk can't create a government agency out of thin air, only a private bureaucracy.
I think you are correct to be skeptical of how all this works. All this nonsense with the national debt and the budget is just nonsense, and politicians who talk about how bad it is generally have some hidden motive for doing it.
Sorry if I am a buzzkill, but this whole thing seems to be a lot simpler than you all are making it out to be...
I work with government bureaucrats most federal, but also a lot of state/city/county. Daily.
>My model goes:
...“processing 1,000 forms”.
Suppose they have 100 bureaucrats, and each bureaucrat can process 10 forms per year.
...
If you fire half the bureaucrats, now they can only approve one drug every 2 years.
That’s worse!
Scott,
This is a poor model. If you have 100 bureaucrats, a good half of them are doing next to no work period. They are effectively processing 0 forms per year, or even a negative amount by bogging down meetings with stupid questions, not completing tasks they were assigned, or socializing with productive people.
They also have the corrosive effect of otherwise productive people seeing them doing next to nothing, not getting punished, so now they too sit around all day doing fantasy football and personal email.
I had a city once hire me to "office space" a department. It had ~80 employees. I interviewed everyone and made notes about their tasks and how those tasks went based on objective measures. They ended up doing some early retirements and a lot of sending people to other departments, and afterwards the 45 remaining people got much MORE work done. From 80 to 45 and it worked better, much better actually.
Less circular finger pointing about why nothing was happening on X or Y. Less items stalled "waiting for Sally to respond to an email for 3 months". Frankly less socializing and standing around the water cooler.
Now some of this is of course layoffs let everyone know you actually mean business. But a lot of it is certain people in government are literally a net negative to the functioning of the office. And while the private sector is not some paragon of efficiency, in the private sector these people are generally found and culled a lot faster. A city can just raise you taxes and hire another person. Or more often, not hire another person and city function X just doesn't happen, or takes 6 months instead of 2 weeks.
Not all the government staff are like this, there are also a lot of overworked super diligent people. And an issue honestly with unfilled positions too, though that again can come back to the same thing (with the HR bureaucracy taking forever).
That said Ramaswamy's plan is stupid. You need to sort the wheat from the chaff. And there is so so much chaff.
I would probably fire the bottom 60%, increase the payscale dramatically, and replace a lot, though not all of them with much better people. Right now important positions managing like multi-billion dollar programs are being managed by government staff who make like $140k/year in DC. You don't see that in the private sector so much.
That isn't even getting into the whole issue with the legislature not needing to pass things that actually work or make a lick of sense. They can just legislate nonsense if they want, and often do. And then the staff need to figure out how to make nonsense work. Once again less common in the private sector because the people at the top are accountable to the same thing as everyone else (dollars and cents) unlike in government where the people at the top have an incentive system that is totally different from actually doing anything effectively.
(also if you happen to read this and also remember that retail pharmacy explainer I pitched in 2018, I published the most recent version of it in 2023 if you want to take a look)
"Upon being read, many of the [Idaho] regulations were not justifiable, for example “rules for a lottery game show that never aired”"
Removing such inoperative rule looks nice of the graph, but decreases the bureaucracy by exactly 0 (as it was inoperative and applied to noone).
Also to be noted that those regular reviews of all regulations have to be done by someone, meaning the same bureaucrats searching for something to cancelled. Which inevitably leads to creation of "department for decreasing of administrative burden" with KPIs, yearly reports, workplans, etc
Re: Idaho, one dynamic that gets missed in these sorts of clear statistics (i.e. number of individual regulations or pages of regulation) are the messy ways the bureaucracies themselves adapt to maintain their status quo.
At a prior job, I had to interact with Arizona Dept of Health during Gov. Ducey’s administration. He’d put into effect many of the same rules you mentioned that Idaho has - re-justifying every regulation every n years, removing x regulations for every new regulation, etc.. The end result, however, was that AZ DHS just made up a bunch of unwritten, illegal “non-rules” that they held you to anyway. In one case, it meant them publishing an “interpretive and explanatory guide” that was twice as long as the actual regulations and included a significant number of completely made up new requirements.
When arguing with that division’s chief about it, she flat admitted that they had to do so because they were restricted from passing new rules. In the end, it made what sounded on paper like some kind of libertarian dream, in fact so dysfunctional that I told Ducey’s policy chief I’d rather open a new site in Chicago or Los Angeles than ever opening another one in Arizona.
None of this contradicts your argument. If anything, I think it bolsters the point that “if we just fire half the bureaucrats we’ll get half the red tape” is laughable bullshit.
My cousin (American citizen) is married to an Australian. He (the Australian) has actually lived here and legally worked here before. They moved to Australia for a bit and for some abstruse reason nobody understands, he now can no longer legally live and work here anymore. They've been sorting his paperwork trying to move back for like 3 years.
They might want to consider just moving to Austraila permanently, considering that the border might get closed by the time the paperwork gets done...
I wonder if it's something along the lines here, from the Indiana University web page:
https://ois.iu.edu/scholars/permanent-residence/marriage/common-questions.html
"I have heard that it may be difficult to travel after I have married an American. Is that true?
Yes. It may not seem logical, but you have to be very careful about travel if you marry an American or green card holder.
If you are in F, J, or tourist status, you are expected to have a home abroad and the intent to return there. When you apply for a visa stamp at a U.S. embassy or consulate, you have to prove ties to your home country. Marriage to a U.S. citizen makes proving these ties difficult, because the assumption is that you will want to immigrate to the United States. If you need a new visa stamp in order to return to the United States, the likelihood that your will visa will be denied is strong—even if you plan to continue as a full-time student."
I don't think that's it because he was like 40 years old at the time he moved over here. He worked for UPS and they had a child here. It was extremely apparent he wasn't a tourist or student. I don't know what kind of paperwork he had at the time.
Now, they did meet in Hawaii so maybe he had some kind of tourism visa when he was there.
The question is not "does DOGE have the power to fire these employees?", it's "can Elon and Trump find the people that can disable their badges and disable their computer, and will congress respond when they do?"
Deal
And all the things you want to be done to Make America Great Again will not be done because the regulations requiring that they have eighty-seven different kinds of permission will still exist but the people with the authority to sign the permission slips will be on an extended paid vacation.
You know you can just... do things without asking for permission right? Especially when you're immune to the law due to having a guaranteed pardon from the president.
You actually can't. Let's say you don't feel that you need a building permit? Fine, but the construction company still won't touch it. You decide to sell an untested drug? Will still have a serious problem getting pharmacists to carry it.
Also, what about when they come after you in 4.1 years after you actually didn't get that preemptive presidential pardon because Trump would rather play golf that day?
>he thinks we're gonna care about the permission slips
You're going to care about permission slips. Because "damn the permission slips, I know Donald and he'll for sure pardon me if some stupid bureaucrat tries to have me arrested" is going to be very much a minority opinion, and the things you are going to want to do or have done are going to require the help of people who are going to be more "show me your permission slip, so I know I won't be going to jail in three years when the Democrats retake power".
I cannot tell you how much I hope you guys actually go ahead with all this, and don't just talk about doing it.
DOGE will have exactly 0 formal power as currently designed.
This is very likely deliberate - a way to shuffle Musk and Vivek off into what is little more than a think-tank, away from actual power.
The IRS is the source of revenue for the rest of the federal government, so I think I'd prefer it does things efficiently.
Queuing theory shows it's actually much worse than doubling the time, because a task doesn't just take the time it needs for someone to work on it, it also takes the time it needs for someone to be free to work on it. Basically once the amount of tasks slightly exceeds the number of bureaucrats, the expected time taken shoots off to infinity. To solve that, one needs excess bureaucrats.
Thank you for referencing queueing theory here. Understanding the basics can really illuminate situations that appear very counterintuitive.
Yes. There is basically a Micawber Principle for any system with a queue:
Admissions 20 per hour, discharges 21 per hour, result happiness; admissions 20 per hour, discharges 19 per hour, result misery.
> admissions 20 per hour, discharges 19 per hour, result misery.
This isn't really a result about queueing. If the amount of work to be done exceeds the amount of work you're able to do, then work is going to remain undone. How you organize the work will never make any difference. What you're losing the time to isn't queueing, it's incapacity.
Note that in your parent comment:
> once the amount of tasks slightly exceeds the number of bureaucrats, the expected time taken shoots off to infinity
this is a queueing result; what's necessary is that tasks outnumber bureaucrats, not that the time required to complete the tasks exceeds your available bureaucrat-hours.
The mathematical study of queuing definitely includes statements the time taken by tasks, without which it would be difficult to do any calculations about queue length. The fact that a queue will grow to infinity if the arrival rate exceeds the processing rate is an elementary result that you will find in most introductory queuing theory texts.
It's so elementary that actual elementary school students have no trouble giving a rigorous proof. You don't even need a queueing concept. If nothing ever queues, you get exactly the same result, although the identity of which work gets done may change.
The point we want people to grasp does involve queues, and is not intuitive: there is a discontinuity at arrival rate=processing rate, in queue length or waiting time. since waiting time is infinite at arrival rate>processing rate, it must tend to infinity as arrival rate tends to processing rate. This is important as it means that 100% utilisation is a bad goal. But it's a goal that is often aimed for in the name of efficiency.
The 'Micawber principle' is trivial as stated, but it allows people for whom talk of infinities and discontinuities would be offputting,to generate the correct intuition naturally. expenditure being too close to income is bad, (the original Micawber principle), so too is arrival rate approaching processing rate ("result misery").
> The point we want people to grasp does involve queues
Sure. But the point you keep talking about doesn't. There's no relation. Your annual expenditures don't change whether you pay for things before getting them *or afterward*.
> This is important as it means that 100% utilisation is a bad goal. But it's a goal that is often aimed for in the name of efficiency.
I don't think this is true anywhere except political rhetoric.
I think they are trying to say that even if the intake is less than the nominal capacity queues still grow non linearly as you approach 100% utilisation if the system has dependent events and randomness in process time.
A toy example:
A worker feeds product into a machine. The machine can't work unless fed product (dependent events). The machine can process 1 a minute and the human doing the loading averages loading 1 a minute but the exact time varies, sometimes 30 seconds, sometimes 90 (random process time)
When the worker is slow, the machine has nothing to do, so capacity is lost. The worker may feed the next products faster than average but the machine can't work faster to make up the lost time.
In one sense the machine and the worker really can process one product per minute. In another sense when you account for the lost capacity above they can't and the closer you get to trying to operate at 1/min the more the queue will build up.
The market will respond by not trying to approve as many drugs (because the wait isn't worth it). Which will prevent unifinte queues, but also the survival of thousands of patients.
This is it! People get sick, people go on holidays, a new programme is introduced which means a flood of new applicants to deal with -you need some slack in the system to handle it.
If you've calculated "on the volume of current work, we only need N admin staff in this office", then you're up the creek without a paddle when the number becomes "N - 1 or more", or the volume of work increases. Files *do* pile up on desks because the one person in the office dealing with this is out and there is no-one free to take over, due to the New Efficiency Drive which meant that when the experienced people retired, no new hires were taken on and trained in.
Penny wise, pound foolish.
As a former bureaucrat, I just want to say thank you for a great informal articulation of how bureaucracy actually works.
Yes, this is very good. I am merely married to a bureaucrat (but one who worked at OIRA, the beating heart of American bureaucracy!)
A few things I'd note:
- the first Trump admin already did the "2 for 1" approach to regulation https://federalnewsnetwork.com/agency-oversight/2018/01/trumps-regulation-czar-touts-success-of-two-for-one-executive-order/ -- this was painful at the time, and there was lots of cheating/use of technicalities to keep things from grinding to a halt. Last I heard people were floating "10 for 1" for this incoming administration. This is consistent with the hyperbolic nature of DOGE pronouncements, too--it's all likely to come down to earth, but not guaranteed to since there are a lot of insane tyros about to be put in power.
- decreasing regulatory page count is an ok metric I suppose, but in addition to clearing out unused cruft (that lottery reg was probably not doing much harm) and some justifiable wins about saloon doors, I would be very curious to see how much of this got done by aligning Idaho regs with external standards by reference. It's a lot shorter to say "use this trade association's safety code" than to write it all out! At past jobs I have spent time analyzing state legislative output and though I don't mean to knock Idaho, I think it's fair to say that small states' government is often surprisingly threadbare. Many are part-time, and it's totally normal for legislators to have to share staff https://ballotpedia.org/States_with_a_full-time_legislature. All this is to say that what Idaho's doing might be good or bad, but I suspect it's something others have thought of, and might well only be practical because of how small the state is.
- a surprising number of those legislatively mandated reports don't get written on time (or at all). It's something Congress throws into laws left and right, and agencies sometimes decide to accept the risk of their boss getting yelled at at an oversight hearing when there isn't enough time to get everything done.
- more work has been done to avoid unnecessary thrash than you might imagine--"economically significant" regulations receive higher levels of scrutiny, for instance
- it's not just nuisance lawsuits: advocacy groups have been doing their part to make rulemaking worse by encouraging the public to spam dockets with thousands of nearly-identical comments about hot button issues. bureaucrats are typically required to review and respond to every substantive comment in the course of the process. they can ignore the junk, but just sorting through it takes time--and of course a lot of it isn't "junk" per se but is prosaic and obvious, but still must be accounted for.
Regarding page count, I rewrote a procedural manual for my job not long ago. I *easily* could have made it 50% shorter... by using more jargon and less clear language, cutting out explanations and examples, and ignoring rare edge cases. I don't think that would made it better! But it would satisfy the demands of idiots that want to cut page counts in half.
Grinding things to a halt is the point, though. Our government has been mandated to do too many things -- more things than we can afford. Something will give. The only choice we have is in exactly what breaks and when it breaks. We have a choice now, but if we wait much longer, we won't have a choice. What cannot go on, will not go on.
I object to saying we can't afford it - it's not like the EPA or the FDA have massive slices of the budget. The FDA's budget is $7.2 billion - which isn't nothing, but it's about 0.1% of the budget. We absolutely could scrape up another $7 billion if we decided we wanted the FDA to get stuff done twice as fast, we just don't have the political will because the Republicans are obsessed with "starving the beast."
Heck, Trump's tax plan is expected to add 3 *trillion* dollars to the deficit over 10 years - how is it possible that we can afford that, but we can't afford a few billion a year to make the bureaucracy function smoothly?
The "twice as fast" is tricky for the FDA, though. They passed the Covid vaccines fast, now some people are arguing those same vaccines are killing healthy young people by inducing cardiac complications and that these dangerous vaccines were rushed through too fast without proper checking. There is some evidence of the mRNA vaccines being implicated:
https://www.bhf.org.uk/informationsupport/heart-matters-magazine/news/coronavirus-and-your-health/myocarditis-and-covid-19-vaccines-should-you-be-worried
Damned if you do, damned if you don't.
We need to spend less. That means having fewer laws and less enforcement. That means fewer people, but it also means less harm to the people. It's not realistic to think that everything the government does is helpful. Plenty of examples of special-interest legislation.
Special interests legislation simply means "two factions compromised." There isn't any other kind of legislation. Regulations exist because they serve the interests of one or more factions of the public. Cutting regulation simply means choosing which factions get screwed.
That's very optimistic, and not very realistic.
"It gives the example of repealing a regulation about what kind of doors pharmacies can have."
I just get the feeling that this was downstream of the Americans With Disabilities Act, and that pharmacies were required to have wheelchair accessible doors, which meant that there had to be a new regulation put in place about "your door must be of this type". I also wonder if there was a lawsuit or some sort of activist campaign, because generally that is what causes such specific regulations to be passed.
A lot of bureaucratic inertia is caused by badly written regulations which are left up to interpretation *plus* ambulance chasing law firms being all too willing to go "Did the local council refuse to rebuild your entire house because you are entitled to have the door frames expanded by 0.5 cm? Call us for legal advice now, you may be entitled to compensation!" advertising.
This means that bureaucrats are *very* hesitant to make decisions which are not backed up in triplicate by CYA, as if you make a decision which grants something, this will cost money and moreover set a precedent for others to claim the same, which means the Department is unhappy with you. But if you decide to refuse the thing, there is the possibility of a lawsuit which may well go against you, create a precedent for others to claim the same, and cost even more what with damages and fees, which means the Department is unhappy with you.
I have seen people absolutely trying to game the system re: the width of doors being too small to let a wheelchair through, so that's why I have my suspicions about that particular piece of legislation.
Oh, the ADA is definitely gamed. You take an unscrupulous attorney, and a cooperative disabled person, and go looking for any business with any ADA violation. The disabled person attempts to enter and claims that their rights were denied by the business. Then the attorney sues them and immediately offers to settle for thousands of dollars.
It could be any business even if the disabled person would otherwise have no reason to enter.
Never going to be any lack of unscrupulous lawyers, either. If Vivek turned his sights on the legal profession, he'd probably have near-unanimous public support, but he went to law school himself, didn't he? 😀
Yes 100%. Scott's understanding of the administrative state has deepened dramatically over the last few years it's really impressive.
As a former biltong business owner, im glad someone sees my pain at the insane bureaucracy around getting succulent cured meats to the people
I've heard that it's bad out there! Thankfully geelvet droewors and wet-cut biltong is only a short drive away for me...
> is not really associated with smallness.
He IS associated with ambitious infrastructure projects though, https://en.wikipedia.org/wiki/Rama's_Bridge, and bureaucracy holds those up too.
Also associated with not transgressing social limits, which is why he put aside his wife Sita for rumours among the people as to her being unchaste when captured by Ravanna.
One time in history when a king really should have gone "Stuff you, I'm the king, I am above the law, and if you don't like it then lump it, peasants". Ah, well!
Rama is also associated with getting rid of people who don't deserve it (Sita).
So Ramaswamy's tweet sounds in character.
Too lazy to hunt the citation, but the Niskanen Center had a piece where a fellow pointed out that the raw number of federal employees had not grown much in the past 60 years. Growth has come from contractors bureaucrats hire, not bureaucrats per se.
So if anything we have been experimenting with the Vivek’s proposal already. Grow the scope and scale of the federal government without growing its headcount or capacity to deal with congress’ shirking of legislative responsibility to clarify statutes, and see what happens. Seems to be bad things to me!
Anyone with experience/knowledge of transit + infrastructure knows that it’s a fundamental mistake to hire consultants instead of just growing the bureaucracy to match needed expertise…
Instead of just hiring a core group of competent engineers, the NYC transit agency just outsources to consultants, who then hire contractors. No one is incentivized to care about efficiency or cost. Every party to the infrastructure work *but transit users* benefits from wasteful spending. So New York builds for 5-10x the cost of the Swedes
"Anyone with experience/knowledge of transit + infrastructure knows that it’s a fundamental mistake to hire consultants instead of just growing the bureaucracy..."
Not true if you are the owner of the outside consultant group being hired. \S
I expect they know it. It's just not to their advantage.
"No one is incentivized to care about efficiency or cost"
It's worse than that -- people are incentivized to inflate costs. I've worked on government funded research projects and from a manager's perspective it is worse to underspend the grant than to fail to meet the original goals. You're punished for being more efficient than you originally planned for.
Though I partially agree with you, I'll try to give the other side of the argument. I manage some government contracts, and when my management tells me to make sure all the money gets spent, part of what they really mean is "money is very tight, don't ask for more than you're going to use, because if you don't need it we can use it somewhere else."
Underspending is frowned upon because budget people at every level share this concern. Inflating the cost at the last minute to make sure you spend everything is a particularly underhanded and wasteful solution to keep from getting dressed down by the boss; the preferred solution is to make a good estimate before the project starts. Unfortunately it's very hard to carve out flexibility for something like a research project where costs are inherently unpredictable. The system demands a budget in advance, and demands you follow it.
That's not really my experience. There is concern that it "looks bad" if you don't spend everything because it means you can't properly estimate your work and costs, and that will reflect poorly when your next proposal is being reviewed. Or, even worse, for ongoing stuff your budget might get cut in the future.
I would agree the grant sin my world are all "spend all the money or you are a bad performer", they NEVER want money back. There is some minor pressure for cost reasonableness, but only on the front end. On the back end they just want all the money gone and to know that you vague accomplished what the plan was.
Yep, clawback. That's why at the end of the financial year, if there's anything left in the budget, you are encouraged to spend! spend! spend! since not alone will the excess be taken back, but next year your budget will be reduced.
And if next year you have extra unforeseen expenditure? Too bad, you managed to be too efficient for your own good. So better to buy stuff you don't need now (because you're not allowed to keep the saved amount on hand to cover anything else or for next year) rather than lose the money going forward.
It's one of those "looks good on paper" ideas that works out terribly in practice.
Obviously YMMV depending on the domain, but sometimes it’s not even that the contractor is particularly bad or inefficient. It’s that the government ends up spending almost as much effort “overseeing” the contractor (and the entire contract bid and award process) as they would if they just did the work themselves.
The bit I took from Edward Snowden's book about NSA contractors probably applies to bureaucrats in general.
There's rules on how many employees NSA can have and how much they can pay them, because these are federal employees. So what do you do when there's more work than employees, and you also want to keep your best people but can't pay them market rate?
After a stint as a proper NSA-employed agent, you let them switch to becoming contractors and hire them back at a higher rate. That's why Snowden was officially working for Booz Allen Hamilton, in a NSA facility with the same job and security clearance as before.
I guess they hire consultants when some regulation stops them from growing the bureaucracy or paying the internal experts as much as they'd like?
Yes, this is a very underrated factor: I'd be more keen to stop them hiring management consultants and contractors at vastly inflated wages than to arbitrarily cut headcount (especially if you're not cutting the actual legal risk or bureaucratic requirements first).
>Yes, this is a very underrated factor: I'd be more keen to stop them hiring management consultants and contractors at vastly inflated wages
If you did this much of the government would cease to function.
Ding ding ding.
That's the problem with the top-down approach by politicians and government ministers versus the on-the-ground view of the people tasked with implementing the shiny new policies.
Not to mention that some of those aforesaid politicians and ministers have nice cosy relationships with the outside consultants, be it "they donate to my election campaign" or "my brother-in-law's cousin runs this".
"outsources to consultants, who then hire contractors. No one is incentivized to care about efficiency or cost."
You should see the saga of the Irish new children's hospital. This has been going on for decades at this point, and it's *still* not ready yet, and the fancy contractors who won the contract are still "yeah it's gonna be late and over-budget even more":
https://en.wikipedia.org/wiki/New_children%27s_hospital
"The Chief Officer of the NPHDB, David Gunning, reported in July 2023 to the Oireachtas Committee on Health that the costs of merely building the hospital, excluding the costs of commissioning it for use, were now expected to approach €2.2 billion due to further overruns. Only 27 of the 3,000 rooms had been completed, and the expected building completion date had been revised to May 2024. By October 2023, that date had been revised again to late October 2024, and it was emphasised that the last date of the building phase would be the first date of an "operational commissioning phase", so the hospital would not actually open before April 2025. The government confirmed in February 2024 that the total sanctioned budget had now reached €2.24 billion, and that this would now be the "maximum allocation", with no further funding to be put towards the project. By May 2024, the date for "substantial completion" of building had been revised to February 2025, with a further six months before becoming operational. By October 2024, the completion date had again been revised to June 2025."
Bear in mind, they initially lumbered into action on this in 2006. Building didn't start until 2016.
So how much of that overrun was due to lawfare vs. changes requested/mandated by the government vs. underestimation by the contractors?
Yes.
A beautiful stew of "government has no frickin' clue what it wants; changes its mind about location several times; entities already on-site take lawsuits against putting new building on that site; lots of consultants, lots of reports, no action for years; finally decide they'll build it on this site, plus two satellite sites (already a change from "one all-purpose centre"); contractors start building and then Covid lockdown hits in 2020; now it's "yeah it'll be open in January February June well sometime" and the budget has spiralled from €650 million to €2 billion and that's probably not going to be the final cost".
That's not even counting "and when it's open, half the wards will be shut due to staffing limitations" which is what I anticipate to happen. Saw it with our own local regional hospital; fancy new cardiac ward got built after massive fund-raising effort, then - was closed for ages because there wasn't the funding in the budget to staff it. Thankfully, that has been sorted out now, but I'm going to believe in the shiny new promises when I see them fulfilled.
Yeah, a new building means a politician gets their name on it. Money for staff, or even cleaning the building? Not sexy, no publicity (unless it's bad publicity for spending too much), cut it.
Most of the military forces in the world are like that too. Lots of money for new tanks and planes, nothing for maintaining the ones they have.
You need consultants because often they are the only people who know what they are doing. The DEI hires in the bureaucracy can't actually get anything done, so you bring in the consultants to actually do the heavy lifting.
I agree a better system would be to pay those consultants less and actually make them federal employees, but many of them couldn't get jobs with the federal government because they don't meet the "desired" criteria (veterans/disabled/minorities/etc.).
If you don't think it is DEI related I would love to introduce you to the racial/gender demographics of the consultants versus the staff in the two main departments I know well. Advancement is a real non-starter for white males unless they are super exceptional, so where do they all end up? Consulting making more than the bureaucrats and doing their jobs for them. It is a pattern that is clear as day, but no one ever wants to talk about it. There is a huge brain drain where the top staff tend to leave after 3-5 years and go work at like Deloitte or some other big contractor because there is simply more money to be made and they will actually pay competent people whereas the government will even promote them much less pay them.
I work in a bureaucracy (though not the US federal one).
Headcount is restricted by budget, and which bucket their compensation comes from. It's a lot easier to write a justification for hiring a contractor "for 6 months" to fill a perceived need than to go through the process of hiring an employee. Many of those employees start out temporary as well.
Of course those contractor positions then get extended. Subject to budget, though budgeted money gets moved around to different buckets all the time.
The way a friend of mine in federal government explained it to me, it's hard to hire new people above about a ~$100k salary, and the maximum pay is well below corporate rates. If you need some sort of tech service, you have to hire a contractor from Booz Allen or wherever. Of course, making it easier to hire $250k/year government employees has its own problems.
Assuming the work exists and has to be done, you can hire a 100k employee. Or you can hire a contractor. But you'll pay 200k or so for that contractor.
Why? Well, when the work is done, you can stop paying the contractor. So you have to pay for their retirement/sick days/vacation somehow. Not to mention the contracting agency's cut. And the days the contractor is looking for another contract.
Of course, this assumes you're hiring the contractor for a specific project, as opposed to hiring the contractor because there's a hiring freeze on employees.
In my specific area there's a "sunshine" law, where employees above 100k compensation are on a publicly available list. 100k was specified in the law, so as the years go by (almost 30 now) inflation makes the list grow a lot.
People look at 100k and think it's a lot, so they put extra rules around it. I just looked and apparently 100k has lost roughly *half* its value since that law went into effect...
I've been travelling a lot recently, and I've noticed that fines in some countries have their own units that can then be centrally adjusted without needing to change the legislation. Researching a little bit, it seems to be mostly ex-Soviet countries, although I have seen it in Colombia as well.
For example in Kazakhstan they have the MCI: https://en.wikipedia.org/wiki/Monthly_calculation_index, and a fine might be 100 MCI, which gets adjusted every year.
To me this seems like a no-brainer, at least in a limited fashion, I'd be interested in understanding the drawbacks.
I'm aware that some places have income-based fines as well, but I think that's more of a political question, whereas keeping fines and thresholds inflation-adjusted seems pretty much strictly better than having to double them periodically? The main argument I can think against it is that fines getting eroded creates a sort of automatic sunset clause, but it has the inverse problem for thresholds.
The downside of this is that the government gets to be inflation-protected in terms of getting income from fines, and likely fees as well.
Since inflation is caused by the government, this seems somewhat unfair.
inflation is entirely caused by the government? Or do you mean more that, fundamentally, inflation of significance is. I ask because my understanding of inflation was that it occurs when, society wide, people bid up the price of goods. Thus a federal reserve increase in the money supply causes inflation by increasing ability to spend, but a society wide increase in wealth or looser lending standards do as well.
The US government pays out a $255 death benefit through SS if your spouse dies. It hasn't been adjusted for inflation since the 1940s. Since the law is still on the books, we are still paying bureaucrats to process claims for this 1940s death benefit.
That's the case in Germany. The unit is "Tagessatz", which essentially is a day of your suspected salary for a day. This also has the benefit of automatically adjusting to your income, so that punishment hurts somewhat equally (though losing 10 days as a plumber isn't exactly the same a losing 10 days as Jeff Bezos, even though Jeff Bezos would be paying millions in theory, he would barely notice, whereas the plumber most likely does).
Australia does this too. Fines are written in the legislation as x penalty units, with the monetary value of a penalty unit adjusted each year in response to inflation. Note the exact value of penalty units varies between federal and state and between each state.
https://en.wikipedia.org/wiki/Penalty_unit
The context of the conversation was about Eric Snowden, and why the federal government hires Booz Allen contractors as sys admins in the first place. My friend's answer was that the government salary is uncompetitive, you couldn't get the talent you need.
Almost no one in the federal government at the professional level makes what they would in the private sector. As an extreme example, the Secretary of Defense, for example, makes $246k a year. I don't want to pretend to know what being SecDef entails, but I'm guessing they should make more than a 20-something project manager at a tech company.
No they wouldn't, but then they're not really in it for the money. Or not money as a wage anyway.
Now expand this to the rest of the federal workforce. I'm not sure a lot of entry level engineers are motivated by the prestige of working for government as GS-7s.
I think what happened instead is, you only get people who are already independently wealthy and just want the prestige, at least when it comes to secdef. or at least, the whole system is biased in that direction.
I tend to think that's what's going to happen here if they do ax a bunch of civil servants through relocations, etc. They'll end up having to hire a ton of people back, except now they'll be employees of politically connected contractors and the government will be spending more on them with even less accountability or ability to control the work.
It's not obvious to me that contractors have less accountability than civil servants do.
Contractors can do campaign donations to protect their funding streams, unlike departments of the federal government.
I wouldn't say less accountability, but they do have different incentives. Even contractors who always work full time on government projects are not part of civil service culture and are not aligned with the agency's mission the way federal employees should be (ideally). There are other advantages to using them though, like budget flexibility, scaling up and down rapidly, availability of niche expertise, and (at least in the cases I deal with) lower employee turnover. Whether it's actually cheaper to contract a job out or do it in house is very tough to figure out.
interesting!
My initial impression is that 'civil service culture' is part of the problem, from my time evaluating fed-funded ngos and their interface with the gov't... but I don't actually know if i have a good reason for thinking this, and it could be pretty localized to my field. I am interested to hear why your impression is the opposite
I'm sure it varies a lot. The people I work with are generally quite capable, and are at least motivated in the sense that they want to achieve the things they're supposed to be working on, rather than just out to further their own careers at their employer's expense. On the other hand, there is a culture of being pretty relaxed and prioritizing work-life balance, rarely working overtime, and being willing to spend lots of time on things like training, seminars, meetings and so forth without worrying too much about it. We work the way we work and things get done when they get done. I think that's a big part of what people complain about with civil service, and although it's nice it does grate sometimes. When I was a contractor it was my manager's job to keep all that stuff away from the engineers and block anything that might distract us from getting real work done. On the whole I prefer the government's way of doing things.
There's massively less accountability on the pay and hiring side of things, that's for sure.
The whole thing seems like a perfect illustration of Chesterton's Fence.
Don't have a Niskanen reference, but Brookings has a similar chart: https://www.brookings.edu/articles/the-true-size-of-government-is-nearing-a-record-high/
Considering that the contractors do the same work as the bureaucrats, I think this is a difference without meaning. We just aren't on the hook for unionized employees and pensions. Its a tool to side step those requirements, they aren't fundamentally doing different work.
"contractors do the same work as the bureaucrats"
No, they don't. It's a fundamental misunderstanding that you can parachute in some MBAs from Deloitte for 6 months and expect them to have any understanding of the actual mission, constraints, intergovernmental issues, etc. that an agency contends with in time to accomplish anything useful.
I find folks who hold this opinion either haven't worked in a bureaucracy, haven't seen what happens when contractors try to 'do the job' of the real employees, or both.
Yeah, that's the "why can't we treat government like a business?" mindset which runs into problems due to the fact that governments are not for-profit. I hate this new idea of calling the people using government services as "clients" or "customers" because they're not; if they don't like/can't get the service they want, it's not like "Aldi didn't have the thing in stock, I'm going to Tesco instead" for them.
And it's also not similar in "okay this person is renting a local authority house, you can treat them like someone renting from a property management service" because you can't. Sometimes I wish you could, because (again) there are people gaming the system and not paying even the minimum rent, and practically daring the council to evict them because they know they can go to court and the judge will order the council to find housing for them.
But in the main, you can't charge "market rate rent" etc. to those people, and you are (technically at least) obliged to provide a range of services in maintenance and upkeep and accommodating needs for them, so it's not at all like "put an MBA in charge of the housing department and they'll get all the backlog of rent paid up on time".
evict them again, and again, and ...
... this whole absolutely crazy total submission to the high wizards of the gavel is madness
as long as the judge doesn't want to personally hold the bureaucrat in contempt of court they should do the thing they think needs to be done.
just as the courts play their part in this inconsistency theater.
eventually the legislative steps in.
instead what we have is historically low trust and approval rating.
In my experience the people from Deloitte or wherever are 9 times in 10 wildly better than the bureaucrats and more helpful even if they have only been there a week.
I don't know I get hired as a contractor all the time to basically do shit that employees should have done. It is like my main line of business. And how it generally goes is something that was someone's full time job to do, or say 40% of their full-time job, and that they have had a year or two to complete.
I get done in a couple months of working 10 hours a week.
One $20 million program I worked on the main lady working on it was making like 1-23 phone calls a month, and sending maybe 5 emails. Full time staff head of this project, accomplishing nothing. "What is happening with subgrantee X?" I don't know I haven't talked to them in a 6 months. There are only 30 subgrantees and your WHOLE job is managing them, you should be talking to them EVERY SINGLE WEEK.
Some of the places I get hired the people are overworked, and I legit feel bad of them. But many they are just incompetent/ignorant or lazy. Paying me $20k for two months to complete something that is 5 years overdue (yes I do a lot of things like that) when the full-time staff haven't gotten it done in that time seems like a steal.
That's the thing - if you don't/can't have the personnel in your organisation all in the name of Lean Efficiency, then you have to hire outside consultants for expert advice on tricky problems and large-scale projects, and of course this costs a minimum of twice as much, because now you're paying them market-rate fees.
Do federal agencies need to be capable of being sued by the general public? If they need more accountability couldn’t that be provided by greater legislator scrutiny? Someone with some basic incentive to make sure the agency is still able to function.
"Grow the government while taking away power from individuals to sue the government" does seem like a way to increase government capacity. It also seems out of step politically for both parties at this moment in time.
I would rephrase that as: “shrink the government by removing the need for it to preemptively fortify itself from privately waged lawfare”
That's essentially what various forms of government immunity is all about. Prosecutors can't be sued for bringing charges. Judges can't be sued for ruling on cases.
If/when people are able to sue, it was often intentional on the part of Congress allowing it, and always should at least have been contemplated because Congress knows how to add immunity to lawsuits in bills.
Those immunity rules were mostly created by the Supreme Court, who apparently thinks that nobody could possibly enforce and administer the law if they were actually SUBJECT to it. It's like eating at a restaurant where the chef says he couldn't do his job if he had to eat the food they serve there.
In practice the rules get applied in pretty horrifying ways sometimes, as in https://en.m.wikipedia.org/wiki/Connick_v._Thompson where a man spent 18 years wrongfully imprisoned when there was exonerating evidence withheld and the Court insists it's important that nobody be liable for this.
Would you demand that a solder must go through a murder trial everytime he kills an enemy combatant?
Soldiers can and do get tried for murder, even during war. Murder is a crime under the Uniform Code of Military Justice. Your example isn't even a very good strawman.
No, but I also don't think that really has anything to do with what I said.
Connick wasn't a soldier, he wasn't making decisions while somebody shot at him, and his actions and the actions of his staff falsely imprisoned somebody for almost two decades. And he wasn't even being held criminally liable.
I think it's very clear that the state shouldn't be able to falsely imprison its own citizens and then refuse to so much as compensate them for it.
In your post, you suggested, "couldn’t that be provided by greater legislator scrutiny?" In practice, the legislature would direct the Executive to stand up an oversight organization to provide this scrutiny and provide reports to Congress. This is actually a common thing that they do! The result is that you grow the government by creating a new oversight organization. This growth occurs even if the oversight organization is good at its function.
But critically, inspectors general and the like probably don't slow things down the way the threat of a lawsuit does. They can be more judicious about holding government accountable for things that are actually bad while not picking a fight over stuff that's basically fine.
Agree with all of this. My point is just that creating offices like this is ADDING bureaucrats, which is the opposite of what everyone wants to do! :-D
It’s fine creating one extra bureaucratic job if it removes two or three existing ones.
> seems out of step politically
Wait what? Trump would love to do that.
I think such agencies make a lot of decisions which affect some people disproportionally. Say the FDA bans a certain food chemical (phthalates, lactose, whatever). This would certainly impact some food manufacturers, and it is good that the have the ability to sue.
However, if they only can get sued by persons disproportionally affected by the regulations, this will distort their incentives. If you have to prove that you have personally suffered negative health consequences of e.g. some mild carcinogen allowed by the FDA (lest they get sued by Big Carcinogen), and realistically you will not be able to prove that, because cancer is stochastic -- if the substance increased your cancer risk by a factor of two, that means that it is entirely plausible that you were not harmed by the substance despite getting cancer, and thus you would not have standing to sue.
I think that the court is the correct battleground if some people want to ban something, and some people want to allow it. However, courts take their time, and there are a zillon chemicals someone could either put intentionally into food or detect in food at any day, and having blanket rules which allow or forbid all substances not explicitly regulated would both have bad consequences. Thus having a regulatory agency which can make decisions (a bit) faster than the court system seems reasonable.
Private lobbying basically already fulfils the same function in representing special interests. It’s just the money goes to lawmakers instead of lawyers. Arguably only a dubious improvement.
Lawyers employ a lot more people than lawmakers.
Not sure if that's good ("job creation") or bad ("soaking up people who could be doing something more useful").
So whenever I am dissatisfied with a decision of a regulatory agency, I should try to lobby the US congress to add a provision like "Also, the FDA must allow/ban phthalates" to the FDA mandate?
I am not sure what it costs to get special interest legislation the public does not care about passed in congress, but I assume that it is much more expensive than suing in federal court.
Also, I have slightly higher confidence in the court system to actually make a reasonable decision than in the outcome of lobbyists interacting with congress.
Does the US not have select committees? In the UK legislators are able to scrutinise agencies directly. Not simply through direct legislation.
In a functioning Constitutional system with separation of powers, yes. To me, the core issue here is that the regulatory agencies have been given huge swaths of power to make their own rules, to be judge, jury, and executioner. In a dream world, all rulemaking would have to get passed as Congressional legislation and executive agencies would only have enforcement powers. All ALJs world be moved to Article 3 courts. And all regulations would sunset every X years.
It's fine to "distort incentives" that way. The FDA's regulations are not supposed to be driven by lawsuits (thus it's not the "correct battleground"). The Constitution gives Americans rights that the government can't violate without being sued, it doesn't grant us the right to sue the government to compel it to act.
All rights come with corresponding obligations to preserve those rights. The right to property means little if the police refuse to investigate crimes against you. A law protecting you from tainted drugs means little if the FDA is corrupt and not doing the inspections that it should. So the proper solution in this case is to get a judge to tell the offending branch of the government to do its job properly.
Did you know that the police don't have a legally enforceable duty to protect you or to investigate anything? You can't actually sue the police to have them do their job--and many times you're not allowed to successfully sue them when they did something wrong! So no, this just isn't at all how things work. Private causes of action are the exception, not the rule, when it comes to government.
People have tried suing the police for failing to enforce the law, SCOTUS ruled they have no legal duty to protect us:
https://en.wikipedia.org/wiki/Town_of_Castle_Rock_v._Gonzales
Sure, it does. Our rights include equal protection - which means that the government has to follow the laws that it has passed.
No, "equal protection" does not mean that. It can give you the same access to the courts as anyone else, but those courts can still say the government isn't legally compelled to act (as in Castle Rock v. Gonzales).
Following the law and enforcing the law are entirely different things.
You are right, it's not equal rights, it's a statutory right.
My understanding is that "they only can get sued by persons disproportionally affected by the regulations, this will distort their incentives* is pretty close to the default rules of standing in federal courts. In order to bring a case before a federal court, you need to have a personal stake in the "case or controversy", in some direct and concrete way.
For more or less the reasons you describe, Congress has often created statutory causes of action allowing interest groups to sue regulatory agencies. And where they haven't, there are workarounds like shopping around for state governments to sue on behalf of the collected interests of their citizens, or finding an individual who is personally inconvenienced by something you already want to sue over and getting them to agree to letting you sue nominally on their behalf.
"I think that the court is the correct battleground if some people want to ban something, and some people want to allow it."
The problem isn't just that courts are slow, it's that they are not experts in the thing they're being asked to ban. The point of regulatory agencies is to hire experts who can make good decisions and, if necessary, defend them in court. If everything started out in court, the courts would have to hire all those experts instead. As it is, when regulations get challenged in court, the decision often reflects a lack of technical understanding because the judge just doesn't know that much about it.
> I think that the court is the correct battleground if some people want to ban something, and some people want to allow it. However, courts take their time, and there are a zillon chemicals someone could either put intentionally into food or detect in food at any day, and having blanket rules which allow or forbid all substances not explicitly regulated would both have bad consequences. Thus having a regulatory agency which can make decisions (a bit) faster than the court system seems reasonable.
Except the FDA does NOT regulate additives very much. Since 1997 (really since the 60's), they allow any company to "self determine" that any given additive is safe, and put it directly in the human food supply.
The FDA on food is the exact opposite of "kill thousands by having ridiculously high 10 year, billion dollar hurdles," in the sense they let food manufacturers put whatever they want into the food supply at pretty much any time.
Since 2000, there’s been only 10 "full process" approvals. There have been 766 new food chemicals since then, which means 756, or 98.7%, have been “self-determined” and put in our food supply with essentially zero vetting or oversight, or means of measuring impacts, or means of recourse if they’ve messed up, either via incompetence or malice.
This was reaffirmed in 2016. The US has ~10k food additives, vs Europe and the UK's ~2k.
From the FDA.gov site:
“The FDA has established a GRAS Notification Program to help ensure that these ingredients are safe for the ways in which they will be used and to help industry meet its responsibility for ensuring the GRAS status of ingredients they intend to use in food. **This notification is not mandatory; however, the FDA strongly encourages manufacturers to contact the agency** and follow the available procedures for our oversight of GRAS conclusions by submitting a GRAS notice.”
(bolding mine)
“By 1997 FDA had tentatively concluded that it could no longer devote substantial resources to the GRAS affirmation petition process. The agency published a proposed rule outlining a GRAS notification process to replace the resource-intensive GRAS affirmation petition process.”
“The GRAS notification program provides a voluntary mechanism whereby a person may inform FDA of a determination that the use of a substance is GRAS, rather than petition FDA to affirm that the use of a substance is GRAS.”
This seems to be one of those things that changed in the 60s/70s. In the New Deal era the federal government could run rampant and there wasn't much anyone could do about it. Later you get Jane Jacobs opposing Robert Moses on freeway construction, Ralph Nader the crusading consumer-rights lawyer, the environmental movement creating myriad opportunities to sue, and of course the civil rights movement setting the norm that the courts are more reliable than the (constrained) ballot box. https://www.slowboring.com/p/community-meetings-arent-democracy
I think the short answer is 'yes' - it is important to allow some check on administrative power. Congress, for various and sundry reasons, seems incapable of performing that function, which leaves the judiciary.
That being said, Congress can (and should) provide numerous offramps that allow the process to be expedited. That could mean providing specificity over what types of things can be sued over, how many appeals they get, carving out specialized courts that only hear specific classes of cases, etc, etc.
If you're committed to the "agencies make regulations" model, it's really hard to hold them to account. If, for example, the SEC did something blatantly unlawful (eg. require all dealing in crypto to happen through OCC-chartered banks), but there was both a president who was hostile to crypto and no strong majority in favour, the only recourse would be a lawsuit.
You could argue that there should be a requirement that the person in question be "affected" by the regulation, but that skews all business regulations in favour of businesses, all environmental regulation in favour of producers etc. (as they're the only ones who can sue if the general public can't). This isn't necessarily going to trend Libertarian either; if only airlines could sue the FAA, they'd slowly reinvent the CAB by the back door.
Judicial review is a fundamental part of the US system of government. Technically Congress can say "this law or federal agency is exempt from judicial review", but there is zero appetite in Washington for that.
It's not a feature of all governments. Many parliaments are sovereign in which judges purely interpret law. They can't repeal or alter laws.
Sure, but as above comments note, the US in say the 60s did not have *this* level of judicial review of government decisions. So it's clearly not a permanent unalterable feature of the American system of government
I would say yes, under basic rule of law principles. The judiciary system is its own thing and no one should be above it (though recent events seem to suggest some think the POTUS should, which is already plenty controversial).
I really enjoyed this article on how bureaucracy interacts with technology, https://www.niskanencenter.org/culture-eats-policy/, with more details form the trenches on how what Scott identifies actually plays out in practice at various levels.
The only nominative Musk-related determinism I can think of, in English, would be that "musk" is a strong smell, something that would be generated when people really work hard and push themselves to the limit. Easier if you leave the confines of English, as in Finnish his first name means "of life" (so if you mix languages his name would be "The musk of life"), allowing for more advanced level of punnery.
A lot of people seem to view bureaucracy as like some sort of a game-equalizing mechanic in 4X games. Once a civilization is big or advanced enough, it starts generating "bureaucrats" that then create "bureaucracy" that then slows the civilization down. Of course it would be bureaucrats doing that, it's in the name - confectioners create confections, bureaucrats logically then create bureaucracy.
If our position is that bureaucrats would gladly make things as simple for themselves as possible to make their own lives easier and the lives of the people they're regulating easier - and the libertarian view of a typical bureaucrat spending their lives going "BwaHaHaHA! I can't wait for another day trying to make the life of some small businessman harder and drive them to despair!" has never gelled with the actual bureaucrats I know - then a solution for tech companies to make following regulations easier would presumably be developing more software for regulators and their clients to follow the regulations as easily as possible. A TurboBureaucrat instead of TurboTax, in a way. 'course there's probably a fair bit of that already in existence, there certainly is here in Finland.
Software and other technology would probably make a lot of agencies more efficient, but the funding and procurement process for those technologies is equally thorny and larded up with rent seeking.
Naturally, but it's not made easier if the natural process of "cutting bureaucracy" involves trying to wring the amount of available funding to be as low as possible.
The experience of government getting software to improve services has not been unalloyed success:
https://www.bbc.com/news/business-67993493
"Fujitsu Europe's boss has admitted the firm has a "moral obligation" to contribute to compensation for sub-postmasters wrongly prosecuted as a result of its faulty IT software.
Paul Patterson said Fujitsu gave evidence to the Post Office that was used to prosecute innocent managers.
He added that the Post Office knew about "bugs and errors" in its Horizon accountancy software early on.
...Between 1999 and 2015, more than 900 sub-postmasters and postmistresses were prosecuted for theft and false accounting after money appeared to be missing from their branches, but the prosecutions were based on evidence from faulty Horizon software.
Some sub-postmasters wrongfully went to prison, many were financially ruined. Some have since died.
It has been described as the most widespread miscarriage of justice in British history, but to date only 93 convictions have been overturned and thousands of people are still waiting for compensation settlements more than 20 years on."
https://www.bbc.com/news/business-56718036
"What is Fujitsu's Horizon system?
Horizon was developed by the Japanese company Fujitsu, for tasks such as accounting and stocktaking.
It was introduced by the Post Office in 1999.
Sub-postmasters quickly complained about bugs in the system after it falsely reported shortfalls - often for many thousands of pounds - but their concerns were dismissed.
The Horizon system is still used by the Post Office, which describes the latest version as "robust"."
So this system was making known errors, both the software company and the Post Office knew this, and yet they still prosecuted people on fake charges and, moreover, the executives in charge have pretty much gotten off scot-free:
https://www.bbc.com/news/articles/cjq5g95e8vpo
When I see people recommending "why not technology like software?" for government services, I shudder.
The smell meaning comes from the musk deer, which secretes an odorous chemical called musk from its musk gland. The musk gland is so called because it looks like a testicle, and the Sanskrit word for testicle is musk, so called because they look like mice (???) and the PIE word for mouse is musc.
Muscles are also named because they reminded someone of mice!
AFAICT Elon Musk's surname isn't related to any of this, but rather comes from several layers of abbreviating the name of the town of https://en.wikipedia.org/wiki/Montfiquet, where presumably his ancestors were from. Still, I thought it was a neat connection.
Wondering if "Elon Montfiquet" could have attained his position as arguably the world's currently most powerful person.
I’m always up for tracing etymology back to proto indo european. More of this please.
"Burning" and "sanding" are both techniques for clearing away excess dead wood. As such, I have a proposal for who should be in charge of DOGE.
Tech helps, but then we have to be willing to reap the benefits of the increased efficiency rather than go "great, now we can do paperwork ten times as fast! That allows us to mandate ten times more paperwork!".
Having worked in the State Dept foreign service, I think they could automate 50% of the work and have 50% less people working very easily. Everything there is slow and feels like it’s from the 1980s. Nothing is efficient and no one cares because it’s so hard to fire people. There is little incentive to perform at private sector levels.
Instead of streamlining, they do the opposite and hire more people, because people like having more coworkers and people like doing less. They also want to hire their spouses, so they actually make up fake jobs for their spouses. This is for morale, but from a tax payer perspective it’s pretty nuts, because each employee costs a lot.
It’s a very slack job full of long lunches and cocktail parties and maximizing per diems. I eventually quit because 1-I was bored/frustrated by the endless paperwork and bureaucracy. 2-it felt morally wrong to be wasting tax payer money like this.
I'm not shocked about the foreign service... I know some people who work for the foreign service of a different country and it's a similar vibe. No wonder half of the U.S.'s ambassadorships are largely political prizes handed out to major donors and supporters. But I will say the jobs for spouses thing is a legitimate problem in an agency that is asking its employees to move around the world every few years. It's very hard for a spouse to put together anything resembling a decent career if the foreign service doesn't do it for them. You can see universities being forced to do the same thing with spousal hire programs.
I know. The spouse jobs make sense for morale, but they’re part of the bloat and some can actually be in positions that impact speed—like office managers. They also have too many local hires for all the same reasons I’ve mentioned, and though the local hires are usually cheaper, they’re not incentivized to be efficient or innovative.
Plus the whole system of moving everyone every 2-3 years also makes things inefficient because you’re constantly spending all your time just learning a new language/culture/job and unpacking and packing and on vacation in between posts.
Also a lot of flying to places to look at things when sending a video or photo could have sufficed, but people like travel and perdiem so they go in person.
Actually there is so much waste I can’t even get into it, but it’s super ridiculous looking back on it.
Ideally, you'd have foreign service operatives who concentrate on one area of the world where there just a few languages and where their previous experience is actually useful. "Area of the world" could be very broad - like all of Latin America, where you'd only need two languages (Spanish and Portuguese), and some things you learn as a junior diplomat ("Third Secretary", in diplo-speak) in Colombia would be useful years later as an Ambassador to Argentina or whatever. Also, you'd build up set of contacts that could be really useful across the region.
Obviously, there are some countries that are much more distinctive in culture and language than others (however good your Vietnamese might be, it doesn't help a bit in learning Thai), but they still have important relations with each other and understanding something about the other countries that are important to where you are posted is of real value.
Just to pick a random British diplomat whose autobiography I read - Craig Murray (who became semi-famous for something to do with opposing torture of people suspected of terrorism, and I'm not being more precise because I forgot the details and I don't want to be sued) was Ambassador to Uzbekistan after working in first Poland and later Ghana as a junior diplomat, as well as working on UNCLOS (the law of the sea) when back in the UK between postings. Yes, they sent an expert on the law of the sea to one of the only two doubly-landlocked countries on earth. What do you learn in Poland or Ghana that is useful in Uzbekistan? What contacts could you have cultivated that are useful now?
And this does sort of happen sometimes, but there are also cases where it doesn’t. I remember meeting an officer who spoke fluent Mandarin and was frustrated that they wouldn’t post her in China and were instead making her learn French because they’d decided to send her to a small post in Africa instead of to China. Because sometimes they believe that if you speak the language too well, or understand the culture too well, then you’re too open to foreign influence.
There's also a prestige/priority dimension here. As you say there are plenty of regional specialists in the Foreign Service. But I've also known people who got into trouble--not career-ending trouble, but trouble nonetheless--and this changed the possibilities for their next post. There's also the question of hardship posts (some people make their careers early doing this, but don't want to be in Afghanistan forever) and differing levels and means of interaction between the diplomatic staff and hosting country.
On the spousal job question: my grandparents were in the Foreign Service during an era that predates this policy. Wives were subject to extremely rigid expectations that constituted an unacknowledged full-time job. My grandmother chafed against these (I shared some of her recollections here: https://tomlee.wtf/2011/02/21/a-foreign-service-wife/) and even, I later learned, wrote a series of sardonic articles about her experiences in Foreign Service Journal under the pseudonym "Mary Stuart". I don't know much about the system that superseded it, but the old ways were impracticably sexist.
Yes. Correct. Even when I entered the service in the mid-2000s the sexism lingered and I had multiple incidents that would now be metoo material. Also, the wives were still expected to constantly host like First Ladies, and there were fewer women than men in the upper levels of the service. It felt very colonial at times even in the 2000s. Also, I was working mainly in Africa, which probably contributed to things feeling more antiquated.
Thank you for the link - it was a fascinating read.
The one thing that struck me was her reference to the SATs - I wonder if she's got them mixed up with some other test, or if they were really that different back then? She describes them involving a composition; the SAT when I took it had an essay, but it was a novelty nobody seemed to take very seriously (except the students, of course) because it couldn't be objectively graded like everything else, and I believe they removed it shortly thereafter. Possible she meant a subject text, or they were really different back then? I don't know much about testing during or right after the war; my father wouldn't have taken it until well after. It's interesting it goes back that far.
Anyway, really interesting to see her perspective. She must have been a very strong-minded lady.
As a fellow federal bureaucrat (well, okay, I'm a mini-bureaucrat in training, a lowly GS-7 clerk), I wanted to echo on to a point here: most of the posters tend to think that bureaucracies should be resident centers of expertise regarding their field, staffed by autist wonks who are deeply read in their field (which is to say, people largely like the poster here). Fundamentally, though, expertise in the bureaucracy's mission isn't what gets your promoted in the bureaucracy, it's expertise in the bureaucracy's personnel system and also in knife-fights with other bureaucracies over various rice bowls and other resources. Tismy wonks don't do well in those kinds of battles has been my experience, so the people who posters think should be in charge tend to end up shunted aside into unimportant subelements.
Also in major impacts to efficacy of the bureaucracy, the need to "error-proof" decisions and actions (not just from external lawsuits but from internal admin grievances) is unpleasant.
In the government tech contracting world, e.g., contracting for NASA or DARPA, they move program managers to a different location every 3 years. I can't imagine any reason for doing this that would justify the massive waste it causes. In my experience, contracts under a million dollars were usually just throwing money away, because by the time you got through the 2-to-4 year process of writing the grant solicitation, choosing one grant proposal, doing the work, and submitting the final report, the person who'd wanted that work done, and who would get credit for it if it were used for anything, was gone; and the entire project would just be dumped in the trash, sometimes without anyone even reading the final report.
So true. And all the report writing — mostly it doesn’t get read by more than a few people, if that, yet uses sooo many paid hours.
Rotating personnel is a standard check against corruption in multiple fields, from financial accounting to police work.
DARPA program managers are explicitly term-limited, that's part of the way the agency was designed to work. The theory is that it forces you to bring in fresh thinkers every few years, and that creates a culture of PMs who take risks rather than becoming careerists who just approve the same non-risky programs and contractors over and over.
Does that create inefficiency? Obviously. Does it have benefits? From my experience I tend to think it has huge benefits: DARPA has a great reputation for sponsoring high-risk/high-reward results. But I can't prove it. I just think it's important to recognize how easy it is to take a deliberate organizational feature and casually misconstrue it as arbitrary government efficiency. (To stress, I'm trying to start an argument about whether the policy is good or bad, just to point out how important it is to fully understand things before you start cutting.)
This is the problem with superficial discussions of government reform. At a superficial level you might conclude "this seems inefficient, let's fix it." With just a little bit more understanding you might find out the thing you're fixing is an enormous Chesterton's fence. We live in the most powerful and richest nation in the world and that's largely because our forebears took state capacity seriously and *didn't* organize things with serial number lotteries.
Ambassadors feel like an anachronism anyways - a relic of a time when communication delays meant that you had to give your representative in a foreign court a lot of autonomy and prestige to negotiate on your behalf. The low-level stuff can be done by State Department workers lower down in the bureaucracy, while the high-level stuff is just done directly between heads of state and their immediate cohort of people.
They're also useful because when your people are furious about something Country X did, you can "summon the ambassador" of Country X to make it look like you're actually doing something about it.
" jobs for spouses thing is a legitimate problem" - This is not a problem for the employer to solve. The employer can just ignore it, and employees will self-select in the direction of higher pay (meaning these jobs will probably need to pay a bit more), and those without attachments, and or those looking to spend less time with their spouses.
This is not unlike basically any other job-specific downside, like danger. You have to pay a bit more, and you get folks more willing to engage with whatever that danger is.
"This is not a problem for the employer to solve. The employer can just ignore it"
The trouble is CHANGING the status quo. As of now, the entire field is used to having their spouses taken care of. If this abruptly changes, employees will self-select out in the middle of their careers, and there will be a period of instability with the average quality of an American diplomat falling even lower.
This is a very thin view of the employment market that misunderstands why businesses ever try to offer side benefits or improve working conditions. Employees make employment decisions based on many factors, with money offering diminishing returns, like everything. Employers in turn make decisions based on what gets them the best employees at the lowest price.
Imagine you are trying to hire a highly-qualified career diplomat to a post in the Congo or a star PhD to a professorship in the middle of Kansas. This person is not married to a nobody. Their spouse is almost certainly highly educated and intelligent. The couple will make their decision together and the trailing spouse is probably not interested in being a trophy husband/wife interrupting their burgeoning career for a resume gap.
There is an amount of money that would probably convince the spouse to suck it up and move so that the couple can save money for a few years. But this amount of money would probable be higher than what both of them could have made together back home.
Now imagine you are the agency doing the hiring. Do you offer a $120,000 employee another $100,000 to convince their spouse to agree to the deal? Or do you hire the spouse into a $80,000 job that also needs to be filled, even though they provide only 75% of the value that the optimal candidate would have provided? This is a no-brainer.
I was also an FSO, long ago. The Foreign Service is a small bureaucracy in a historically low-productivity field and many of my former colleagues were (and are) hard-working, though you can question how useful much of their work is. Spouses get hired because (i) it’s cheap (ii) most FSOs are quite bright and they and their spouses could get decent jobs in the States if they stayed there. So unless you have a spouse who just wants to stay home and care for kids (but nannies are really cheap in most of the world) there’s going to be a demand for jobs for spouses.
Yes. Definitely correct on the brightness and hard working nature of the officers, but I can’t help but think the whole system needs an overhaul. Most employees were frustrated by the slowness and inefficiencies also. But then a lot also milked the system. Or this is what I saw… could have just been the specific posts I was at, but I sensed it’s widespread.
This is a great case study actually. State budgets are incredibly popular to cut, and it's also popular for each new Secretary to create a new dedicated special office to absorb the few personnel remaining, so personnel dedicated to traditional mission have been basically vanishing over time. Some career tracks pick any post they want because of chronic understaffing.
And yet also... the core remaining, without the resources to do effective diplomacy, are a certain breed. Agree maybe a third are self selected for the benefits and just riding out the bipartisan disdain for their work, another third are ground down by the impossibility of driving effective change and become cynical with the same result. And the last third are wonky savants who understand an unsettling amount about state capacity and driving policy changes for the benefit of the US and humanity. And this last group is responsible for a phenomenal number of US wins that nobody will ever report on or notice.
So it's both true that there's a lot of personnel wasting time and also that if you cut personnel, which we have already done, it directly reduces the department's ability to do really clever and effective diplomacy.
At what multiple do we fund the military relative to diplomacy? And then when we look to our bag of hammers to solve policy problems we see an awful lot of nails. Then the public wonders why we get entangled abroad so much.
Better firing authority would be great, but should also be joined by better hiring authority, and greater accountability for results. Those seem better than mass firing to me.
100%
Is it a very slack job until the day there is a major war or other international crisis?
This seems like the type of agency where the amount of work is not constant, and you desperately want to have enough excess capacity to handle the surges quickly because they correspond with the most important things you ever do.
Yes and no, because when there is a major crisis almost everyone in the DOS gets evacuated in advance anyway and they send in the military.
My sister is actually married to a language officer in the Danish military, and got hired as a spouse so they could live together. Seems kind of necessary thing to have if you ever want to keep your workers past their wedding.
Also, speaking from the same example: It might be that most of the time the foreign service / military / diplomats etc don't have enough to do because the world is mostly peaceful and not involved in too many international crises. But you still hire that many officers fluent in Russian, because the day one of your submarines accidentally bumb into a Russian sub under the polar ice you really are going to need all the diplomats you can get your hands on.
But that's the level that is *not* going to be fired, is the problem. Vivek may get it pushed through to fire the receptionist in the local office or the Grade III doing filing, but the diplomats' spouses won't be touched because that's way too hot a potato.
So the inefficiency and bloat continues, while service on the ground declines.
Yeah I know a lot of civil servant, and lots of the infeffiency really isn't because of anything related to suing.
I suspect Vivek is thinking of a version of this. Each worker can actually process 1000 forms. They process 10 because that achieves the departments performance objectives. If half were fired, theyd groan and process 20. This might not apply at acute crunch times like the translator example, but would day to day
> Nothing is efficient and no one cares because it’s so hard to fire people. There is little incentive to perform at private sector levels.
This is my experience as well.
> I was surprised to see someone with such experience in the pharmaceutical industry say this, because it goes against my understanding of how the FDA works.
The obvious explanation is he is just lying through his nose to score points in some intra-MAGA power struggle. It's like elected Democrats who talk about corporate greed - in private most of them would probably admit that they know punishing people for wanting to have money is not actually a good way to solve market failures, but that's the rhetoric that polls well so that's the rhetoric you'll get in public.
I agree with your logic, though I think lying is the wrong framing. He knows that what he proposes will never happen.
I think it's unlikely to happen as-described. That said, I don't think it's necessarily the case that he feels that it's only in his interests to to propose measures which will not harm the effective operation of the government, or that proposals which do will not be implemented.
Trump says a lot of stuff that seems to be maximally offensive/unhelpful/impossible on the way towards achieving a much smaller goal closer to what he probably intended in the first place. Vivek was directly following Trump's style and approach when campaigning.
Best guess, Vivek wants to cut the federal government and bureaucrats specifically by significant amounts. But not 50% and not arbitrarily. He likely wanted to scare the agencies into doing a better job themselves, rather than facing an obviously unfair and highly disruptive "you lose your job because your SSN ends in an odd number" approach.
That is lying though. Saying you'll do something you know you won't or can't is lying.
Having listened to him for X hours on podcasts (the person he pretended to be on the debates was fairly cringeworthy; he's completely capable of playing a part for personal gain), and how many idiosyncratic complicated thoughts he has that may or may not work but are integrated into a coherent world view that doesn't map to political or personal incentives for him to have, I have a lot of good faith skepticism about this.
I predict that if Scott and Vivek got on a podcast for three hours and talked, Vivek would either change his mind or pass a turing test of explaining why he disagrees in a way that sounds reasonable to a neutral audience.
'Craft an explanation for why your policies are internally consistent and not self-serving that is convincing to the average person' is not a very high bar for a smart and experienced person with inconsistent self-serving policies, though.
An easy way to convince me of both good-faith and competence would be to say good-faith, competent things in public.
Where is the boundary between a post that can be counted as hyperbolic but making a worthwhile point, and one that is so ludicrous and daft that it should get the poster laughed out of the room? I would argue that Ramasvamy's suggestion is on the latter side of the divide.
Not only does removing every other bureaucrat at random halve the capacity just on a "people who can get stuff done" basis, it will probably reduce the capacity close to 0 for a while, because at least some of the people you throw out will be specialists with arcane knowledge necessary to keep the system running, and many will be the liaisons between different branches who now suddenly have no idea who to talk to when some issue related to another division comes up.
If you think some of the work done by bureaucrats is unnecessary, by all means, find out which part of the work that is, and optimize it away. Until you're willing to do that, making blanket suggestions like that doesn't really help.
Yeah in some ways this amounts to saying "I am about as useful and smart as a random number generator"...
Or perhaps that the current system is so fucked up that large changes even in random directions would be a significant improvement over carefully-directed incremental changes.
It's wild that we live in the richest, most powerful country in the history of the world, but our current system is also that fucked up.
Not really: if the US weren't so unassailably powerful, it would have been forced to make drastic changes long before its systems got so fucked up.
Yeah, that's why America, the rich country, has the fucked up systems, while Russia, the poor country, has the really strongly functional ones. Come on man. America's system of government is the most blatant Chesterton's Fence in the history of the world.
I've never been to Russia, but I bet there's enough corruption to make the system work fine. They'll have regulations about what kind of doors pharmacies can have, but everyone understands that when the inspector comes by to inspect your doors, you need to pay him off, not try to demonstrate to him your doors meet regs.
The US's problem is the abysmal lack of such salutary corruption, "the people's deregulation."
Tacitus said it well, "The more numerous the laws, the more corrupt the state needs to be."
perhaps everyone in congress really IS a reptilian
at a certain point you have to dismiss the ludicrously implausible
Or maybe certain people have succesfully convinced everyone that the system is totally fucked up, and only they know how to fix it.
It's been shown again and again that random number generators are *better* than experts in some domains. There are multiple examples in Kahnemann's book 'Thinking fast and slow', but I think the classic example is that managed funds tend to underperform index funds.
But he didn't get laughed out of the room. He's one of the leader's top lieutenants now.
Besides, I don't see how it's that stupid. If the bureaucracy is too understaffed to actually approve anything, they're also too understaffed to actually enforce their rules. It's a temporary measure until the department can be eliminated and replaced entirely.
Right, he didn't get laughed out of the room. Neither did that other dude who promised to have people on the way to Mars by 2022, and neither did their overlord who promised to end the Ukraine war on day 1. Somehow that points to a much deeper, much more serious problem than any failures of US bureaucracy.
The laughing-people-out-of-the-room department is desperately overworked and understaffed.
There are a lot of people who think they'll personally benefit from living in a US without state capacity, even if the overall economy shrinks by 60%. Based on history I think most of them are wrong, but you can't argue with personal greed. There are other more capable people who see eliminating the civil service as a way to eliminate one check on untrammeled executive power; they fully intend to staff a bureaucracy, just one filled with politically-appointed loyalists.
The key is to probably establish a secondary bureaucracy that evaluates the first for efficiency. Every form that the bureaucracy requires should need to be justified by a secondary form explaining why it is necessary work, which will be evaluated by this new department.
This already exists, and is called OMB. Go read about the Paperwork Reduction Act and see if it has made government better or worse and report back on your findings.
Perhaps a third layer, so every action done by the OMB has to be justified by a form that explains the benefit. These forms could be evaluated by the new overseeing department.
But then who would review the forms reviewing the forms reviewing the forms to make sure they're reducing the paperwork burden on the public?
It wasn’t meant literally or seriously. Its purpose was to show how good a troll he could be as part of his audition for VP.
He is a high honors graduate of the Andy Kaufman School of Performance Art.
Think of Andy during his intergender wrestling champion phase preparing to wrestle a woman in some southern backwater venue holding up a bar of soap and roll of toilet paper to taunt the audience by implying that they need remedial instruction in hygiene.
Edit:
As a side note Andy worked out the mechanics of that bit with another performance artist, Laurie Anderson, acting as a shill playing a random woman in the audience.
What if ~100% of what they do is unnecessary?
What was the size of the regulatory code in 1912?
What was the life expectancy in the US in 1912?
I mean, sure, if you want people to die of workplace accidents, building fires, collapsing biildings, traffic accidents, airplane crashes, food poisoning, counterfeit medicine, air and water pollution etc, go ahead, remove all regulations. But don't come crying to us after your children's funeral.
Most of the improvement in life expectancy and safety has more to do with technological and medical improvements rather than regulation. Of course, the regulators are happy to take credit.
Could it be possible that the regulations that are actually necessary to solve these problems consist of a small fraction of the regulatory code? and that the rest do nothing at best, and make things worse at worst? Were these things *really* that much worse 100 years ago?
How many regulations actually make these things better? Seems that regulations on nuclear energy have made air pollution worse, forest clearing protections have caused more houses to burn down, and medical regulation has resulted in preventable deaths due to medicines being delayed (see the covid vax - finished in January 2020!).
Airline deregulation resulted in many more crashes except, oh wait, it didn't. The price dropped like a stone and it got safer. Hmm. Maybe they're just full of shit?
Another example: if you could buy a new build Toyota Corolla with 1980 regulatory standards, it's likely to not be *that* much safer, but would probably cost ~$10K. Maybe the fuel efficiency would be a little lower, but I haven't even mentioned oil drilling and pipeline regulations. Would anyone buy anything else? That such a product is not allowed to exist is money stolen *from you*!
Yes, technology has improved. So has the capacity of the chemical and food industries to poison us with convenient but unfortunately toxic substances, the capacity of the construction industry to build skyscrapers that kill 10000 rather than 100 people when they collapse in an earthquake, the capacity of the financial industry to lure in more suckers and defraud them of their life earnings, etc. Products have become more complex, and so have their supply chains, exponentially increasing the potential for fraud and deception.
"Of course, the regulators are happy to take credit." Probably with some justification. People in other countries have access to the same technology what the US does, at least in principle. If you follow the news, you hear a lot more reports of people dying from accidents related to shoddy maintenance, shoddy construction practices and horrendous workplace safety in poorer, less-regulated countries (including countries that have regulations on the books, but are too corrupt or incompetent to enforce them) than in the US and Europe. Sure, those regulations cost time and money and effort. If you don't want to afford that, feel free to move to, I don't know, Bangladesh?
"Could it be possible that the regulations that are actually necessary to solve these problems consist of a small fraction of the regulatory code?" That is a possibility. It is also a genuinely hard question to answer, since it would require going through the code, figuring out the history and justification for each item and judging whether it's worthwhile. Again, I would support that effort, but I am not on board with assuming the percentage of pointless regulations is "~100%" from the start.
I am not a bureaucrat, but I work as a software developer, and I can tell you that we have a similar phenomenon of dealing with bloated, borderline incomprehensible code. Sometimes fits of "no one needs that stuff, let's just throw it out and start from scratch" pay off; often, they are followed by lots of bug reports from angry customers and moments of "oh, so that's what those lines of code were for."
I think there might be other factors in Bangladesh...mainly the poverty.
And I'm not convinced that regulations are what got us out of poverty. From 1750-1913 we went from the first steam engines through to Belle Epoque Europe and Gilded Age America. Basically, if it's a popular tourist destination, it was probably built before our current massive regulatory code. And during this span, were there industrial accidents, bridge collapses, and the like? Absolutely! During this span, things seemed to be improving in the safety department over time, with the exceptions being in areas that were extremely poor.
So, if the regulations block economic growth, it's very possible they're killing more than they're saving. While ~100% of the code being useless is likely hyperbole, do keep in mind it plays to the regulators' interest to keep the code as long and esoteric as possible. The analogy I would use is software developers writing unmaintainable disgusting spaghetti code on purpose to increase their job security.
Also, I have to call you out on: "the capacity of the financial industry to lure in more suckers and defraud them of their life earnings" - if this was a real motivation, they'd ban sports betting and other gambling...they haven't. Indeed, they've massively expanded these while making it basically impossible for normies to invest in early stage startups. You could have made a killing getting into Apple or Microsoft early - but you were locked out of getting into Facebook, Stripe, or SpaceX early. Was that really for your own good?
"And I'm not convinced that regulations are what got us out of poverty." No. Regulations are (part of) what brought security and health, after the worst of poverty had been taken care of. More money doesn't buy safety all by itself. Of course, while you're dirt poor, you're more willing to put up with horrible workplace conditions, pollution etc, but once people can afford nice houses and nice cars, they can also afford the regulations that make sure they can enjoy those houses and cars alive and healthy.
"if this was a real motivation, they'd ban sports betting and other gambling..." - it sure was a real motivation, and it worked. When's the last time you heard of a properly regulated, old-fashioned bank going under due to fraudulent dealings, and all their customers losing their money? When's the last time you heard of a large crypto company going under due to fraudulent dealings, and all their customers losing their money? What's the fundamental difference? Wasn't De-Fi supposed to be just like regular finance, just with the opportunity to make big money unburdened by those pesky regulations?
Now I don't know what led to the liberalization of sports betting, but I think it was a shitty idea. (The difference between banking and gambling, of course, is that gambling is always understood to be optional, but you need some way of participating in the financial system to receive money and pay your bills.)
Regarding the startups, I disagree with you as well. Most startups fail utterly. If you're so intimately familiar with a startup that you're convinced that they will be a massive success with good reason, and you know that you could handle the loss of the money, do the work, talk to them, do the due diligence, become an investor. If you're a somewhat informed investor, buy shares of a fund that invests in that startup. If you're a normie who just happened to hear a convincing sales pitch... then it's no better than gambling. (Also, opening up the field for normies would change the incentive structure for the worse. Again, look at the crypto space: lots of startup-like projects collecting investments from normies, many of them with some well-known name or other to back them up. What percentage of them end up not just as unfortunately failed businesses, but outright scams and rug-pulls?)
A false proposition implies any proposition, so it's true both that if ~100% of the work they do is in unnecessary then firing half of them will be fine and that if ~100% of the work they do is unnecessary then firing half of them will still be a disaster because it will unleash the wrath of Nyarlathotep.
That's an interesting choice of date, because the Triangle Shirtwaist Fire, the most central example of "why safety regulations exist," happened in 1911. Would I like to go back to an era where it was legal to lock your employees inside the building and let them burn to death? Gee, let me think about it!
Ah yes. Because some regulations actually are useful that means that they all must be useful.
It's like how after 9/11 they stopped future hijackings by just locking and reinforcing the cockpit doors, but somehow also got everyone to go along with an absurd system where idiots molest you and micromanage the amount of liquids in your bag and make you take your shoes off for no reason while we pay billions for the privilege.
Then someone says: "maybe we just just go back to the pre-9/11 airport security and just keep the doors locked" and people respond "Wow! You want to have another 9/11 to happen!"
I guess it's "common knowledge" that a lot of regulations are results not of lawsuits, or fear of lawsuits, but of inter-department warfare, where one group of bureaucrats try to grab more power by introducing newer, longer, stricter, more urgent forms. I haven't ever worked in government, or a big regulation-heavy company, so I don't know how true is that. But it's certainly the popular perceptions of bureaucracy. So maybe DOGE can target such regulations and be successful.
But in reality, Musk's goal is probably to eliminate most oversight altogether, and damn the consequences (as long as he gets to Mars before his death, he personally doesn't care about the rest of us; and after his death, Mars can build its own bureaucracy from scratch). He just can't say it out loud. So halving the number of people staffing various regulatory agencies might work: it will bring work to a halt, and so DOGE can point fingers and say "well, now those people are completely useless, they haven't approved (a single drug, a single rocket launch, whatever) in a year, let's get rid of them altogether!".
"But in reality, Musk's goal is probably to eliminate most oversight altogether, and damn the consequences" - bingo. It's one of the most brazen acts of corruption I have seen in a civilized country in recent memory. Dude bought himself a goverment position that can get rid of the regulations that he perceives as holding back his companies and also gets to have a say on government contracts for his companies and their competitors. Conflict of interest, anyone?
Well, as a non-American, I'm kind of interested to see how it turns out. I grew up on books like Heinlein's "The Man Who Sold The Moon", and maybe Musk will become "The Man Who Sold Mars". Living out those 50's libertarian tropes.
I remain confused about what "living on Mars" is going to accomplish. Even if it's possible to build anytime soon, a Mars colony is going to be dependent on critical materials shipped from Earth for all of Musk's lifespan and likely decades beyond. You definitely want to make sure that you have a stable political environment here on Earth. (In some sense you're even more vulnerable than someone living in a compound here on Earth, because every nation with space-faring weapons has a veto on your survival.)
I imagine it'd be enough for him just to get there, to prove that it's possible and maybe re-ignite interest in space exploration, or even a space race. I noticed he does talk about future Martian nation, but only rarely and vaguely, so I don't think he has a grand plan for that. Just go into space, become a hero, be buried on Mars, and the rest is for others to work out.
Besides, I don't think it is possible to achieve a "stable political environment here on Earth" in Musk's lifetime, anyway. Certainly not with his personal best abilities. So his choices are down to "forget about Mars, and concentrate on something more practical and boring" and "full steam ahead, and damn the Earth", and, well, he's the kind of person who chooses the later every time.
From a human perspective, there are various possible economic motivations (all pretty blue sky [red sky?] but something will probably work out), and many more scientific ones, but the biggest in the long-term is just the backup capacity of having more than one life-support system for large numbers of humans.
From Musk's perspective, there's the honour/hubris (take your pick) of the founder, plus someone has to ship all those critical materials and that's likely to be a good source of business for rockets for some time, plus as far as I know he's human and so should share those motivations as well.
Repealing regulations is also certainly on the menu.
Cmonnnn Jones Act....
There is no incentive in the current (well, incoming, but it seems like Trump is already in charge) administration for improving the United States ability to trade with other countries. The longshoremen bargain is best understood as purposefully slowing down our ports, as a very easy to implement tariff.
This doesn't seem as relevant to the Jones act in particular, which mostly serves to restrict intra-United States trade.
Certainly there are other regulations acting as ersatz tariffs, but I don't see how the Jones act is centrally that.
On my tombstone may it be written 'To the right my wife, to the left a spot for Jones'
It would require an Act of Congress, although I think Trump could grant exemptions under circumstances (which would almost certainly be litigated if they were too generous).
From what I've heard, the reason why reform is so hard is not just opposition from the domestic shipyards remaining plus the unions, but from the Navy and its favored politicians in Congress - they really like the idea of having a US-made fleet of civilian cargo vessels that can be turned into a merchant marine in case of a conflict, and also are very afraid that the few remaining shipyards will go bust and wipe out the US's ability to make military ships.
I wonder if he could issue letters of marque and claim that the ships commissioned are performing some national security function.
He can't though, unless the actual law empowers him too. Which I'm not aware it does.
Well, he can break the law as he pleases, then it's up to congress to impeach/remove. Will they?
He can't get persecuted, but the court can still order him to follow the law, and all other government employees don't share in his immunity.
They will indeed have immunity after he grants them a preemptive pardon (Thank you Gerald Ford!)
As for the court - "Roberts has made his decision, now let him enforce it"
Sure, but by that logic Trump could ignore an impeachmen as well
I just checked, and the Jones Act creates a private cause of action. So Jones Act can just be enforced by the courts without the executive branch doing anything.
20% do the work, 20% are so bad they obstruct the 20%, and 60% are just waiting for retirement. Every GS-14/15 I have ever talked to over the past 40years. BTW- of course anyone who interacts with the bureaucrats says they were helpful. What sort of idiot says they are useless if you have to continually interact with them. Also GS 14/15s- do not irritate a bureaucrat in the 60%. You have just given them something to do until retirement- cause you problems.
I have a relative doing work for the government who is a true believer in the cause the department is doing, who thinks the government is good and trustworthy, and a significant portion of their job as they describe it to me is routing around the useless unfireable coworkers so they can accomplish their needed mission. Give her a magic button to fire all these people and she would.
This happens in every organization. Eventually you have a small percent who get most of the real work done, a large group that performs their minimal function to not get in trouble and a small percent who are actively detrimental. The large group is mostly just "putting their time in" "waiting for retirement" or whatever. Many may actually be reducing the productivity of the organization in various ways.
The problem with the federal government is that it's intentionally hard to fire people and the ones who can force that to happen lack ways to legibly determine who is good and who is not. So instead churn and the natural process of removing the unproductive is stalled out, gradually building the numbers of people just filling the ranks. Most of them are at least doing what they've been hired to do, so if their jobs are designed well they can at least keep filling out the forms and passing them along - in line with Scott's article. Better employees would be working on ways to reduce the needs for forms and streamlining the processes, etc.
That Musk was able to fire ~80% of Twitter's workforce without the company breaking speaks well to his ability here. But the federal government is much larger and has way more legal and cultural roadblocks to removing people. Musk also can't just ignore the complaints of people upset, because the government is required to be responsive to the public.
Federal civil servants aren't legally unfire-able. They could terminate the 20% if they wanted to, but when they stick around it's usually because their managers don't really give a shit either - don't want the hassle of retraining people or going through the termination process, etc.
Even if that were true, the plan here is to fire half of the 20% who do the work.
Nothing Scott has said here is wrong wrt the proposal in the tweet, even if we buy your premise.
Optimistically, resultant disruption of routines makes most of the remainder of the "actively detrimental" set easier to spot and remove, since half of the folks covering for them are gone, while those who turn out to be irreplaceable get promptly reinstated as soon as the "don't know what you've got 'till it's gone" effect kicks in.
The government will surely be an employer of choice in the future for high performing employees if they randomly delete half the current high performing employees.
Yeah it's like that quote about marketing "Half the money I spend on bureaucrats is wasted; the trouble is I don't know which half"
I am very down on the federal staff and I wouldn't be that harsh. But I would say something like:
35% good, 35% completely replaceable and in way over their heads (often over-hired/promoted), 40% actively negative to getting things done.
Is there a Laffer curve for the bureaucratic state? Adding marginal bureaucrats from zero may improve a country's "output." Adding marginal bureaucrats to a bureaucrat-heavy country may increase sclerosis and decrease "output." In a plot of output vs. # of bureaucrats, the start of such a curve is positive in slope, and the far end is negative. A continuous curve must therefore have some maximum between those two points. Vivek is bolder than saying we're bureaucrat staffed beyond optimum, he's saying we're sufficiently beyond the max that reducing count by 2x improves output. Scott asserts otherwise. Scott, any thoughts as to what your curve looks like?
I think this is a very interesting question and I'd love to know about international analyses. Some countries are small and have small governments. You'd think that some government work would scale with country population but some wouldn't. Are these small countries doing less stuff than their bigger counterparts? Is the country worse as a result?
Even if that model held, don't overlook teh fact that it may be a ratchet.
IE just because adding 25% more bureaucrats at T1 makes you less efficient at T2, does not mean that firing the same number of bureaucrats at T3 will make you more efficient at T4. If the system found jobs for those extra bureaucrats to do and made important function route through them, removing them might just crash everything.
Which is not to say that you can't carefully unwind those effects to reverse the ratchet to the optimum level, but 'fire half of people at random' isn't going to do it.
A large contingent of the federal work provides services, like health care to veterans, screening passengers at airports, or helping the public navigate things like paying taxes or receive Social Security. Will these functions become more effective with fewer employees?
I think you may not have fully steelmanned the "thesis where red tape expands to consume the resources available to it".
What happens if with fewer bureaucrats they justify a regulation with a document only 500 pages long rather than 1000 pages long? When there's a lawsuit, does the judge just automatically say "not adequately justified!" and throw out the regulation? I don't think so. The judges will reduce the standard they expect.
At least they'll reduce the standard sometimes, when they personally think the regulation is needed. One might object that this gives too much power to judges, to selectively decide the justification was adequate. But don't they already have that power? If they don't like the regulation, they can say that 1000 pages wasn't enough, that they should have produced a 2000 page justification...
You seem to have misunderstood "number of pages." It is, in the post, a proxy for "a document adequate to convey appropriate information." Just halving the information, does not halve the threshold for "adequate." Otherwise, repeating the process, every document could be one letter long.
Maybe if the regulation is so important, they can justify it with fewer words? If the usefulness would be obvious and wouldn't require them to torture logic?
This is how laws were written in the 19th century. But presidents started using dubious logic to get around the law and/or extend their authority, so Congress started writing longer bills to hem the president in.
When someone complains about 500 page bills they're basically saying "I prefer to let the president decide everything." That's great when your party is in control, not so good when the other side gets a chance.
And the SCOTUS Chevron decision means Congress will have to write even longer bills than before.
laws != regulations
So it's true that nothing is infinitely inelastic; if an agency has half as many people they will probably eventually cope by reduce the length of their reports somewhat, and if judges keep getting shorter reports they will probably cope by reducing their standards somewhat.
But the failure mode of this insight is thinking that only the things you *like* are elastic, and other things aren't. The number of drugs approved each year is elastic, that can go down. The number of mistakes made is elastic, that can go up. The amount the government pays out on lawsuits it fails to properly defend against is elastic, that can go up. Etc.
So yes, it's probably true that when you pull the legs out from under the entire system and cause everything related to it to malfunction, some of that malfunction will happen in the things you dislike and reduce them some amount. But a lot of the malfunction will also happen in the things you like.
I mean it just depends. Bureaucrats have pretty wide latitude to just ignore parts of the law if it is too burdensome to enact, and that happens a lot until there is eventually some lawsuit. And sometimes the lawsuits never happen and no one ever notices.
Just one example there is a requirement that all outside hiring for large swathes of grantees (I am going to be vague here) always need to first market/consider "low-income" people.
Yes this includes if you need an architect or a lawyer or a nuclear engineer. What does a low income nuclear engineer look like? Certainly not someone you would want to hire generally I would guess. And this is true even for say people building a structure. Why would you seek out low income ones? Also it sort of binds against the prevailing wage rules besides.
Anyway the rule is garbage and stupid, and so what happens is grantees promise to do X, some even stupidly try to do X, but the bureaucrats never ever check X and the grantees figure that out eventually , and soon everyone is just filling out forms and saying they do X, and the bureaucrats are saying they checked X, but no one is actually doing it because it is asinine.
That happens a lot. Maybe 10% of the time for any given rule +/-? Some small portion of the expertise I sell to government entities as a consultant is having the experience on what rules the federal government is serious about, and what rules the government just pretends to be serious about and/or has actively forgotten about.
Thanks for a great post. One thing it is missing is some glancing at what has already been tried at the federal level. Idaho's approach of changing the status quo from "regulate more" to "make sure the regulations really meet a bar" is kind of the approach Trump45 took to, at first glance, noticeable effect: https://www.cato.org/regulation/winter-2023-2024/was-trump-deregulator
It's also about ignorant politicians who have no experience with managing processes, so they write the laws without understanding the costs of them.
Look at the resume of your average politician. Barack Obama, Joe Biden, Donald Trump. Before they came to office, what large, high-volume organisation did they manage, or were even part of? And I can say the same about the UK. They come from law, journalism, trade unionism. None of them have worked in retail banking, Amazon, WalMart.
And if you don't work in these fields, you don't understand things like KISS. That in the era of computers, it's complexity, not scale that is the big cost. That if you have 1000 pages of rules, someone has to write the code for that, staff have to be trained in it, lawyers, accountants and so forth have to learn it all. A legal case will happen because there is a different tax on biscuits and cakes in the UK, and how is that defined? A better idea: tax both, or tax neither. Or even better, remove sales taxes entirely and just raise a tax elsewhere to make up the shortfall. Then you can fire the whole sales tax department.
>Barack Obama, Joe Biden, Donald Trump. Before they came to office, what large, high-volume organisation did they manage, or were even part of?
Trump: building and operating various hotels and casinos. I get the impression his involvement in the actual operation of the businesses (as opposed to promotion and financing) was relatively nominal, thought.
Biden: Had been a Senator for decades, including being chair or ranking minority member of various committees. It's quite possible for a back-bench legislator to remain ignorant of how the federal bureaucracy operates in response to legislative rulemaking, but it would require a particularly dense and slothful Senator to remain ignorant over the course of something like Biden's career.
Obama: back bench legislator, and before than an attorney at a civil rights law firm. You have a good point in his case.
Bush the Younger: Governor of Texas, and before that CEO of an oil exploration company and owner of a major league baseball team.
Clinton: Governor of Arkansas, and before that Attorney General of the same state.
Bush the Elder: Director of the CIA
Yes, and none of those are operating a large organisation with millions of customers. They were politicians, except Trump. And his hotels and casinos were not exactly a huge success, were they? He made his money in real estate which is mostly about large deals with small numbers of customers, not large scale operations.
I don't think you can simultaneously argue that high government office doesn't count as experience "operating a large organization with millions of customers" and that such experience is an important qualification for competently executing such office.
But supposing for the sake of argument that only private sector experience counts, Bush the Younger should get partial credit for the Texas Rangers, which aren't a huge organization (I think MLB teams have on the order of a few hundred employees between the players, coaching staff, ground crew, and front-office staff) but do have millions of customers.
On the subject of Trump's business record, I would be one of the last people to defend it, except to note that it exists.
Being a politician isn't operational experience. It's oversight. And I don't draw a line between private and public sector. Someone who worked in the PR department of a private business also lacks operational experience. I would take someone who managed a local authority over those people.
Ah, that makes more sense. Thank you for clarifying.
Or is your point that public sector institutional cultures are so severely and pervasively broken that experience as an agency director, state governor, or senior senator is usually experience at doing things the wrong way with entirely the wrong mindset?
If so, I'm more sympathetic to that line of argument even if I don't think I'm willing to endorse it entirely.
No. I've worked in parts of the public sector for some very good people, people who have operational experience. They would make better politicians because they understand what works and what doesn't and how to fix things better than politicians do.
They mostly know how to fix things using tools that aren't available to politicians. Their track record when they become politicians, is mediocre at best.
The State of Arkansas isn't a large organization with millions of "customers"? Not even Texas?
Not buying it.
Also worth noticing that Biden and Bush the Elder both spent eight years as literally the President's understudy, which should be about the best possible apprenticeship for that job.
There's a difference between being a politician (oversight) and management (operational).
The Governor of a State is management in pretty much the way that the CEO of a large corporation is management.
100% a huge problem with the bureaucracy is the laws themselves which are passed, which sometimes literally don't make sense or are internally inconsistent.
Some of the COVID relief was like this. Where it was like "here is a big pile of money (sometimes billions) and a million rules for doling it out, and it needs to be out like in a couple months". Like brand new programs that are multiple billions and they are meant to be up and fully functional in say 60 days. So what happens is just LOTS of rules get ignored and fraud protections skipped. Because something has got to give.
If I had to guess, I'd say Vivek believes that bureaucrats can actually do twenty forms per year, but are only doing ten because there's no accountability and they're basically impossible to fire.
I guess there's nothing stopping bureaucrats from refusing to do more than ten forms per year even if their workforce is cut in half. You'd have to reform in other places.
This. The problem for Vivek is that they will simply refuse even if they could work more, and since Vivek volunteered to be the scapegoat through his mass layoffs he will have to take the blame. It's about incentives.
You'll also get a lot of Mount Rushmore syndrome.
"The Washington Monument syndrome,[1] also known as the Mount Rushmore syndrome[2] or the firemen first principle,[3][4] is a term used to describe the phenomenon of government agencies in the United States cutting the most visible or appreciated service provided by the government when faced with budget cuts"
Ah, I see this gambit works the same the world over. In the UK, if it had a name, it would probably be called the Trooping of the Colour syndrome:
When the Defence Ministry is told to make yet more cuts, the military often responds that the Trooping of the Colour ceremony must be cancelled this year, as there simply won't be the manpower for it! Cue press outrage, and the cuts are reduced.
(Trooping the Colour is an annual ceremony performed in London, in which regimental colours are paraded before the troops of a household regiment, in the sovereign's presence. It's original purpose was to remind the troops of the colour flag's appearance, so they would recognise it in battle. But these days the ceremony is no more than a popular tourist attraction.)
Another trick used by Home Office staff in passive aggressive efforts to resist orders to crack down on illegal immigration is to scour the length and breadth of the country until they find a blond, blue eyed, Hitler youth looking young man to publicly eject from the country because his parents' visas were slightly out of date when they arrived from Australia or South Africa twenty years previously, or were missing a vital date stamp or comma!
Can't you just fire them for that?
Not without an extraordinary amount of effort, which will not be sustained across the Federal civil service. The government uses pretty much the exact opposite of "at will" employment. I've seen exactly one civil servant fired for being so lazy that he basically didn't do any useful work, and it took a solid year of his manager's focused effort documenting the laziness and arguing the case with the bureaucracy.
Exactly. At best people get reassigned or sidelined into a do nothing position.
You could fire the least productive ones : but it would take.a bureaucracy to identify ones.
Apparently, efforts to analyze and optimize U.S. Federal bureaucracy is one of the largest sinks of time any money in the U.S. Govt. Mandates keep coming down requiring more and more documentation and analysis of cost and time. In all honesty, the same is true in my multi-billion dollar company (we are sort-of a startup). We have to get multiple quotes even to spend $1000. We probably end up wasting $5k in salary to protect against an excessive $1k quote.
I wouldn't treat # of bureaucrats and red tape as independent. The more bureaucrats there are, the more red tape the system can support. If you drastically cut staff, the first order effect is these orgs will be massively backlogged. The second order effect could be to figure out how to re-tape everything to work with reduced staff.
The right way to do this would be a first principles approach where you figure out the regulatory scope and sizing to achieve some goal. But as it stands, the regulatory offices operate on blind trust and are mostly immune to electoral politics. So I think there's a safe presumption that they are way overgrown and you can blindly downsize them by huge percentages. Short term breakage is preferred to some 5 year transition plan that somehow leaves everything larger and more convoluted than before.
Scott identified the problem in the article, which is that government bureaucracies aren't doing things for the fun of it, they are doing the things that the legislators told them to do. (Writing annual reports on non-clinical drug testing, analyzing all the kinds of pthaletes(sp?) and deciding which are poisonous, that sort of thing.) Fire half the people, and the remaining ones are still are required by law to write that annual report and analyze the phthaletes.
So the solution rests with Congress reducing what the bureaucracies are required to do.
Is that really in the actual laws? Or is it just an outgrowth of what the policy folks have decided to interpret the laws as allowing them to get involved in?
Doing a naive text search of the entire US code section 21 (the Food and Drugs section) reveals no uses of the word "phthalates", for example. And the only two uses of "non-clinical" appear in text saying that people who want exemptions for rare diseases can ask for recommendations as to what they should do to test them. Expanding that to "we have full control over non-clinical testing" is, well, not something you can blame on Congress. Except to the degree that Congress doesn't actually write specific rules and just says "yeah, the secretary will make rules on what this means."
If the report on phthalates contains ten pages of core functionality and 990 pages of "yes we thought of that, please don't sue" ass-covering, maybe cutting the workforce in half will mean they put out the same ten pages of core functionality and only three or four hundred pages of ass-covering... without actually getting themselves sued any more often, nor to meaningfully greater effect, because the resources available *to the people who'd be inclined to sue them* haven't actually increased.
Go read about the judicial history of the National Environmental Policy Act and report back on your findings and whether you think this would still work.
I think the key theoretical tool missing here is the distinction between line and staff.
In the example here of just filling out forms towards a 1000-form goal to approve a drug, that's clearly some kind of line function for the bureaucracy. They do the main frontline productive work in terms of the demands on the bureacratic institution, like the FDA. I think most criticisms here hold true for the line -- you better think about why they need to fill out those forms to begin with, etc.
But I think people tend to instead think of the bureaucracy as non-line, staff functions. Especially if you're in the private sector, where that's basically what you mean by bureaucratic overheads. And then they generalise that to the entirety of public bureaucracies, since they sort of parallell a staff function for society-at-large.
Not only do these staff functions easily bloat, capture leadership etc way beyond what's justified, but also impose even more work on the line, as a sort of bureaucracy-within-bureacracy. Please fill out this form. Please attend this seminar. Even something like IT does this: a lot of new IT systems offload work on the line IKEA-style (instead of a supporting admin staff doing meeting room booking, you have to do it yourself via a nifty system).
Recent years' discussion on DEI and HR departments is the culture war spearhead of this, but you can also discuss, idk, agencies' individual communicatins departments and so on.
You still could probably not just fire half of staff function personnel at random, there are still externally imposed limits and so on. But I think to a first aproximation "bureaucracy as bloated staff numbers" holds true for them in a way it doesn't for line, and this post seems to assume all public bureaucracy functions as line staff.
Obviously Vivek is being hyperbolic but let's take a real example of a country that is doing something similar to what DOGE wants to achieve.
In roughly a year, Milei's government has cut 7% of the government workforce, roughly 30,000 public sector jobs.
Is the country falling apart? In my personal experience, having spent 23 days in Argentina (leaving as I write this comment), the country seems to be doing OK.
> 7%
the guy wielding a chainsaw and called an extremist made a 7% change?
Why even bother considering being a moderate, it seems all politics is slow.
He's been in charge 12 months, cut him some slack. The chainsaw sits at his desk, and he will continue to use it.
But his cuts have already yielded a budget surplus.
Was that 7% selected randomly?
I think 7% is just what the government has been able to achieve in 12 months, I am not sure what the goal was / is
A country like Argentina has a huge amount of people in unprofitable SOEs and obviously make-work government boards that it’s easy to find places to trim the fat. They aren’t exactly there to process millions of regulations.
In the US bureaucracy is huge because it’s essentially a massively parallel computer. The biggest civilian employers—over half of the workforce—are the VA, HHS, Dept. of the Army, Navy, Air Force and DoD. Cutting these jobs doesn’t decrease complexity. It only reduces processing speed.
Most of the federal government is military or homeland security functions . I'm not sure Argentina has to worry about this stuff too much. Also, doesn't your point kind of disprove itself somewhat? Isn't the US doing OK?
Vivek and Musk have experience in the private sector, where you cut jobs first and then the remainder figure out how to work more efficiently because their bosses are driving for increased productivity, with compensation structures aligned accordingly. The proper question is whether that works in the public sector, where as you point out some of the inefficiencies are due to externally imposed requirements and also where compensation structures are aligned with seniority, not efficiency. But it seems to me this is a better framework for understanding DOGE challenges.
Not addressed in the post but relevant: A lot of these processes and regulations were put in place to AVOID having "faceless bureaucrats" making decisions with high impact on the lives of citizens. So if you want to start turning 1000 forms into 100, you're going to have to grant a lot more discretion to faceless bureaucrats.
(I realize that "passing laws" and "deciding things" aren't the same thing, but they're both types of power and authority. If the folks who don't want them passings laws are fine, actually, with giving them more discretion then they currently have, that'd be great, but that's what I assume they want.)
I think this is really important. Almost all formal regulations are put in place to constrain the discretionary power of the organization. If you could trust people to make the right decisions based on a short statement of values, there would be no need for detailed rules explaining how things need to be done. Behind the right wing desire to cut bureaucracy is really the desire to take power away from the organizations that make up the state. But if Congress insists on having regulatory organizations like the FDA, FAA, etc., then there will be a push to write tons of regulations to prescribe exactly how the organization should carry out its mission.
IIRC the Republicans have appointed more judges than Democrats in recent years, so they have an incentive to move decision-making power from the bureaucracy to the courts.
I think that might actually be a good thing. With the Progress Studies folks, we've seen a lot of nostalgia for how quickly stuff could get built before the 1970s - that was back when the regulatory process could consist of "1. Go to bureaucrat in X office to get approval; 2. They grant approval after review; 3. Project goes forward".
The closest thing we've had to that in recent years was when the FCC was given the power to tell obstructionist local governments doing NIMBY fear-mongering about cell phone towers to pound sound - "put up the tower, and we'll give you some leeway in where it can goes as long as it gets put up on schedule".
Indeed! Give more power to the unelected, faceless bureaucrats! :-D
But on the other hand the bureaucrat in question was often Robert Mose, and the stuff that got built was horrendous city-ruining garbage in many cases.
Or simply have the officials involved in fewer things. If the FDA was a pure research organization, instead of an approval, research, enforcement, and rulemaking organization, you'd need a lot fewer people and it wouldn't matter so much
Of course, that would have other effects...
They actually often have a lot of scope for discretion but don't want to use it because fucking up some law badly and making a big mess is about the only way they could ever lose their job.
Ramaswamy's approach to bureaucracy seems to be based on Parkinson's Law formulated in the 1950-ies. Citing the Wikipedea, it states that "work expands so as to fill the time available for its completion", and "the number of workers within public administration, bureaucracy or officialdom tends to grow, regardless of the amount of work to be done. This was attributed mainly to two factors: that officials want subordinates, not rivals, and that officials make work for each other." The evidence includes, in particular, "increase in the number of employees at the Colonial Office while the British Empire declined (he showed that it had its greatest number of staff when it was folded into the Foreign Office due to a lack of colonies to administer)." So it least in some instances cutting down the number of government employees is justifiable.
Sure, but the US government now has more functions and fewer civil servants per capita than 50 years ago. Seems like removing civil servants won't address the issue right now.
I've always been in the incentives camp. My wife's dad had a horrible experience working for the government: He just simply worked for 8 hours and did all the applications that he managed. But that's wrong, you see! Everyone else did some low fixed number of applications per day, so they got angry at him for making them look bad. In the end, he spend the first 3 hours or so every day doing the work he was supposed to do, and spend the rest doing whatever. I've heard similar reports from multiple people who worked at the government, and to a lesser degree in very large legacy companies. People just simply have no reason to work hard *OR* efficiently.
I've seen similar.things in the private sector.
As I said, especially large companies in a field with significant moat can exhibit some of the same tendencies, since the incentives can be likewise screwed up.
That's been my experience in every job I've ever had; even the very well compensated ones.
If you work too hard, all you do is screw yourself down the line.
Even in my contractor days If I worked too efficiently with time and materials, I would get lowballed in future negotiations.
>Even in my contractor days If I worked too efficiently with time and materials, I would get lowballed in future negotiations.
That doesn't even make sense, though; You're not competing with some imaginary version of yourself that always works exactly as hard and as efficiently as yourself, you're competing with other contractors. If you work harder and more efficiently than them, you can still ask for the money other contractors demand and take the difference for yourself. If they lowball you, the correct response is "F U" and go work for someone else.
Which is the reason why contractors generally have the opposite problem; They tend to overwork themselves to keep up. Which incidentally is the experience every single self-employed I know. With one exception: Working for the government or other monopolies as a contractor. But that's again the same kind of market/incentive failure I'm talking about.
Your customers remember what you did, and bargaining isn't logical.
In situations where things are time sensitive getting it done fast might be worth extra; my experience was emergencies get done on time and materials, non emergencies get done on contract and coming in too fast and too efficient is bad, though I will note the worst of it was from working for giant corporations.
They would see that someone was eg. doing PM on a piece of equipment in 3 hours instead off four, note that down, and next time the contract came around they would itemize the hours and fight for a specific flat rate.
Basically, it was never worth it to not quote a fat margin and then spend a couple hours holding the walls up, instead of quoting a thin margin and getting in and out.
This was replicated across all the major companies we worked for in this field, excepting one which is known for not being particularly capitalistic internally and being family owned/operated, even though they are pretty big.
Having recently spent all day at the DMV only to get absolutely nothing at the end, I've become radicalized beyond the capacity for rational thought.
https://www.youtube.com/watch?v=UFnhHtIGkN8 but in real life
Do they have express DMVs where you are? I recently went to this one and was shocked at the efficiency: https://www.yelp.com/biz/license-express-new-york
After waiting all day, the clerk said we'd already lost as soon as we got in line at the start of the day because we haven't made an appointment. Hearing the crowd's grumbling since the soonest appointment is three months away, she said "you gotta call the phone number, not use the website."
I get home and call the phone number. They tell us "we can't do anything here you can't do on the website." The in-person clerk didn't care, but she got the crowd out of there thinking they had a solution.
A big part of the issue is being underfunded and we *do* need more workers ie more bureaucrats. Probably at higher pay because at minimum wage you only get people who decide part of their job satisfaction is telling people who've been waiting all day they get nothing.
In my experiences with the DMV in the US, they should be exhibit A for the proposition that work expands to meet the number of people available to do it.
My understanding is that DMVs in the US insist on an in-person meeting to do all sorts of things which could easily be online forms.
dmv commenta are weird to me because in the two states I've had to deal with them, I've never had a bad experience. even at the super busy location I did my first driving test, they had a long line... that was moving quickly. no more inconvenient than shopping at costco.
Yeah, in Ohio I've regularly walked in without an appointment and been out before the end of my lunch break. Maybe it depends on the location.
Seriously. My DMV is incredibly efficient, friendly, and fast. You do most everything online, and if you need to go in you make an appt online, walk in, and you're done in 10 minutes. It works substantially better than many private sector services.
This is the case now in Illinois. Our DMV used to be no less rage-inducing than is depicted in popular media but it is now a miracle of efficiency at least for residents doing routine stuff. I know several people who've had that experience (am married to one of them).
I myself had to go do a not-routine thing late last year because of stupidly allowing my drivers license to expire. Back in the day that would have meant a long line and re-taking some old tests and etc so that's the ordeal I was prepared for.
But in fact it was smooth and simple -- even with some (understandable) proving-who-you-are steps I was in there no more than 30 minutes total. Also no civil servant gave me a fish-eye for being a sloppy knucklehead which years ago they would have....how and when that change happened I have no specific idea. Must have been during the past 15-20 years though just based on my own experiences.
I always hear about Americans queuing for hours at the DMV, but what do you actually do there? Do you have to go in person to register a new car or something? The only time I ever had to go to our equivalent of the DMV was when I imported a foreign car.
I think it's fairly exaggerated - I've been to the DMV maybe a dozen times in the decade and a half I've been driving, and it's never been a terribly long wait - often very fast, maybe 15 minutes - I don't think I've ever waited close to an hour.
Getting a license, *maybe* renewing a license (though I think I can do it online), getting a new license if you move across state boundaries (probably one of the more unique 'American' aspects). Last time I was there was updating the title on my car to add my wife and I was in there for maybe 20 minutes total and half of that was the actual paperwork.
Probably would be worse if I lived like... downtown New York City (which is hugely overrepresented in media portrayals), but I don't know, sometimes city bureaucracy is more efficient, sometimes its not.
I can do a lot on line, but certain key things, like getting someone their first-time driving test or ID, need to be done in person.
You need a state ID to vote, even if not driving, that you get from the DMV, and not having appointments is a way of low-key disenfranchisement and then saying "whaaaaaat, why can't you get an ID???" (And I'm generally in favor of voter ID! But this is some bullshit.)
>You need a state ID to vote
That varies by state. California, for example, works on the honor system. The more binding requirement is that a lot of the banking system requires ID to open accounts, and that you need to show proof of identity and citizenship (or visa status) for most non-self-employed jobs.
Off the top of my head, I've been to get my learner's permit and first driver's license, when signing up for a new license when moving to a different state, once for registering a private-party car sale, once to sign up for a RealID Act compliant license (*), and once to process a legal name and gender marker change.
(*) A federal law that went into effect recently that set higher uniform federal standards for state-issued IDs to be usable for federal purposes (mainly for air travel and entering secured federal facilities)
Routine renewals and in-state address changes can usually be done online or by mail. And if you buy a car from or sell a car to a dealer, they are almost always the ones responsible for DMV filings.
My experiences with DMVs varies wildly. The offices I've been to in major suburban areas in California (one in an Orange County suburb of Los Angeles and another in a Bay Area suburb) were every bit as awful as reputed, mainly because they were crowded and understaffed. The one I went to in a Seattle suburb an another in a semi-rural exurb south of San Jose were much better, similar to Retsam's experience.
In my state they partially privatized the DMV offices and surprise they work WAY better!
My reading was that Ramaswamy was just appealing to voters who hate bureaucrats and want to see them suffer.
To be precise: To the voters who don't hate bureaucrats because of the red tape, but because they see them as a nobility of a kind, because they are not bureaucrats themselves and their kids are not going to become bureaucrats either.
A couple thoughts to add as someone who a) is in the process of building a residential home and b) spends much of my time working with large organizations, trying to help them "get things done".
On a) - if you have not built a home, it is an informative exercise that is an exception to your comments above. There are numerous regulatory bodies who have a job to "say yes", but the degree of rigor they apply to the many, many regulations is absolutely directly related to the number of staff / time they have. The more people & time they have, the more it slows down the approval process.
I think you're thinking of regulations (the "checklist of 100 things for the FDA") as a firm set of guidelines with clear rules & each one represents a fixed unit of work that must be completed. That is almost never the case - the rules are most often vague & have multiple interpretations & level of rigor. A town with 1 person will check for egregious violations and say "proceed". A town with 10 people can easily fill 4 months checking every nuance, multiple reviews, multiple internal meetings, etc. (See: HOAs for a multiplier effect on this.)
On b) - there are massive coordination costs that happen with large & complex organizations that basically grind them to a halt, even if they have clear objectives & vision. This is why organizations are constantly going through exercises to reduce layers & manage spans of control.
You could remove half of most large corporate organizations and still maintain the same output. Are government bureaucracies more or less efficient than the average corporation?
HOAs, with which I have a lot of personal experience both as a member and as a board member, are actually the opposite problem: they have _zero_ salaried professionals whose job it is to review and approve/disapprove proposed actions.
The rule-enforcing power of HOA boards is effectively absolute so long as they stick to enforcing their own rules. HOA boards do occasionally stray outside their remit and get into real trouble but that is much rarer in fact than in legend. Sticking with the 99.9 percent of cases in which a member simply dislikes how, or how quickly, the board decided on a matter falling within the HOA's responsibilities, the member is without practical recourse.
HOA board members are unpaid volunteers; there is no knowledge or experiential or educational qualification for election; being an HOA board member is a thankless role which quickly feels for a non-retired person like having a second fulltime job. Those realities mean that in most HOAs most of the time elections aren't even contested, in fact some HOAs struggle to find enough warm bodies willing to take a turn on the board.
So: unappealable decision authority lodged with a small group of un-fireable volunteers whose only qualification for that responsibility was being the only people egotistical enough or naive enough to say yes. Hey what could go wrong....
Well lots of stuff, as everyone who's ever served on or gotten crosswise of an HOA board knows too well. It is though a _different_ malign logic tree from that of big bureaucracies.
Vivek's argument might be not that halving the number of federal workers will somehow remove red tape, but more that it will save money and have no downside because a lot of people in the federal bureaucracy aren't doing anything useful at all. Hence his 'nothing will break' assertion. His model isn't that if you lay off half the form-fillers the others will start filling forms twice as fast, it's that ~half of the people working in those offices aren't even filling out the forms and are either useless or actively counterproductive to the mission of the organization. I think he's wrong (I don't imagine the FDA staff is 50% DEI consultants) but I think that's what he's arguing.
Agreed. If you added up all the people who are doing nothing and are counterproductive, what percent of the workforce would that be? Vivek is saying something along the lines of "a clear majority." Such that you could cut a large number of random people and, on average, not reduce efficiency. If those numbers were a minority, cutting half randomly would hit a lot of productive people. If it's literally everybody (and for some organizations this is possible, especially if you fundamentally disagree with that organization's goals) then you could cut at or near 100% and only see upside. Republicans have talked about removing the Department of Education for years. At this point it's an organization that hands out some grant funding and tells local schools what to do. You can cut the grants (and therefore reduce taxes?) or move them to another organization pretty easily. The rest is even easier to remove because Republicans really don't agree with the rest of the DOE's purpose.
To clarify, the grants are a big deal to public school budgets, but they don't necessarily have to be. State taxes could cover the deficit, paid for by a reduction in federal taxes. Or the federal government could just continue the grants with less direction and oversight so that schools have their budgets maintained.
He can't be arguing that half the people in those jobs are doing nothing while the other half do the work, because he wants to fire people at random, not based on performance.
As the head of the FDA DEI department, I will be hiring only DEI employees with even SSI #'s. FU Vivek. \S
Then a marginally more-qualified DEI manager with an odd SSI# sues you, and Vivek pulls a https://www.giantitp.com/comics/oots1266.html , possibly after publicly flipping a coin, so everyone in your department qualified to run the defense is out.
"a recent success story: Idaho". Idaho needs to be cloned many times.
“I believe very strongly that people on the left are too prone to do things that are emotionally satisfying and not politically useful. I have a rule, and it’s true of Occupy, it’s true of the gay-rights movement: If you care deeply about a cause, and you are engaged in an activity on behalf of that cause that is great fun and makes you feel good and warm and enthusiastic, you’re probably not helping, because you’re out there with your friends, and political work is much tougher and harder." - Barney Frank
I would absolutely love an administration to come in and significantly shrink the bureaucracy in an intelligent way, but because the task is so immense it would take a lot of sober, grueling work that I don't see people in this administration being willing to do. They're focused on publicity, not work, and maybe there's some gain from propagating the idea of large government downsizing, but Ramaswamy's tweet shows they're not serious. I hope they prove me wrong.
Somewhat related: given that immigration seems to be a top priority, I sure hope they don't apply this scorched earth method to bureaucrats handling immigration processing. Wait times are already insane, cruel, and probably illegal.
I don't think Vivek is right about half, but he is right about at least 20%. And we know this is true because the bureaucrats themselves have said so when trying to angle for 4 day work weeks.
Several studies arguing for the 4-day week have shown that productivity is not reduced by reducing the working days by 20% (almost only in regards to government work oddly). So it stands to reason that reducing the workforce by 20% will have the same negligible effect.
Part of the reason people get as much done in less time is because of inefficiencies caused by being tired and burnt out. Most humans work more effectively if they're not overworked. Reducing staff wouldn't have the same effect.
That is one theory. Another is that the "work" expands to fit the time.
> (you can’t prove you’re treating the animals in a certain way because there are no animals)
https://en.wikipedia.org/wiki/Vacuous_truth
I work in navigating federal bureaucracy (pesticides instead of drugs) and this matches my experience pretty closely. The agency has been steadily losing staff over the last few years, and the result is actually worse than Scott's toy model would predict because approving applications is only, say, 75% of what the program does. The other 25% (or more) comprises statutory obligations like periodically reviewing approved pesticide actives, figuring out how to do (and then doing) Endangered Species Act assessments, etc.
No one is going to sue the agency over not approving pesticides fast enough, but they will absolutely get sued over the Endangered Species Act. So if you cut the number of bureaucrats in half, that entire 50% reduction comes out of approving applications, which now take 3x as long to process, not 2x.
Also: non-clinical/non-animal methods are an issue in pesticides too, and my understanding is that this is largely driven by NGOs/public pressure and, to some extent, companies wanting to be able to say they don't test on animals. Thing is, replacing animal methods is hard! Biological systems are complicated, and you don't want your in vitro test system to miss some adverse effect that would show in a test animal.
I should add that there is absolutely a problem where idle (or not sufficiently overworked) bureaucrats will find new things to worry about that almost certainly are not contributing to the public good. Still, I'm inclined to think this is more a problem of legally mandated priorities and institutional culture than number of bureaucrats per se. We live in a large and complex society, bureaucracy is probably necessary to maintaining that, and keeping the bureaucracy well tuned is, apparently, not trivial.
I do a lot of defense of bureaucracy below, so I will say, this is also true. It can be really hard to tell from the outside, but I'm 80% sure that another office I'm aware of has this exact problem, or a related one that's feeding on itself. The office manager insists on reviewing all final work himself, so there's a giant bottleneck, which he tries to fix by hiring more people in his office to feed stuff into the bottleneck.
But they don't have enough work, so they respond by going looking for problems, or brainstorming up nightmare scenarios, or doing vast amounts of research on piddly questions that should be answered in five minutes.
I said it below and I'll say it again, federal hiring/discipline/firing is a giant mess. I have no confidence in DOGE to fix it, but if they somehow manage it, I'll be glad. My fear is that (to the extent they accomplish anything) they'll simply make it easier to fire/discipline for political reasons, which probably won't be great. Whether it'll be worse than the status quo?
Bureaucrat here (another country): This is such an excellent article that very succinctly expresses what I've been trying to explain to my non-bureaucrat friends. Swap drug approvals for anything else (let's randomly say procurement!) and it's the same issue.
The idea of cutting people is easy, the reality of changing the rules is hard (auto-sunsetting clauses are a great idea though).
This makes me lean towards the idea that the DOGE goal is fundamentally to break things further so as to have it all collapse and I guess start over somehow.
I'll say, if you want auto-sunsetting, you need to combine that with streamlining the process of creating new rules, otherwise you're likely to end up with a situation where you don't know how to issue required approvals--because the autosunsetting I've seen only hits implementation regs, not the statutes they're trying to implement.
Oops, I'll note another issue with auto-sunsetting--it most likely restarts the clock on challenges to regulations. Because you're always reissuing within the statute of limitations. So, to take a streamlining example, NEPA authorizes agencies to identify categories of actions which are excluded from NEPA due to their low potential to cause effects. That's done by regulation. I am unaware of any new CATEX's identified in the last 20+ years (note, doesn't mean it hasn't happened, just that I'm unaware) because such a change would be open to litigation, but the old CATEX's aren't, because the statute of limitations has generally passed (note, actions can still be challenged as not properly falling within, but that's a different issue).
If you have to re-issue every five years, that means folks get another crack at suing every five years.
Yes, think you're right here. Auto-sunsetting is functionally unworkable with the way courts interpret the Administrative Procedure Act. Depending on the sunsetting period, you could probably federal government is unable to issue any rules whatsoever or spends so much time trying to issue and renew rules that it can't do anything else.
The only place I would challenge the reasoning here is the density of actual work per bureaucrat. That is, if government employees are anything like typical Corporate America employees (which they often are), they spend half or even most of their time sitting around doing nothing, or waiting-as-an-activity.
It's not like government bureaucrats are at their desk working on one single form, nonstop, for 45 hours per week, getting cramps in their writing hand. Most of the time it takes to accomplish anything in bureaucracy isn't the work itself - it's when you respond to an email and then wait. It's when you submit a form and then wait. It's when you do your little tiny task to advance the process and then send the item back into the machine and then wait.
So although I agree with your point that bureaucracy isn't "measured in bureaucrats," I don't think it's true that cutting bureaucrats would actually slow all things down. In fact I don't even think it would slow most things down. The time loss of bureaucracy isn't in paperwork being done by hand - it's in paperwork sitting in cyberspace until someone gets around to it. And that can be made more efficient by un-stupidifying the system, and isn't mutually exclusive with cutting people out of the system.
It reminds me of the Kissinger's madman tactic to encourage negotiation.
"Listen, I know you want to negotiate a solution that will work for everyone, but my boss (Nixon), he's a madman! He's got his finger on the nuclear button, and I don't have control over what he does..." thus incentivizing negotiation.
Maybe there's someone at doge who will be saying
"Listen, I know we both want to make the FDA more efficient using some reasonable reforms, but my boss (Musk), he's a madman! He wants to fire half your staff, completely randomly, and then expects you to be twice as fast with approvals..." thus incentivizing cooperation.
The goal isn't to actually use the nuclear button, but to use the threat of it to get a better negotiated settlement. Maybe the over looming threat of a 50% reduction in FDA staff is enough of a threat, and credible enough of a threat, that the FDA will be very keen on working something out with the new administration that just makes them more efficient, rather than something that destroys their whole system. I'm sure there's a lot the FDA, and other bureaucracies for that matter, can do to make reforms ineffective, but if they're eager to work with the new administration, to avoid the nuclear option, then maybe the reforms will be a lot more effective than they otherwise would have been without the threat.
I think there's something to this idea. Whether it's explicitly what Musk/Ramaswamy is thinking or not, I'm sure whatever operatives DOGE end up having will naturally fall into that rythm. Even the softer form, "look, I'm on your side here, but my bosses would kill me if I accepted this deal" is the kind of classical negotiating tactic that can work even if everyone sees what you're doing.
In a way, this mirrors the debate about Trump's tariffs as a negotiating tactic vs actually probable policy outcome. I'm not sure what to think about that one, though.
And this works all the way up to the top: Musk himself can negotiate framing himself as the reasonable man while HIS boss, Trump is the madman.
...But the problem isn't that the FDA isn't efficient, it's that they have employees that cost money to employ. Even if they were more efficient, the department would cost just as much money to run.
Same amount of people, doing twice the tasks either greatly improves the throughput of the FDA, making the bureaucracy run more efficiently and slow commerce down less (which is good) or the higher productivity per employee allows for the system to function, while laying off half the workforce (which saves money).
As I understand it, DOGE is not about saving money by firing a bunch of federal employees, as even if you could literally fire half the bureaucracy, you still wouldn't make a meaningful dent in the deficit as it's a small proportion of overall government spending (the majority being welfare, healthcare and the military). The intent is to improve the efficiency of commerce, so projects and economic activity that would otherwise be unjustifiable (try building a new highrise when you have a multi-year long environmental review), become justifiable again. The increased economic output grows the GDP, decreasing the deficit & debt in nominal terms, grows tax revenue, decreasing the deficit in real terms, and slightly reduces spending the bureaucracy (which isn't relatively high), which decreases spending in real terms.
If the plan is just " Let's cut the number of bureaucrats in half! That will fix the deficit!" it is extremely stupid and there's literally no possible way it will work (like trying to afford your mortgage by going out to eat once instead of twice a month). The more complex argument at least has enough uncertainty and plausibility that we can say it might have a real impacy.
Why do you believe that FDA bureaucrats are going to agree to work twice as hard suddenly for less pay than their private sector peers? If you have some ideas on how this is a sustainable or implementable solution without losing the entire workforce with outside options, I'd love to hear about it.
Because we either cut down the number of tasks they need to do for a given level of throughput (perhaps digitizing or simplifying many forms), or otherwise increase their efficiency.
The key word is "efficiency" not "working hard". I'm not an expert on the topic, but from my general knowledge of government bureaucracy, there are very many tasks that are pointlessly complicated.
For example, I recently ordered a form from a government agency (something extremely simple, basically just confirming that a bit of information is still true in state records), and the only option to receive that form is via mail. They don't do email or even fax. As a result my business is stuck waiting for this form (which is a precondition for applying for another government form), and we're stuck waiting a week for the USPS to deliver a letter. That's one week where we aren't starting this new project, a week of lost profits, which compounds across millions of companies all waiting their individual weeks. (The form also cost $125 + $100 to ensure it takes less than 7 days to arrive).
This could be digitized by a mediocre software engineer in literally a week. Maybe the cybersecurity requirements would make this take longer, but digitizing this form would both save the manual work needed to actually send it, and the customer support I called to confirm that the mail was sent. That would save thousands of man-hours per year, which could either justify an increased throughput, or decreased payroll.
As a long-serving government bureaucrat (but military, so I'm different, I guess?), the problems I see are:
1. Regulations that are sensible individually but ludicrous when combined. This tends to be cyclical - politicians complain about how long process X takes, so we cut some steps, then they complain about the output of process X, so we add some steps...
2. Regulations that have a higher cost than benefit, but the cost is diffuse while the benefit is clear. ("Measure twice, cut once" is fine 99% of the time, but 1% of the time measuring thrice would have been better. Those 1% failures can be spectacular, though, and can inspire calls to Do Something.)
3. Lack of incentive for effective implementation. In business, there is both upside and downside risk for most decisions. In government, there is generally only downside risk, so the default is to follow the safest course possible.
This is mostly good, but I'd also add:
4) Regulations (or policy requirements that aren't formal regulations) that have been put in place instead of punishment (this could arguably be 1 or 2). The big one that jumps out to me every so often is conference policy, where the entire thing exists because one agency did a major fuckup in a quite corrupt way planning an extravagant conference for themselves and so every agency suddenly had a 'conference policy' that was usually so broadly written that you suddenly needed high ranking folks signing off on every training course anyone travelled to. Great use of everyone's time.
Social security numbers are not random, the leading digit is assigned by area so some geographic regions would be much harder hit than others
It was firing by the last number, not the first.
I'm hearing comments that Vivek's comment was obviously hyperbolic. I don't discount that as a possibility. But it may also have been a negotiating ploy to ask for more than is wanted. And... ...what fraction of Twitter's staff did Musk lay off?
Over 80%.
https://www.businessinsider.com/elon-musk-chops-twitter-down-1000-employees-2023-5#:~:text=Getting%20closer%20to%201%2C000%20workers%20means%20Twitter's,Twitter%20dragged%20on%20and%20became%20more%20troubled.
Isnt at least some bureaucrats nessery for enforcement
If you fired 100% of the irs, would tax enforcement practically exist?
Oh no, that would be terrible.
In Luxembourg they had to enforce some new tax directives to prevent money laundering. and tax evasion. But since the country makes a lot of money out of quietly permitting those things, they did not really want to enforce them. In the end they hired three people to do it, so they were enforcing it on paper. But three people were nowhere near enough to actually enforce the rules, so they mostly went unenforced.
Depends on the law. For example, you can only sell pharmaceutical drugs with FDA approval. If you dissolved the FDA, that doesn't mean that individuals couldn't still sue you for selling them drugs without approval after claiming that it made them sick.
There's also lots of NEPA lawsuits against solar and wind farms that block the project because the environmental review wasn't through enough.
Really, a big part of the problem is that enforcement is pretty privatized in the US. So dissolving the bureaucracy doesn't absolve you from getting permission and you absolutely *will* get smacked down by private actors. Sometimes you'll get smacked down even *with* bureaucratic approval under the arguments that the bureaucracy was too lenient on you.
"In this moment, the student was enlightened"
It used to be that nations just can't have enough tax enforcers to actually tax their people. It's actually the Hallmark of developed nation to be able to effectively tax more, either because of might or pure efficiency. In a way, if the superpower AI can make and enforce the ultimate tax regulation, it'll benefit everyone and we can fire 100% of the IRS.
The number of people needed to enforce taxes is a function of the complexity of the tax system for the most part.
A simple poll tax is simple, you get a bill and you either pay it or you don't. Much as I'd hate to wake up the Georgists here, a land value tax is pretty simple too.
All the complexity and difficulty comes in once you decide that income is the thing you're going to be taxing.
Yeah there are a lot of simple measures that are employed by pre-modern states. The obvious weakness of poll tax is that you have to tax the rich and the poor by the same measure, and have to cater for both. If you use the poor as standard, you miss out a lot on the rich. If you use the rich as standard, you'll bankrupt the poor. If you use the middle, you'll annoy both. This results in less effective taxation. More complexity may mean more tax income, but it can also backfire horribly
> Much as I'd hate to wake up the Georgists here, a land value tax is pretty simple too.
A land value tax is pretty complicated in terms of bureaucrats, since you have to value all the land. All the proposals I've seen by Georgists include "First, hire a ton of people to assess the land, and have them do it again every year".
It's not that red tape expands to consume available resources, it's that having plenty of resources does not incentivise finding more efficient ways to deal with red tape. Halve the number of employees, and suddenly people will stop using fax machines, and start using GPT more.
Minor point in terms of the post's overall themes, but I'm pretty sure the nonclinical testing mandate is an animal rights thing
My understanding is that companies claiming "No Animal Testing" were using compounds that were previously tested on animals--say for example in China or India--and submitted based on the foreign acedemic study as "safe."
That may be an overly cynical view, but I can't imagine Lush's legal staff being comfortable with using their customers as test subjects.
Good points overall, but I would like to push back a bit on the claim that ban-based bureaucracies wouldn't benefit from a cut.
You mention that the crypto industry is demanding regulation, but this is only *after* multiple bureaucracies started opening up expensive lawsuits against crypto companies of all sizes, in a seemingly capricious way. Most companies settle out because they don't want to be bothered, but the cases that actually go to court tend to side with with defendants, and the judges in some of the cases are not too kind to the bureaucracies who are engaging in these capricious lawsuits against people and companies.
If you were to cut 100% of the SEC, crypto would likely rejoice because now most of the lawsuits against crypto companies would go away. If you cut 99%, there would still be some lawsuits, but the number would likely be low enough that the crypto industry would just absorb them for lower cost than dealing with regulatory red tape. If you cut 1% of the SEC, you still have basically the same number of lawsuits and it maybe takes them a little longer to actually create regulations.
Another thing to consider is that when companies "call for regulation", it is worth noting that this is *exactly* how capture happens. The little companies aren't demanding regulations, it is all of the giant companies who can hire a team of people to implement whatever regulatory requirements there are, and who mostly companies just looking to profit, not looking to enact some meaningful change in the world (true for biotech, crypto, social media, journalism, etc.)
Finally, specifically on the crypto side, this particular bureaucracy (SEC) seems to be particularly bad. Even people within the SEC have openly opined about how the SEC should provide clear rules and lay off on the enforcement until that is done, but the SEC leadership continues to not actually fulfill their mandate of protecting investors and instead just harasses the crypto industry. I suspect many enforcement based bureaucracies are formed of well meaning people trying to fulfill their mandate, but there are at least some that are run by people who don't care much about the mandate, and instead answer to some other master.
One could likely make similar arguments about the DEA or the ATF, depending on your political leaning. Both of these spend most of their resources on enforcement rather than rulemaking, and when regulation is "demanded" it is either by big companies, or so people can stop living in fear of the agency suddenly cracking down on them in a capricious way (see cannabis industry for example).
I agree that the vast majority of stuff in crypto is somewhere on the spectrum of scams. The projects the SEC has been targeting though seem to skew unreasonably heavily towards the few legitimate businesses/projects in the space, while they ignore the rampant scams that are happening.
For example, going after Coinbase (https://www.sec.gov/newsroom/press-releases/2023-102) who has been trying hard to follow the unclear rules, rather than going after people running blatant pump & dumps.
I suspect the SEC's underlying theory is that there is no legitimate purpose to cryptocurrencies in general, so if they can successfully shut down the biggest ones, rest of it will just be a matter of mopping up.
I think it doesn't help the perception of crypto that all the scam blackmail emails require "send this amount of bitcoin to this address, if you don't know how to buy bitcoin look it up online".
That doesn't help crypto to look like it's legit, even though it's not Bitcoin's fault.
How is it not Bitcoin's fault? The exact qualities which make it such a great option for blackmailers, drug smugglers, and the like were a deliberate, integral part of the original design, and that outcome was reasonably foreseeable. Compare https://en.wikipedia.org/wiki/Attractive_nuisance_doctrine
I have the very cynical view that much of the request for "please regulate us" is so that the existing companies can get ahead of that, have the regulations written to favour them, and then they can stifle new competitors who might arise in future by dropping a ton of red tape on them.
Good points on the detail level - but with a birds eye view Elon and Vivek might have a point or two... My own observation in halve a century of life, universe and everything I identified a general law: Mass is attracting more mass - if its in the universe (it all ends in a giant black hole), life itself (cells clump together, build multi-cell organism, which build even larger conglomerates and so forth), bureaucracies, companies, human organisations in general - smaller things clump together and get bigger and bigger over time...
The rare situations a bigger entity is not getting larger are depletion of resources or explosion/war/death.
If you look at each individual example or situation further growth seems inevitable and all the counter arguments for "just cutting the cancer off" sound dull. But practice shows that more often then not someone is just ignoring all the evidence of keeping things as usual and just getting rid of it all. An guess what - "the cycle begins again" to cite my favorite singer/songwriter.
The past years I turned from a huge fan of Elon and his achievements into an outspoken critic. But we must admit - there is a huge track record of awesome stuff that worked besides all the nay-saying. And regarding "Firing halve of the workforce of an overblown org and good things will happen" I look at former and current Twitter (7.5k vs 2.8k, 1.5k in 2023)
But the big elephant in the room for me is this insane military budget. But when we try to learn from the Romans we clearly can see that the Imperium can easily grow some further years till it will implode finally.
Shared my view with ChatGPT and it has some interesting points and views to add:
https://chatgpt.com/c/677fdbb9-8400-800c-ac6f-9b73bdb62a0e
So, I'll say, there are massive problems of government efficiency which need to be addressed, most obviously in procurement and hiring on the bureaucratic side, actually passing bills on the legislative side and remembering that no one actually fucking elected them to impose their preferred policy preferences on the judicial side. Eating Policy (https://www.eatingpolicy.com/) is good on this. I wish DOGE would indeed make government more efficient, but I am...not optimistic on their chances of success given their fairly clear lack of interest in how any of it actually works or the legislative action which would generally be needed to fix it.
Not that I'm a fan of trump nor musk no other oligarchs pushing their private interests down our public throats.
But - in such cases like public or private company internal bureaucracy growing like cancer getting everything to an halt I more and more get the impression that creative destruction might sometimes be the only way out. And either its a strong man not giving a shit on all the "concerns" - or its something even worse like a natural disaster, a war like all the ones we've seen the past decade or two or a civil war like in Ruanda were one group tried to erase the other one after long times of suppression and corruption. We've heard a lot of positive things about Ruanda since woman are now the majority in Politics, Government and Decision making. Similar to Germany after its had lost WWII and most man and infrastructure and woman had to take over. Very depressing as a man getting the impression this might be a actual pattern on earth.
Eh, I'm unconvinced by that argument, but I think the key point your missing is that most (though not all) of the red tape isn't being created by bureaucrats, it's being created by congress/executive action/court action.
Now, it's not the case that none is created by bureaucrats, there's quite a lot that is, and it drives me up the wall, my worst moments are when I'm looking at some regulation or policy document that imposes a bunch of bullshit on top of the bullshit that congress gave us, but, like, I'm guessing what DOGE will actually focus on is:
1) Federal contracting rules, especially as related to Musk companies. I don't know enough about this to comment on source of problems.
2) NEPA, especially as it relates to approving various things Musk wants to do. This is some congress, some executive action, a lot the courts and a tiny bit agencies (indeed, the main thing agencies have done by regulation is identify what is categorically excluded from NEPA...which oh boy do I not want sunsetted).
3) Federal hiring and firing--again, don't know enough, but I believe this is mostly legislative, a result of attempts to professionalize.
> (indeed, the main thing agencies have done by regulation is identify what is categorically excluded from NEPA...which oh boy do I not want sunsetted).
If the sunset clause just says things need to be periodically re-examined, stuff that's obviously worth keeping can get rubber-stamped every few years. If it's a mandate to "remove two for every one you add," might be able to squeeze out a few by consolidating them into broader categories, which then save future trouble by including things that never quite came up before but are, by reasonable standards, sufficiently similar.
So, the problem is that if you rubber stamp a renewal, you're going to get sued under the APA. There's a reason that agencies haven't added to their CATEX's (at least the ones I'm familiar with) because the moment you reopen those to litigation and they're no longer protected by their age, you're going to be stuck in litigation over them forever.
Even from an Irish perspective, if they could cut away a lot of the underbrush from contracting rules, it would really help. You do need to keep the balance between "this is a cosy way for me to funnel taxpayer money to my brother-in-law by hiring his company to do all our contracts" and "this imposes way too much burden on us since the situation here is different so these rules don't really apply".
Oh, it's terrible and then you get contract challenges on almost any big enough contract. To be fair, contract management is always going to be a major problem.
My favorite one, though it's not contract specific is the conversation I always have with new people where they say 'X is categorically exempted from NEPA, so we don't have to do anything, right?'
And I have to respond with hearty laughter and 'No, don't be silly, we still have to do a bunch of work to demonstrate that it's categorically exempted from NEPA.' It's not enough to be right, you need to have been provably, with existing documentation, right at the time you made the decision. Which is fine for 'we're spending 5 million dollars rebuilding a road' and insane for, we're spending 50K doing tree trimming on our parks.
"It's not enough to be right, you need to have been provably, with existing documentation, right at the time you made the decision."
Tell me about it. Regulations for the sake of regulations, in order to CYA from any possible lawsuits.
There's to-and-fro here in the comments about "is it beneficial to have a litigious society or not?" and I do think people are forgetting that "They could build this stuff in the 1950s, why can't we?" is also because back in the 1950s, there were way fewer activist groups and ordinary people didn't immediately reach for the phone to contact a lawyer about everything. In 1950 if little Johnny falls off the swings in the public park and hurts himself, even up to breaking a bone, well that's all part of life and growing up. Today, you bet the parents will sue the park, the manufacturer of the swings, and anyone else they can think of because it shouldn't have happened.
Thing is, you say the military budget is insane and that is what should be trimmed. (I broadly agree, because of course there is bloat and featherbedding there). But another commenter wants to trim the bureaucracy so that the one important element of government can be preserved - which is the military.
So even there, arguing over "we should cut defence spending in half/no, we should increase it to twice the level!" is going to be the problem over "what gets trimmed where from whom?" for reducing the bureaucracy.
> Maybe he is working off a thesis where red tape expands to consume the resources available to it (as measured in bureaucrats)
One argument for this is that countries with a fiftieth the population of the US (like, say, Paraguay or Singapore), which presumably have approximately a fiftieth the bureaucrats, still end up issuing approvals for things on, more or less, a similar timeframe (usually faster, the US is unusually slow at things). There are some economies - smaller countries presumably have fewer individual projects to approve for things like new factories, and mooch off the FDA for things like drug approval (e.g. some governments might just have one guy whose job it is to read the FDA approval notice and go "yeah, looks solid"). But these are still far from linear, and yet we don't see approvals for things in Paraguay or New Zealand take fifty or even ten ties as long as they do in the US.
So, the issue is 'what are the requirements for approval' and though bureaucrats (hi!) can and do create some of those requirements, most are made by congress. We're a rich country, so we choose to place more requirements on approvals than most other countries. The obvious flaw with your size argument is that increasing population increases the number of approvals requested, not just the number of bureaucrats. For the FDA this is only somewhat true, but for, say the Clean Air Act, or Clean Water Act, as discussed below, the number of approvals grows pretty much linearly with population/land/land ownership.
The easiest fix would be to increase bureaucratic discretion across the board, including judicial discretion, that allows every bureaucrat to unilaterally bypass forms and red tape, first in exceptional cases, then in more cases as bureaucratic efficiency increases. The problem is that legislators don’t like to give executive and judicial discretion because it cedes power. Legislators are the problem. https://wwnorton.com/books/fewer-rules-better-people
Your book experience and necessary research seems like it would be relevant to the discussion. Could you offer more insights?
What could or should Trump do this month to dramatically reduce red tape, bureaucratic sclerosis and inefficiency?
The primary driver of inefficiency is decades of lawmaking requiring the civil service to abide by a variety of rules (The Federal Acquisitions Regulation, National Environmental Protection Act, Administrative Procedure Act) and the removal of bureaucratic discretion within agencies from "street-level" bureaucrats. Trump has minimal control over this, and if he replaces the bureaucrats with party loyalists, it'll create an even worse problem. First, partly loyalists are just bureaucrats who are taking a different top-down mandate, now from a single executive, and the same problems arise; they have to check a decision against a mandate and will be unable to break from the mandate when a situation requires it. But secondly, the best bureaucrats are the ones who do not blatantly violate the laws, but ones who know how to bend them in particular places so as to get a job done while avoiding a lawsuit. The Trump-style loyalists do the opposite; they openly and brazenly announce their violations, and that makes their decisions subject to litigation that ties things up for months or years. That's not efficiency. Unfortunately, this is mostly a legislative problem which require legislative solutions.
That would be solved by replacing all of the bureaucrats with party loyalists, no?
I predict that's going to make the problem worse, see my response to Swami. The person to read on this stuff is Francis Fukuyama.
I still don't see why they can't just... ignore the litigation. There's no reason they have to cooperate with leftist courts, just like the US doesn't need to let the Hague try them for warcrimes.
You are something else. 🤣
> If you don’t, you end up like Ramaswamy, who seems to think that halving the number of bureaucrats will halve the number of forms that need to be filled out. I think in his worldview, the FDA will think “Now that we have fewer bureaucrats, it would take forever to complete our current process, so let’s simplify the process.”
I read it a little bit differently. Much of the DOGE effort is focused on budgetary issues, and trying to find ways to cut the deficit. If you fire half the bureaucracy on day 1, you're saving a whole lot of money from the very start.
Meanwhile, promulgating new rules to repeal old rules (the formal process by which "the number of forms" gets reduced) is a long, laborious process that can take months. (Fixing *that* is up to Congress.) This is likely to also happen, but it's more of a medium-term goal. Cutting costs is a short-term thing that can happen on day 1.
And if the rumors on remote workers are even partially true, they don’t even need to fire anyone. Just require 90% of workers to come to the office 5 days a week, and a significant number will quit.
TBH I think that's a bad idea for the same reason I think private-sector RTO mandates with the same goal are a bad idea: it's blatantly illegal. Creating unpleasant working conditions to push someone into quitting so you won't have to fire them is known as Constructive Dismissal, and it's a good way to get sued.
It'll be interesting to see whether you could get a five-day workweek interpreted as "unpleasant working conditions." I would be surprised, but I guess it depends on what kind of jury you get.
Remote workers work a 5-day week too. I'm one of them. My entire team is. But forcing the extra expense and stress of a commute on people who are accustomed to working without one — especially in areas that get winter! — is clearly unpleasant. Especially when, as Swami suggests above and some businesses have been distressingly clear about, the RTO mandate is being done with the by-design intention of making things unpleasant and causing workers to quit. Attempting to willfully commit Constructive Dismissal with a clear mens rea is likely to be pretty persuasive to a jury, even if they don't personally find the means by which it's being committed to be all that unpleasant.
Yeah, yeah, sure they do. I'm familiar enough with what "work from home" means that you're not gonna take me in. But you might be right that a typical jury would be convinced you're working just as hard, if not even harder, as if you were in your office.
If you're trying to insinuate that people working from home are not actually working, you're not familiar enough with it at all. The data is pretty clear on this point: productivity is typically *higher* for workers who work from home. Exactly why is open to interpretation; my personal take is that it's a result of higher morale. But the facts are really not in dispute.
I was working from home, now I'm back in the office. Could I prove to a jury, if necessary, that I worked as hard from home as in the office?
Yes, easily. Provide evidence that payroll ran on time, accounts were processed, emails answered in a timely fashion, returns submitted, etc. as per my normal work.
Did I sit at the desk every single minute of that time? No, but I was *more* productive precisely because in the slack periods, I got my domestic work done during the day instead of having to cram it all in after returning home from work in the evenings. That meant if I was working on something that ran over normal working hours, I could work on it until it was finished, instead of having to rush out the door to make sure I could get home and not miss my transport.
Now I'm back in the office, do I sit at my desk every single minute? Yeah, more or less. Do I *work* every single minute? No, because there are slack times. So I surf the Internet, do online shopping, etc.
If people don't have the 'work ethic', then it matters little if you force them to return to the office. You will then have to have a supervisor or manager standing over them to make sure they are working, which is a waste of the manager's time and will either slow things down (as the manager is not getting *their* work completed) or will necessitate hiring on 'standing over the workers' managers, which is not going to reduce headcount and expenses.
"They're making us go to the office and work! These are inhumane conditions! We'll see you in court!"
Is there a way to, collectively, bring about a culture where fewer people get sued?
I had an odd personal experience where I went to a fitness activity (think squid game but pretty banal) and they had a long completely unenforceable contract that basically said 'if you die for any reason you can't sue us', for some reason I got grumpy about this and didn't sign, and the rest of my friends had a great fun completely risk-free time.
Unenforceable non-competes I think similarly have a lot of value, where you can create a non-enforceable culture of people being willing to take risks and collaborate with a much lower chance that you will lose key assets to competitors or employees leaving to make a startup with your intellectual capital.
I feel like I would genuinely be better off if I could sign a contract that allows my doctor to kill me without risking liability, because I think in general the doctor wants to help people, and is laboring under terrible incentives on that goal. Even if this contract wasn't 'real' even if it was unenforceable, I think the doctor would give me better treatment if they thought I was the type of person to sign this contract.
"It means they’re stuck in regulatory limbo."
I suspect that's the goal here. You will need dear leader's approval to get out of regulatory limbo so your industry can survive.
Imagine doing "whatever it takes" to cut $2T in federal expenses, or doing something as silly as laying off 50% of all government employees. If that costs you more than $2T in output per year, growing at the pace of real GDP, then the initial cost cutting won't have been worth it. Scott asks the right questions - WHO is worth laying off? WHAT regulations are worth removing from the books?
I'm sorry but if you're patting yourself on the back for removing a regulation designed for an unaired show, you might as well start putting on the clown paint on your face. Isn't it more likely that some form of the Pareto distribution is at play here, with a minority of laws causing a majority of the latency and unproductivity of the government? Similarly on the human capital side, isn't it more likely that a select few people at every layer of the federal machine are responsible for a majority of the stalling and enacting of red tape?
Also, if we're saying that the federal government is a burden on taxpayers, it's not so much because people feel like they're paying an ethereally high amount of taxes, it's because people feel they're paying too much and getting too little in return. Whether that comes from the state of infrastructure around where they live, the curtailed speed of innovation they're seeing in their industry or whatever, people ultimately see their taxes as an investment and they're unhappy with their return. It seems to me that blanket budget cuts and layoffs would increase chaos and do little to reduce the faults of the bureaucracy.
If you want a better return on your investment, you typically have to accept more risk per unit of return. If you want a meaningful change in how people perceive government spending, you're unlikely to get there with something as "simple" as laying off 50% of the workforce - you need brand new ideas that won't please everyone, and thus are much harder to actually be put in place. You might want to start introducing more competition for government jobs, by increasing salaries meaningfully and reducing job safety, like making it depend on regular performance reviews. You might want to start having clear goals for people instead of letting hundreds of thousands of people coast their way to retirement, or create fake jobs for their spouses (stuff I read in comments on this post). But this would involve seeing the government slightly more as an enterprise rather than a safety net. The majority of its spending could still be done to support lower income people and at-risk populations, but by making it harder to get involved and retain a job once you have it, you'd naturally generate competitive dynamics and, hopefully, better outcomes for the people's investment in their country.
Note this isn't a perfect solution, in fact I could already see many flaws as I wrote this comment, but the brunt of my point is that you're not gonna get to a visibly better situation by slashing useless regulation (because you need to slash "currently in use but problematic" regulation, which is going to anger people on the being slashed side), or by massively reducing the workforce (for reasons explored in Scott's article), or by doing any "easy" stuff. You could of course optimize the process by singling out actually problematic regulations and erasing them, picking the low hanging fruits, but once you've cut the 50 most annoying regulations, you'll still have a whole lot of people screaming at the size of the government.
By what metrics are you judging efficiency in this case? Yes, removing pensions and entire departments would be efficient if your mission is simply to cut costs. But when you cut cost in your own life (say deleting a bunch of subscriptions, or changing your energy provider for a cheaper one), are you not being very intentional with the costs you've chosen to eliminate? If I took a kitchen knife to your subscriptions, maybe I'd end up removing one that, upon further inspection, provides you with a lot more output than it costs you. Maybe if I force you to destroy your current home energy infrastructure, I cut costs immediately but then you have to make a large investment in Tesla electricity generation that has a 10-year payback period.
There's part of me that wishes they actually would try some insanity like firing half the civil servants. 98% chance everything breaks, Ramaswamy looks like an idiot, and everybody appreciates the important work of civil servants a bit more (with the downside that lots of things are now broken, but hey, I don't live in the US and I appreciate their sacrifice to teach themselves and the rest of us a lesson). 2% chance everything's fine, Ramaswamy looks like a genius, lots of public money is saved, the rest of the world gets a remarkable example to follow, and I get to have my opinions changed.
Win win.
Small typo near the end:
"Or are they more like phthalates, wwhere environmental groups and industry groups fought each other"
I worry that "repeal two old regulations per one new one" would have a perverse incentive, unless you can 'bank' repeals. Otherwise, you'd want to delay removing regulations until you need to pass a new one, rather than the ideal case of removing regulations as soon as they're determined to be unnecessary.
Presumably part the goal is just to get folks combing through the existing regulations, running a sorting algorithm weighted by willingness to remove.
Lot of what's found will be things that are mostly useless, but also not urgent, and a list of those can be accumulated as easy 'permission slips' for adding future regulations... but that list is also a clearly non-renewable resource, so it'll force some degree of cost-benefit consideration when any new additions are being considered, meaning a higher floor for the incoming material's quality.
When such a search turns up some genuinely terrible regulation which deserves to be removed ASAP, it can simply go on the top of the stack, get promptly paired with whatever was already latest under consideration to be added. Even if the new addition turns out to be useless, and the excised blight only counted as one of the two required, that's at worst break-even on the stock of permission slips.
Ongoing process of cost-benefit analysis hopefully means there'll be beneficial but, again, non-urgent proposals sitting around in an informal queue, getting debugged and fine-tuned while they wait for such an excuse, rather than half-baked concept sketches pushed straight to production and forgotten about on the basis of "something must be done => this is something => something has been done."
Former lawyer at U.S. Environmental Protection Agency (8 years in the New England Regional Office) with three specific (and distinct) examples showing the naivete of Vivek’s view.
Example #1, an example of the folly of the supposed distinction between agencies whose job is to “ban” vs. “approve” things: The Clean Water Act, which has the stated goal of ultimately eliminating all discharge of pollutants into waters of the United States, prohibits discharge of pollutants into those waters except with a permit that applies (I am greatly simplifying) a combination of technology-based and water quality-based standards to that discharge. A newly built facility that wants to discharge pollutants needs to apply for & obtain a permit, and then the Clean Water Act requires that it must be renewed every 5 years, but an expired permit is valid until the permitting agency issues an updated permit, and most permitting agencies have a backlog. The renewal process involves (again simplifying) looking at technology-based and water quality-based standards. (Also, both the original and all renewal permits are subject to public notice and comment.) Sometimes a renewed permit is just reissued because there’s no reason to change anything; sometimes it gets more stringent (e.g., because the water quality of the receiving water has degraded). A company that wants a new discharge wants the agency to move quickly and approve a permit; a company with an existing facility, existing discharge that it doesn’t seek to change, and an existing permit would be happy if the agency never gets around to it.
The people who write Clean Water Act permits (engineers, supported by biologists and lawyers) are the same people whether it’s a new permit (which companies want approved quickly) or a renewed permit (which companies, simplifying, want to never happen); writing new permits vs. renewed permits aren’t separate skills. In fact, it often makes sense for the same permit writer to be attached to a particular permit over the course of their career, for institutional knowledge.
So, simplifying, if you cut 50% of permit writers, you’ll make it harder for new construction and facilities (whose discharge might not even be that bad if they’re using modern methods) to get approved, and easier for older, out-of-date facilities to continue very bad discharges.
Example #2, on why banning & approving aren’t separate agencies: I worked on both the Clean Water Act and the Clean Air Act. (At EPA's DC headquarters, people focus on just one statute or even part of one statute; at the regional offices, where the rubber meets the road, people are sometimes less specialized.)
Clean Water Act permits are weighted more towards renewals (because of that 5-year renewal requirement, and the statutory goal of eventually eliminating all discharges to water), whereas Clean Air Act permits are weighted more towards new facilities (simplifying, new big facilities with an air pollutant discharge need a preconstruction permit requiring them to meet technology and air quality-based standards; older facilities face far fewer requirements). If we want to simplify, we could say that most of my Clean Water Act work was more about cracking down on older facilities (“banning,” though not really – a discharge is almost never banned, it’s just subjected to more stringent requirements than the facility owner would prefer) whereas most of my Clean Air Act work was about approving new facilities (“approving”). If Vivek had fired me based on my SSN, he’d have eliminated both "banning" and "approving."
Example #3, on why sunset requirements for regulations can actually increase bureaucracy for no good reason: You cited Idaho’s automatic sunset requirement for state regulations. Here’s a counterexample to consider. The Clean Air Act requires states to develop State Implementation Plans (a package of regulations) to attain air quality standards that they’re not presently attaining, and EPA approves those regulations. New Hampshire developed some regulations to control e.g. volatile organic compounds (I don’t remember the details – stuff like requiring gas stations to have certain technologies to limit VOC release), and EPA approved them – they met Clean Air Act requirements. The problem is, New Hampshire (I guess like Idaho) required all state regulations to sunset after I think 5 years. But if New Hampshire had let its volatile organic compound regulations expire, EPA would have been required to step in and taken federal control over New Hampshire’s gas station regulations. Nobody wanted this. So, like clockwork, every 5 years New Hampshire would need to re-enact its volatile organic compound regulations (time spent by the New Hampshire Department of Environmental Services staff), and EPA would need to approve them (time spent by EPA). In most cases, this was pretty routine, because NH wouldn’t change the regulation much or at all, but we did need to check – we couldn’t rubber stamp its proposal without actually reading it. Now consider that NHDES needed to do this for many air pollutants, not just VOCs. Thus, as a result of NH’s requirement for regulations to sunset, a considerable amount of time by both state and federal regulators was wasted in re-enacting and re-approving regulations that everyone agreed were perfectly adequate the first time around. This actually ended up consuming a fair amount of my time.
By the way, if Vivek had eliminated my job, EPA would still be under a legal obligation to approve the (new, but not really) NH regulations, and if it didn’t, environmental groups could have sued EPA, in which case a different EPA lawyer from my office would have been required to stop whatever they were doing – perhaps approving an air permit for a new facility that would bring jobs and electricity! – to work on the lawsuit.
I may think of more examples and if so I’ll add them.
This was fascinating, thank you for telling us about your personal experience.
Re example #2: what is your personal view on the best use of your time, between the more "banning" side of it with regards to the Clean Water Act and the more "approving" side with the Clean Air Act? Is the complexity regarding both the same, or is one much more straightforward than the other? If that's the case and let's imagine approving is both the more personally gratifying and straightforward process, couldn't we optimize future Acts by having a more "approving" template in mind for them, instead of spending a lot of time penalizing older facilities? Similarly, do you think a more "approving" slanted procedure would lead to more suing and unrest, since the older facilities with the most problematic pollutants would see their life majorly extended (even more in the case where the new Act is seen as very stringent and costly)?
Re example #3: can you see the benefits of a more specific 5-year sunset requirement, based on the (probably many) cases like the one you're detailing here? It would have a less dramatic impact on the number of pages of regulation in a State's code (fewer brownie points with voters for the Littles of the world?) but it would allow the sunsetting of specific regulations that are not anticipated to generate more work than they're eliminating. Or is your view that sunsetting as a tool is fundamentally misguided?
Thanks for the kind comment. Again oversimplifying: as a general matter, new things pollute less per-unit-of-whatever-generated than old things, because technology improves. And the statutes are designed to impose more easily impose more stringent technology requirements on new facilities (where it can be priced in from planning) than on old facilities (which may be harder to retrofit).
I'm not sure approving is more personally gratifying or straightforward (these are, after all, facilities that will emit pollutants, and neighbors will yell at you at a public hearing for a permit to build a new one that's going to pollute in their neighborhood) but -- here air and water can be very different -- often approving a new facility that pollutes a moderate amount will help drive an older facility that pollutes much more out of business. An example would be in New England, the construction of new natural gas power plants (which pollute! and this pollution needs to be controlled, often more than the company wants to do! which requires bureaucrats) led, from the 1990s forward (through economic forces that are sort of predictable, though out of EPA's control) to a decrease in usage of older, less efficient oil and coal power plants (which pollute much more, and can be harder to retrofit/control), thus resulting in air quality improvements.
On #3, obviously not all state regulations are subject to a federal approval requirement - most aren't! But my guess is that people overestimate the # of regulations that are genuinely totally obsolete and can be completely repealed (whether automatically or by deliberate action) and not need to be replaced with something similar that addresses an updated version of the same topic. But that's the type of ordinary regulatory updates that all agencies should do anyway - sunsetting the old one on an automatic schedule doesn't really help, since it takes time away from considering which regulations need to be substantially modernized vs. which don't, and forces regulators to waste time re-enacting regulations that were perfectly fine.
People may also over-index on things like "number of pages in regulatory code" as an important measure. But nobody prints out the entire regulatory code (except as a stunt for a TV ad) and has its size as a burden on them. What matters to a regulated entity is not the number of pages, but the actual burdens of the regulations involved. You could write a one-sentence regulation that says "All burning of coal, oil, and natural gas is hereby and forever banned in this state without any exceptions" and it's very short and crisp but it's obviously way more burdensome to all industries than a 30-page regulation specifying control technology requirements for pollution control.
It's certainly true that some regulations are written in wordier fashion than others. It's also true that some regulations are written in more difficult-to-understand language than others. Notably, simple language isn't necessarily shorter -- sometimes making a regulation "easier to read" to a layman makes it longer. This is a value tradeoff.
As for truly obsolete regulations that serve no valid purpose anymore whatsoever, in some platonic sense it would be nice to get rid of them because they're taking up some type of space and presumably imposing some type of burden on somebody, even if rare. But it's usually a very low priority. By analogy, virtually every state has weird laws that everyone agrees are obsolete, like banning spitting on Sundays or playing music involving more than two guitars. They take up pages in the statute books, but legislatures rarely repeal them because it's not worth the time and effort. An automatic sunset provision doesn't really help with this because the time wasted re-enacting regulations that were fine to begin with (harder if you've fired 1/2 your bureaucrats) almost certainly exceeds the time it would take to sort through and propose & finalize repeal of outdated regs (which is also harder if you've fired 1/2 your bureaucrats - with a skeleton staff, you'll focus on the absolutely most essential things, and repealing obsolete regs that aren't bothering anybody will never be that).
There are other ways to reduce regulatory burdens on industries while obtaining environmental quality improvements -- for example, cap-and-trade is a proven method for reducing certain types of air pollution faster, and at lower cost, than "traditional" regulation. However, it, too, requires bureaucrats - someone has to design, operate, and enforce the system.
Another excellent comment, thanks!
This is excellent, definitely highest value added comment thus far. Helpful to see insight into the actual practice.
How can you use Space X to illustrate your point? Is that a straussian thing?
Space X is hounded, amongst other things, by bureaucrats seeking our reasons to sue them!
Less bureaucrats would very likely help that!
So much comes down to the insanity of the American legal system. Whether it's constructing a building, implementing a new regulation, or approving a drug we let anyone with deep enough pockets slow action to a crawl and cost whoever is trying to take action ridiculous amounts of money to push it through.
For your amusement, an effort post I did over at Slow Boring on the approval process (though I will note, if you reduce resources enough, people may just be able to do stuff without getting caught/stopped, whether that's good or bad in any individual case, I think it's pretty bad overall for general order reasons):
Okay, effort post. Let’s walk through a small federal project/approval (apologies for the length of this). This is fairly generic and isn’t a copy of any individual project I’ve dealt with, but rather is a composite which doesn’t reveal any confidential information. I will occasionally add complicating factors I’ve seen arise at various points in brackets, but those are fairly uncommon.
Let’s choose a project most agencies will have to deal with, assuming they manage land. A power company needs to run a power line across their property (note, this simplifies things, if we’re talking a federally pushed project, then there’s a whole bunch of planning that is happening at early stages to figure out where/how to do something, getting this to line up with NEPA is a horrible mess as you can’t evaluate environmental effects until you know what you’re doing, but can’t know what you’re doing until you’ve evaluated environmental effects, so you do this stupid little dance of identifying alternatives, then coming up with reasons to screen them down to a reasonable number to evaluate, then evaluating those, at least for larger projects).
This will begin with a real estate application, generally, requesting a permanent easement to use federal land. [[Or it will begin when they got caught just doing construction prep work on federal property and then act very confused when its pointed out they don’t have any rights to do that yet, after all they got permission from the city/state/maybe even federal grant money or some other permission like a CWA permit, see below, so surely they can just build on federal property!]]
Having applied, the first thing that happens is that the realty specialist (or equivalent, terms vary significantly by agency) will need to figure out how to pay for this work. In some agencies there’s a pot of money for issuing such permissions, in others it's funded by the local office, in others it's funded by the applicant. If there’s just a pot of money, the question is ‘is there money for this action now, or does it go on the pile for next year, assuming we get money then.’ (remember you’re usually dealing with one year money and your not getting any more until next year, hopefully, assuming you’re not furloughed, or under a CRA where you’ll get fractional funding).
If it’s funded by the local office, you need to reach out to them and make sure they’re willing to fund this one, not some other local action. If it’s funded by the applicant, you need to figure out how much it’s going to cost. Sometimes there will be set schedules of costs, sometimes there won’t. A standard choice is ‘we can give it a general estimate and get to work with that, or you can pay us for a more specific estimate, which will cost X thousand dollars.’
Given this is a private action with no benefit to the federal government, we’ll say they’re paying the costs themselves. Let’s say this is a reasonably successful company, they want this powerline, they say ‘just give us a general estimate and we’ll give you the cash, give us anything left over at the end, and if you need more, tell us.’ The realty specialist says okay, send us a check, or make an EFT like this for ~10K. Then they wait ~1 week while the company does that. [[This is quite efficient, many companies are extremely uncomfortable making payments with no guarantee that this is all it will cost, or that they will get their easement in the end. And so there’s often a lengthy discussion about how this works, both educating the applicant and repeatedly confirming that no, you can’t promise anything at this stage.]]
Okay, so, we’ve got their money...except we don’t actually have their money, to be able to use their money, it needs to get from the finance folks, back to the realty folks, so it can be divided up and used to fund the people who will actually do the work. This is horribly inefficient and can take extremely variable times depending on a bunch of factors I’ve got no visibility on. Expect at least a month. No one can do any actual work on this action until they’ve got funding.
Now, they’ve got money to review the proposal. Step one, talk to the field folks to see if this location is available for power lines. Has it been designated in such a way that it’s closed to development? Is it needed for other purposes that the power line would interfere with? Is this smack dab in the middle of a park used extensively by recreators using drones? Is there anything on the ground that means we should tell them no, or we should place limits on it (if it’s in a park, can we route it around green areas? Can we schedule construction work to not interfere with rec season, etc.) In fact, this frequently happens earlier, as this is the most likely way for the answer to just come back ‘no, this shouldn’t be allowed’ in which case you don’t want to bother collecting funds and usually field folks can give you a general thumbs up/thumbs down fairly quickly as they’re familiar with the areas they work.
[[Note: Thumbs down is not necessarily final. Almost all of the things I’ve mentioned are field preferences, or policy concerns, both of which can be overridden if you’ve got a good argument and are willing/able to elevate your concerns. Almost all policies have the potential to have exceptions granted to them, by their issuing authority, if no one else, though even attempting that will blow timelines all to crap. I don’t think most applicants realize this, but firm nos are rare anyway.]]
But this time let’s say it’s a field, owned as a storage area near a facility, so they’re not worried about having a power line there, so long as it avoids the actual storage lots. So, having cleared that hurdle, the proposal is sent up for what’s frequently referred to as ‘EC [Environmental/Cultural] Review’ or ‘Compliance.’ These are where all the laws that Slow Boring talks about are going to be considered. The big 3-4 are: NEPA, ESA, CWA, NHPA. NEPA is your umbrella statute. An action like this is almost certainly going to be categorically exempted from NEPA so you aren’t working with an EIS, or even an EA, you just need to demonstrate it fits within the scope of the categorical exemption, and that no special circumstances kick you out of that. On its own, this won’t take too long, but you generally don’t complete this process until you’ve finished everything else so you can be sure that none of the others kick you over to an EA.
ESA is concerned about endangered species, if there are none affected, and a power line like this is not likely to have such effects, you’re fine. If not, you’ll need to consult with the relevant service (either U.S. Fish and Wildlife Service, NOAA Fisheries, or both, depending on the species/location). There are theoretical timelines on this, but unlike the NHPA, discussed below, failure for the services to respond does not free you to proceed and there are no legal consequences to dragging things out on their end. They are, at least the offices I’m familiar with, chronically understaffed and overworked, so the 135 days for formal consultation is more a joke than a reality. But, this one we’re able to get to ‘No Effect’ so we don’t have to consult at all. We note that in our NEPA file and move forward. If we did have to consult, we’d likely get to ‘not likely to adversely effect’ and do informal consultation which theoretically has a 60 day timeline and they’re okay about meeting that timeline, as far as I know [[Note: no timelines start running until they determine they have a ‘complete consultation package,’ this starts with a biological assessment assessing the effects to the species, but if they have any questions, they’re likely to view the package as incomplete and assert their clock hasn’t started to run, this is one way they’ll defend themselves on meeting their timelines, sure the actual consultation took two years, but we only had all the information we requested in the last two months, so really, we were more efficient than you could expect! This can either be true, or total bullshit, depending on circumstance.]]
CWA is the clean water act, it’s concerned with water/wetlands. Here we’re going across a dry field! Yay, no concerns. If we were, say, on the side of a river, they might need a clean water act permit if they needed to put any fill into the water, (and maybe a section 10 permit if they were interfering with navigation, say by stringing a low power line across a navigable river but that’s a different issue) similarly, if they were going through a wetland and needed to partially fill it to stabilize a pole base they’d need a permit, from the Corps of Engineers. This would probably qualify for a nationwide permit, but that’s not our problem. [[Unless it is because we’re viewing this as one action and the Corps wants to rely on our compliance work, rather than do their own, or our compliance people want to make sure we’ve got the Corps documentation so we can prove our action complies with the CWA, or both--we may well need a meeting, or even to formally appoint one party to be lead federal agency, or just agree to work together and sync up our permissions. Setting up this meeting and making sure we don’t get crosswise with each other can take time.]]
NHPA is concerned with historical preservation. This is a field, so we’re not concerned with historic structures (probably, but note things like levees/irrigation ditches/etc can count if they’re old enough we’ll have to evaluate them), but we are concerned about archaeological properties, as they’re going to be digging holes and doing construction work. So, our archaeologists check our files, do we have enough survey information to determine if there are any archaeological sites present. Shoot, we don’t, at least not in this area. It’s a storage yard, but the fence and handful of pads went in prior to the NHPA, or someone just installed them without talking to anyone and so there hasn’t been a survey done. We could go do a survey ourselves, on their dime, and will if this is high priority/urgent, but this is a small project and our archaeologists time is better spent reviewing the results of five surveys rather than going out and conducting one, so instead, our realty specialist is told, ‘nope, we don’t know enough, have them file an ARPA permit application to do an archaeological survey of the area they want to install the power lines.’
The applicant gets that answer, hires an archaeologist, [[or more likely, comes back to complain about having to apply for a permit to apply for an easement, then asks who they should hire and is extremely annoyed when we cannot make a recommendation, and then eventually hires a contractor]] who applies for an archaeological resources protection act permit from us. That comes in and is reviewed by the archaeologist who checks to make sure that the work they’ll do is going to be sufficient to meet our needs and will generally consult with the local tribes, as we’re going to be consulting with them on the overall project, so making sure that this is going to produce what we and they need, just makes sense. Any time we do that, we give them a minimum 30 days. Plus our own review, plus getting to the top of everyone's pile of crap, ~45 days, if you’re lucky.
The contracted archaeologist goes out to do the survey, but this is just standard work for them too, and they’ll need to get out to this random field, do shovel tests, some borings, then come back and write up a professional report. ~60 days if you’re lucky. [note: the applicant has some control of this, but they’re likely contracting this out and most contractors aren’t just sitting around with big holes in their schedule.] [[Note: if they actually find important/interesting artifacts, or even worse, human remains, then timelines are likely gone to shit.]]
While that’s happening, folks aren’t sitting idle, the NHPA stuff can’t proceed, but you can work on drafting the categorical exemption, just leaving the NHPA bit empty, to be filled in at the end. Okay, ESA isn’t an issue, nor is CWA, we fit the categorical exclusion--shit, I forgot the Migratory Bird Treaty Act. Okay, but this is an open field, but, there are a couple of trees on the edge of the property, we can’t disturb nesting birds. Okay, easy solution, either you do it outside nesting bird season, or you have a bird nesting survey done before you begin work, to make sure you didn’t disturb the birds. We can put that requirement in the easement (or this may be broken into a bigger construction license and a smaller, permanent easement, as the footprints for those two actions will be different, but let’s simplify it and say we’re putting it in the easement, after all, then we can say the same thing for ongoing O&M, make sure they don’t disturb birds).
Okay, archaeological report is back, our archaeologists review it good news, nothing was found, and it’s in good enough shape that we can use it, without having to go back to the archaeologist who did the work and tell them they fucked up their report (or worse, their survey). We’re able to say no historic properties effected by this action. We write up that report (or stick our cover page explaining our determination on the ARPA report), and because basically all these actions are streamlined to minimize everything, we do all the steps that the NHPA imagines you’re doing sequentially at once here, we propose the area of potential effect to be the footprint of the easement, we describe our research to find the potential historic properties and we propose our determination of effect, ‘no historic properties effected’ and we send it to the state historic preservation officer and all the regional tribes (that is, tribes which occupied this area historically, not just the ones which happen to be in the region now). When communicating with the regional tribes, we also make sure to offer government-to-government (now being called nation-to-nation, in order to annoy me and make me learn new acronyms) consultation. If they say yes, all timelines are shot, because we’ve got to schedule an in person meeting between our leadership and the tribe’s leadership to address their concerns. This can take days to months depending on the priority placed upon it and schedules.
The SHPO never gets back to us, because he doesn’t care about this action. But, 30 days pass, so under the NHPA, we can proceed, since he didn’t respond. Most tribes didn’t respond, one did to say they agreed with our determination. We follow up with all the other tribes to make sure we’ve made good faith efforts, they don’t respond again. [[Note, if they do respond and disagree, we need to try to resolve that disagreement, before finalizing. How long that takes depends on how much everyone prioritizes it. We don’t have to agree with them, but we do have to evaluate their input, which means getting it, which can take time, and responding to it, which will take time, especially since they always have the option of going, ‘I don’t like that response, our council requests government-to-government consultation on this matter, which freezes everything until it can occur.]] NHPA process is finished, it’s plugged into the categorical exclusion, which is finalized and signed. The whole package goes back to the realty specialist who pulls out the one requirement being placed on it limiting construction/maintenance activity around nesting birds, adds that to the template easement document and then pulls out the stuff he was doing while compliance was working. [[Given this is a standard one, the realty specialist most likely already knows that EO 14026 (minimum wage for federal contractors) and EO 13706 (sick leave for federal contractors) do not apply to this ‘federal contract’ and so we don’t need to circle back and discuss the way we do for weirder situations.]]
So, this is a permanent easement on government property, they need to pay consideration for this. [[unless you’re the Corps of Engineers and they’re a rural electrical utility, in which case section 1172 of WRDA 2016 means they don’t have to pay consideration, they have to provide documentation proving they meet the standards of the exception to consideration]] How much? We need an appraisal. This is a small action, so we can probably use informal procedures, but that still takes time and money, but we’re running at the same time as compliance and they’re going to take longer, unless the appraiser is really slammed. Okay, so we’ve got our appraisal, we’ve got our compliance terms, oh, right, we need an actual legal description of the easement area, don’t we? Okay, we go back to the applicant, and ask for that. They get us a survey, but we can’t just accept that, they could have given us anything, so it needs to be reviewed by our survey folks and it needs to be turned into an exhibit we can attach to our easement. Again, this is not going to take longer than the compliance, unless someone’s really slammed, or really slow.
This is a low risk easement, not like an oil pipeline, or lease or anything, so we aren’t concerned about RCRA/CERCLA...probably, and so aren’t going to do, or have them do an environmental site assessment to make sure there’s no issues in the long run, so the realty specialist, besides the above, is mainly acting as a project manager and communicating any questions/needs back to the applicant, who is getting increasingly frustrated throughout this at how long its taking and that everything they’re doing is costing them money, on both sides of the transaction.
But, we’re finally done with compliance and we’ve got our draft easement! Now it needs to go through peer technical review to make sure you haven’t made any mistakes/missed anything, then my supervisor’s technical review, then it goes to the attorneys which is where those fuckers [[me!]] look at everything I’ve done and ask annoying questions like ‘why don’t these numbers match up?’ and ‘hang on, let’s not just copy and paste their project description into the easement, I know they’re asking permission for a two-strand power line, but let’s just issue them a power line easement, we don’t care how many lines they hang from it, right? Let’s not set ourselves up to have to amend if they hang an extra line. Okay, go confirm with compliance that that doesn’t change anything, then change the wording and keep all these communications in the file!’ And they answer all my questions and then it goes to the person who can actually sign it. But we have to sign it second, by policy, so you send it out to the applicant.
Then you answer their questions about the certificate of authority and the need for them to both notarize the signature and have a corporate secretary confirm that the person signing has authority to sign on behalf of the company. Then they send that back, along with their consideration check and the person authorized to sign on behalf of the federal government signs and that goes back out. Then you file everything and discover that you’ve got $37 left in the account that they provided and it will cost ~$200 to refund their money and so you’re going to be stuck with this accounting issue forever and have to explain it to every new accountant for the rest of time. Oh, and the company is having a problem, because the county is asserting that they won’t record the easement, because the exhibit map is in color. Can we please re-do it with a black and white exhibit map?
And you’re done! Call it 45 days to get funding, 45 days to issue ARPA permit, 60 days to work on ARPA permit, 45 days to turn that report into a determination and get it complete and 14 days to take that finalized report and get to final document which goes out to the applicant. We’re sitting at ~7 months. Now, you can cut a big chunk out of the middle if ARPA isn’t needed, that’ll get you down to ~3.5 months. But you can also extend out massively if GTG is sought, or anything archaeological/ESA is present/effected.
This system is obviously real messy, because it’s trying to address everyone’s potential concerns on every action and the statutory expectations are built for larger projects, but apply to everything. There are ways to try to streamline, but I’ll point out, simplifying NEPA doesn’t actually solve the problem, because NEPA’s your umbrella statute, even if something gets entirely exempted from NEPA such that you have to do nothing, you’ve still got to comply with everything else and a lot of work is being hidden behind the lag time that is NHPA consultation.
Note, that the NHPA consultation looks like your big time suck, and the smaller the project, the more likely that is to be true, because you’ve got this statutory 30 day consultation clock. But if your dealing with a big project, the NHPA timelines don’t get THAT much longer (worst case is negotiating an MOA to resolve adverse effects to historic properties, which sucks, but is not going to be anything like as bad as the timeline on an EIS) while your NEPA timelines can extend out to basically infinity with a large project.
Also, I know there are a million acronyms there. I defined most of them, but if any are undefined, then that’s because I decided leaving some unexplained would be more realistic to the applicant experience. That’s definitely it, not a mistake on my part.
So, where is there room for improvement? Well, lots of places. An obvious one is that the government’s fiscal understructure is extremely messed up as is the actual software/architecture for moving money around and distributing it. This is extremely complicated though and outside my area, so I don’t have a lot of great suggestions.
NEPA on these small projects is very rarely an issue, the big delays are going to come mainly from NHPA and ESA with their requirements to bring in other parties (the services and the SHPO/tribes, respectively). Any time you’re doing that, timelines stretch, painfully. The NHPA does include a ‘if they don’t respond within 30 days you can move out’ while ESA doesn’t, but that just incentivizes responding within 30 days to raise some issue and prevent the clock from running, so I don’t think it solves the problem, generally (also, putting that on ESA timelines would incentivize shrinkage of the services so they just couldn’t respond). Instead, I’d focus on reducing the circumstances under which consultation is required.
Removing the existence of informal consultation under ESA and moving all determinations under ‘no adverse effect’ (‘no historic properties effected’) to the pile of ‘don’t have to be consulted on’ would be helpful.
If congress and the president could resist the urge to create a bunch of rules and a bunch of individual exceptions to the rules, things would also be more streamlined, but that’s pretty clearly a pipe dream. Then there’s other stuff, like having a single government-wide land use fee/consideration schedule might simplify things quite a bit (though the BLM charge rates being set incredibly low, then stuck there is not an optimistic example, though of course, these funds generally don’t go back to the agency, so this doesn’t have significant operational effects).
For tribal consultation, the threat of GTG consultation requests hangs over every project and timeline, you could simply cancel those rights, but I tend to think the better option would be some sort of standing meeting, as its the ambiguity which messes things up. The tribes have asked for GTG on this and it will be discussed at the next meeting in two months, isn’t a message anyone’s going to love, but it’s better than, it will be added to the next GTG which we’re still trying to schedule. The problem is the number of agencies and number of tribes and the tribal position that GTG only occurs on an individual tribe-agency setting, in person, preferably in their headquarters. I don’t have a great solution for this. This one comes straight out of an EO though, so is probably the easiest for another administration to mess with, but fortunately, tribal issues seem to actually be depolarizing these days in an interesting way.
The other solution is to remove the threat of litigation on this stuff by amending the APA...but for little projects there’s not THAT much gold plating, because the whole thing is so small, there’s just not that much room for it. But even the smallest actions have a lot of individual moving parts, which have to be run through, even if no one can sue us over them, as congress, or the regulatory agencies have required them.
Any questions?
None, that was extremely thorough and interesting. Thank you for writing it!
"If congress and the president could resist the urge to create a bunch of rules and a bunch of individual exceptions to the rules, things would also be more streamlined, but that’s pretty clearly a pipe dream."
Oh man, if we could just get the politicians to sit down and shut the fudge up, wouldn't it be lovely? Alas, it will never happen!
I still like to talk about the time we got a (long-promised and perfectly usual) funding from central government for housing provision, and *three* local politicians from three *different political parties* all issued press releases to the local papers as to why it was they, them, personally, who got that money released and thank you voters of the county, remember this for the next election as to who gets things done round here.
Need I add *none* of them had anything to do with it, as it was routine disbursement of funds due to the council? (They were the same when I worked in student grants; not these particular politicians, but in general yeah - local councillor to local representative would ring up, ask "Did Joey Murphy get the grant?" and if told 'yes', would then tootle off to Joey's mam to tell her "By the way, your boy is getting his grant to go to college, now no need to thank me (nudge-nudge,wink-wink)".
Democracy in action!
Instead of halving the number of bureaucrats, I propose we halve each and every bureaucrat. This will increase time spent working since they won’t be getting up from their desks and walking around all the time. Guaranteed reduction in number of steps.
In the long run, this will also disincentivize becoming a bureaucrat, and thereby force a more practical and considerate reduction of red tape.
You're halving them along the wrong plane: sagittal would be better than transverse.
The DOGE mission is mostly theatre and can only achieve incremental or even non-material progress. The main benefit of DOGE will simply be awareness in the public's mind on the reality of the size and waste of the administrative state, whether or not we have the ability to meaningfully solve the issue in the relative short term (5-10 years). The reversal of Chevron, DOGE, and the well documented failures and drawbacks of the growth the federal government are all steps in the right direction. I really struggle to understand how the massive growth of government in the last 100 years has been a net positive to American life. The same is true at the state level. Top down awareness could create a level of outrage and information that fuels a bottom up change in behaviors around how we legislate and govern.
So much for pith:
I actually worked at the worst bureaucracy imaginable for 5 years. We did professional licensing for non-medical/non-legal services (barbers, realtors, contractors). My particular job was doing two separate hearings for every applicant who had a criminal conviction, making a recommendation to the board members actually making the decision, and then all the applications were approved. In the case of barbers, every single application was (and should have been) approved. There was no reason at all for this job to exist.
Whenever I hear people talk about bureaucracy, I understand where they're coming from, but they always seem to assume that either the bureaucrats don't know what they're doing is stupid or don't care. Scott Aaronson's conception of "blankfaces" comes to mind.
The truth is that every person involved is acting rationally. We constantly attempted to lower the burden of going through this absurd process. We even skirted the "everyone has to be interviewed" law for minor offenses on the theory that nobody would care/have standing to sue us if we did.
The conservative legislature constantly did things to make that "get around this stupid law of having to interview everyone" process harder. They'd insist we hire non-lawyer staff to run the hearings and oversee the process so they could use non-lawyer pay scales. Those folks were more likely to blindly follow the process. They'd demand we "get rid of half our regulations." They'd do anything other than *DO AWAY WITH THE LAW THAT REQUIRED US TO DO A RIDICULOUS THING* or *STOP LICENSING PEOPLE WHO DON'T NEED TO BE LICENSED!*
Why wouldn't they just repeal this law? Because there's a huge difference between not caring if rapists are barbers and standing up in front of a room full of people and saying "I think rapists should be allowed to be barbers." This communicates a bunch of stuff other than the practical point - similar to communicating on several simulacra levels described by Zvi Mowshowitz and others.
So conservative lawmakers with dual incentives to appear tough on crime and tough on regulation pass nonsense laws that look good but do nothing. The Board for each profession is a citizen board made up of licensed people - they have incentive to make the process hard and even if they're individually conscientious of this bias and careful (which many, but not all, were!) they don't have the legal training to make it easy. The director has incentive to appease his legislative and board member masters by appearing to work hard to cut red tape while making no fundamental change to how quickly licenses get approved. And I, poor sod that I am, have incentive to keep my job and not do or say anything that could be grounds for a lawsuit, *including* explaining to people/acknowledging how absurd the process is.
We ultimately end up in a mirror universe where 200 people perpetuating a crazy system that makes no sense is the only way to preserve their livelihood, their social position, and their integrity.
Once you see this in one place you see it everywhere. As an example, when my child was born my parents told me about a free service the fire department in my area ran that would ensure we had correctly installed her carseat. The service had been discontinued because of liability concerns (if they got it wrong, they could be sued). They directed me to the hospital who told me the same thing. Ultimately nobody in the state (other than friends/family) was willing to give me a quick opinion on whether I was keeping my kid safe because of concerns that they might be held accountable for it. In this case that seems fine (I don't really think I needed a second opinion), but it's emblematic of the bigger issue. Things don't get done out of a general desire to avoid having responsibility fall on any one person. Even the responsibility for avoiding responsibility is diffuse.
The only way to solve this would be to inculcate a sense of courage, not occasionally, in individuals, but society-wide, in institutions. And I have no idea how you'd do that.
Very thoughtful, thanks!
I like those "sunset" provisions, so that dumb rules and regulations can be allowed to just quietly expire and go away. Most bureaucracies tend to accumulate a pile of procedures and rules over time to deal with edge cases and odd situations, and then some of them stick around mostly being useless and ignored until someone sues.
One bummer is that DOGE seems to be moving on from rhetoric around mass firings of head-count to immiseration to get people to quit - IE, move the department across the country, get rid of workplace flexibility, etc. Which is a very effective way to get rid of the most capable people in the department with plenty of other options, while keeping both the folks who stay in the lines but suck at their jobs, and the folks who are so mission-driven about it that they won't quit (note that their idea of the department's mission and Way of Doing Things probably does not coincide with DOGE's).
And those people won't be able to do anything, including enforcement of regulations. Problem solved.
I appreciate the article, but I hope the model that's wrong here is the model that says Trump, Ramaswamy, and Musk are serious about what they say. I hope they're trying to scare people about X when they really want Y. Along the lines of nominating Matt Gaetz for AG and Kennedy to head HHS in order to get their second nominee accepted. Ramaswamy is too smart to imagine this is a real way to cut bureaucracy.
You need to take Parkinson's law into account (https://en.m.wikipedia.org/wiki/Parkinson's_law)
Bureaucracy tends to grow, regardless of the amount of work to be done. Some people in industry had good experiences with some bureaucrats, but that doesn't necessarily mean that all bureaucrats are doing useful things.
From the link you shared: "the number of workers within public administration, bureaucracy or officialdom tends to grow, regardless of the amount of work to be done. This was attributed mainly to two factors: that officials want subordinates, not rivals, and that officials make work for each other."
I would argue it's possible to fight back against, or rein in the two main factors of this law.
The fact that officials want subordinates instead of rivals could be alleviated by creating incentives (monetary or prestige-based) around rivalry, and disincentives around furthering subordination.
The fact that officials make work for each other is even more pernicious, but maybe there's something to be done with a mix of performance reviews, reducing job security and having tasks ("work") be reviewed through an explicit lens of goals that have the highest impact on the people/enterprises/infrastructure served by any given department, thus making a lot of "work for the goal of work" less likely to be distributed.
Actually, the per capita seize of the federal civil service has drastically shrunk since 1945 and been completely stable for the last 15 years. “Analyzing the federal workforce as a percentage of the total U.S. population for the past 15 years reveals the workforce has represented approximately 0.6% of the population. This percentage remained stable as the population and workforce have grown, indicating that the federal workforce has kept pace with population growth. However, it is still a significant decrease from 1945, when the workforce represented a historic 2.5% of the entire population.” https://ourpublicservice.org/fed-figures/a-profile-of-the-2023-federal-workforce/#:~:text=Size%20of%20Federal%20Workforce&text=Table%20with%203%20columns%20and%205%20rows.&text=Analyzing%20the%20federal%20workforce%20as,2.5%25%20of%20the%20entire%20population.
I think auto approval after a cutoff date is fully in the Overton window now. Either the regulatory agencies will figure out how to radically expedite the process, and even lobby for those changes, or the process will be expedited by the cutoffs
Is it just me or are Scott's and Ramaswamy's arguments rather tangential to each other?
I read this tweet as a call to reduce the federal budget by reducing payroll expenditures (this is the most common use of the term "downsizing" in my experience). His other tweets suggest he also wants to reduce red tape, and theoretically it's possible that he believes there's a direct causal relationship between those two things. But on the surface, looking at this article and the tweet contained therein, it seems that the two are thinking about different variables in a large and complicated equation.
I think Scott’s article would remain relevant even if that was the case, in showing how those two separate goals of Ramaswamy often are opposed to each other, and pursuing one might come at the expense of the other.
This is true, but I think that relevance is diminished by the fact that (unless I missed something by reading this while waiting for morning coffee to kick in) he doesn't even acknowledge the main thrust of Ramaswamy's argument, much less mention the trade-off.
It looks to me like Scott is just missing the point entirely; this is incongruous because he's smarter than I am and not prone to bad-faith argumentaton, but I can't reconcile it otherwise.
Yeah I keep reading because Scott is usually thorough enough to at least also discuss Parkinson's law, but here it's not mentioned at all. In his anger it seems he forget to steelman.
Similar to Peltzman's observation, I think there is a design optimization principle where benefits on one frontier, when used optimally, tend to be converted into benefits on multiple frontiers.
For example, if you suddenly gain the ability to make the same sized engine in a car go even faster and be even more powerful for the same price, you might instead convert those capability gains into several different benefits: you might make a smaller engine for cost savings and have more cargo space with the same performance characteristics in the same vehicle.
This same optimization principle works in reverse. If the cost of an engine doubles, or you have to fire half your staff, if you're being intelligent you'll probably absorb the cost across many frontiers.
So I would predict something like an FDA that approves somewhat fewer things in a worse way (where "worse" may open them up to more liability) and that also cuts back quality, patient safety, (A double edged sword as an employee at a medical device manufacturer - we are more careful for the sake of compliance but at great expense), etc.
However, in order to predict the distribution of the capability losses, and how much fat the FDA has to cut, we need to have some level of understanding of what the FDA is doing internally. I honestly don't respect anyone who tries to answer what will happen without trying to understand what is happening internally at the FDA. For example, I have seen metrics saying there was a 70% increase in FDA employees since 2005. Okay, but at the end of the day do they need them or not? That's less than healthcare GDP growth since 2005 IIRC.
I think it would be far more reasonable for the government to pay an independent consultant to map out what is happening at the FDA and proceed from there. Of course this is impossible because of partisanship, but one can dream I guess. But yes, ultimately if you want to cut the red tape, figure out where it is, do that directly (change incentives?), and then fire proportionally to how much you actually overhired, if you did. It's impossible to fix complex systems by treating them like black boxes and then smacking them with a hammer over and over.
I wonder if the crux of the issue is the courts, rather than the bureaucrats. Under common law regulations grow unboundedly, because new precedents impose requirements on regulators as they get sued.
Inquisitorial systems seem intuitively better at adjudicating complex regulatory regimes. They only consider the letter of the law, so they can process cases faster. And because they can’t make binding precedents, they avoid the vicious cycle of regulation -> litigation -> precedent -> new regulation.
It would be interesting to see a comparison between, eg, Louisiana and Mississippi to see if this actually holds.
Common law systems are also inherently cruel and unjust, because they unduly benefit the wealthy and connected.
If I had any influence on government budgets and staffing, there would be four important questions that I'd like to hear every agency answer:
If your funding was increased by 5%, what would you be able to do better?
If your funding was decreased by 5%, what would get worse?
If you only had 10% of your current budget, what would you still be able to do?
If you had ten times the budget that you do now, what new things would you be able to accomplish?
When I was a government contractor, the contracting process was extremely wasteful because there were too /few/ bureaucrats (program managers). Small contracts are much more-efficient than large contracts, at least if measured by number of publications produced (a bad measure, but it's hard to find a good measure that can be applied to all contracts in the same way). Government agencies award mostly big contracts because big contracts are much more-efficient at spending money. If you have to give out $500,000,000, it's a lot easier to spend it on one project than on 10,000 projects. Same reason venture capitalists would rather invest $10 million than $1 million.
I once was the PI on a $100,000 project, reporting to a DARPA program manager who was also managing a $100,000,000 project. He notified me at the start of the project that he would not answer phone calls or emails from me and would not read my project reports; and near the end of the project his office told me not to bother writing a final report, as nobody would read it.
I love this article in general.
But re this: "Vivek . . . seem[s] more qualified to wield chainsaws than to understand legal minutiae."
I am a Vivek stan so I just needed to chime in here and disagree.
Vivek Ramaswamy went to Yale Law School. That's usually ranked as the #1 law school although sometimes I see people calling it #2. I applied and didn't get in. Everyone I know who DID go to YLS seems to me to be a hard-working genius.
To conclude, even if Vivek hasn't practiced law since law school, he is probably qualified to understand legal minutiae.
#Vivek2028
I’m curious about how that “eliminate two regulations for every new regulation made” has worked out for Idaho. Republicans have been trying to push this sort of catchy policy at the federal level and it seems insane to me.
Maybe at the beginning you’ve got some low-hanging fruit - but you don’t want to touch more of that than you have to, because you want to save it to delete when there are new rules you want at some future point. And then there’s a further question about what counts as “one regulation” - can you delete two old regulations and write a single new regulation that includes both of those as well as the new one you are writing?
Why is it sane to think that in a changing world, you generally don’t need to change any rules?
The crypto industry seems to disagree.
Really? I’m not aware of any country that manages food safety, energy security, financial stability, or construction safety with a few stable rules and no regulations that take account of changing technologies.
In fact we had this very same rule at the federal level during the first Trump administration. I guess it didn't leave much of an impression?
I remember it being discussed - I didn’t realize it was actually implemented!
https://en.wikipedia.org/wiki/Executive_Order_13771
Maybe they need a Lord to make the career civil servants and holdover political appointees bend the knee. Or a Musk Ox to trample the ones who don't? I'm thinking the nominative determinism is the easy part.
As a bureaucrat, all I can say is, Oh, You sweet summer child.
I think any organization balloons beyond necessary size. Mid-level managers are often rewarded for growing their teams, and rarely for working more efficiently with fewer people (I managed 20 staff! Hire me for more money!). Most management staff are not incentivized toward efficiency. You would think that money inflow problems would enforce efficiency, but I've never seen this realized in 10 years of government; its always threatened and never happens. There is always more money for whatever program we want to continue. There are many programs and initiatives and committees that aren't particularly evidenced based and never really produce anything. Calls to make bureaucracy more efficient result in longer forms (more social determinants of health questions, more demographic questions for representation, more evaluation questions for program feedback, all justified and in the name of doing better work, but none of it shorter or easier for the client). A sense of safetyism or racial justice can be used to justify just about anything, and no one tolerates objections to those values.
Seriously, I used to work in a unit that delivered free health services to low income folks, like 70% of my work was forms. Making sure that people didn't get too many services, and qualified, and justifying spending on every single person to the federal government with a double receipt trail. I have a lot of questions of how much more good we could do by just dispensing care with fewer qualifiers, so that we remove the huge administrative costs of government programs. Yes, there would be some cheaters, and also a larger chunk would actually go to the public.
Just like the Paperwork Reduction Act should have been named the Bureaucratic Job Preservation Act. And generally increased the length of each form, because now they had to add a page saying "we complied with the PRA".
Yeah, in my current job I am anticipating with much joy *another* layer of bureaucracy laid on, this time allegedly to reduce the form-filling for the service agreement we engage with. I sure hope it reduces the form-filling, but the fact that they went for a whole new set of acronyms and a pilot programme testing out the new form before scaling it up nationally and webinars to inform us about the upcoming training for the implementation etc. doesn't fill me with confidence on that.
Mostly I think the new year will be same as the old year: providing the same information to three different bodies, all tasked with implementing chunks of policy in order to cut costs and bureaucracy (but of course really in practice increasing both). And on top of that I have a whole new set of five letter acronyms to learn off and try and remember which refers to what.
"I have a lot of questions of how much more good we could do by just dispensing care with fewer qualifiers, so that we remove the huge administrative costs of government programs."
The irony there is that all the extra box-ticking is imposed in the name of economy and efficiency and cost-cutting. Some politician gets up in the Dáil (or Congress or Parliament) to question the government of the day on reducing the tax-payers' burden by cutting costs and ensuring public funds are being spent as intended. The Minister (or Secretary or whatever title the relevant administration official holds) replies that naturally under *their* administration, the public finances have never been better and moreover they have increased the number of services available to the hard-working people of the nation.
Then they go off and tell the civil service to draw up a report on public expenditure, which means the civil service has to bother every entity providing public services to dig out and present the information to them on top of their existing work. And then the report shows money flowing out, so now the minister tells them to set up something to reduce expenditure.
And that means you, me, and the other public-facing service providers now have to present double and triple receipt checked individual forms filled out to demonstrate that Mrs Murphy is getting €100 a week as she is entitled to receive and not €101, and if she is getting €101 you are to seek repayment of that extra €1 over the next five years.
"Also, doesn’t Congress pass like one bill per year now?"
Now google "Chevron deference" and see what an incredible mess the Supreme Court has created. In the interest of destroying the "administrative state" regulators are disempowered to use expertise and evidence to decisions. So the paralysis continues and expands, new causes of action are offered to claimants and the regulated, and the judiciary is empowered to make these calls instead. Congress could take action to direct and clarify intent, but Congress is no longer a functioning body.
>I was surprised to see someone with such experience in the pharmaceutical industry say this, because it goes against how I understood the FDA to work.
The tweet is from November 2023, when Ramaswamy was running to be the Republican nominee for President.
He was campaigning as a Trump-like populist, so (like Trump) anything he said had nothing to do with any attempt at presenting policy. Instead, the strategy is to make a statement that emotes something the desired constituency agrees with ("bureaucracy bad >:("), while also dialing up the insanity of the statement to 11, ("let's fire 50% of bureaucrats at random xD") so it goes viral when opponents argue with it, rightly calling it insane.
This is what propelled Trump to massive popularity in 2016 and got him his Republican nomination, but didn't work for Ramaswamy because all the populist voters were already all-in on Trump.
That origin of the Tweet makes any steelman of the position ultimately pointless, even as an intellectual exercise. The Tweet fundamentally doesn't come from an intellectual position of "how to make the government more efficient," it's just emoting and virality-seeking.
This article is interesting in intuiting the principles of governmental bureaucracy, but is also the equivalent of wondering why someone would throw a perfectly good milkshake at Nigel Farage by asking: "was it not a good milkshake?"
I don't think this is actually his position, but maybe the strongest steelman of Vivek's tweet is the accelerationist argument, basically:
'The best way to fix this problem is through careful consideration and dedicated effort to reform the system, but that's been true for more than a century and no one has ever done it. It's never going to happen. My plan is idiotic and will ruin everything and cause a gigantic crisis, but people will be forced to *actually* get off their ass and respond to that crisis, and the reforms they pass to fix the crisis with half as many bureaucrats at their disposal will end up being better than the current system in the end.'
I don't actually think that would work, but it's probably the version to argue against.
Why won't they respond to the crisis by just hiring twice the number of bureaucrats?
Or, if you can block them from doing that, do nothing and blame you for causing and exacerbating the crisis with your refusal to hire the people they say are needed to fix the problem? Their primary incentive is not to *fix* problems, but to avoid being *blamed* for them. And here you are, volunteering to take all the blame.
I mean I think in practical terms the answer is 'because the people I have undue influence over control all three branches of government for the next 2-4 years and I think I can push them in useful directions.'
Like I said, I personally don't think this would actually work, but that's the steelman version I think.
They'll let the crisis play out for two years so they can take control of the legislature in the midterms, and then allow fixes only so long as they can take most of the credit and so win the Presidency in four. So, 2-4 years of abureaucratc paralysis with all the unpleasantness that entails, and then unchecked Democratic control until the GOP somehow rehabilitates its image.
Bad plan. Lose-your-hat bad.
Granting that firing half the bureaucrats would paralyse the red tape, could that paralysis be then used to cut the amount of red tape? Presumably the system would find a new equilibrium that's not "nothing is ever done"
I teach Legislation and Regulation at NYU Law School, and I am assigning this crisp, funny, completely correct essay to my students next spring.
I’m a bureaucrat, and think this is a good analysis. One thing to note is that a lot of lawsuits against Agencies go to other Agencies. There are about 2,000 Administrative Law Judges in federal Agencies that are responsible for a lot of these roadblocks. The Merit System Protection Board, Federal Labor Relations Authority, and Office of Federal Operations in the EEOC, are specifically the reasons it’s nearly impossible to fire bad performers in the Federal Government, and the President gets to appoint members to them. Overhauling those 3 would greatly improve the government’s efficiency in my biased opinion.
How often are the bureaucrats themselves responsible for creating the red tape? Obviously Congress doesn't pass every single regulation themselves, that's the whole point of the bureaucracies. Firing bureaucrats doesn't reduce the number of existing regulations, but it could be necessary to slow the rate at which new ones are added in order to achieve any net reduction.
> Two years ago, after consulting with industry, the FDA finally banned 23 phthalates but said that the other five were okay, releasing a 58 page decision explaining its decision. Two days ago, the environmental groups sued, saying the remaining 5 phthalates are still bad.
If they're going to get sued anyway, it seems like the report is kind of pointless? Or is just supposed to be more efficient to do all this pre-emptively rather than waiting for the lawsuit? Seems like now they're just doing all the legwork twice. It sounds like the real problem is the fact that anyone can sue anyone else for anything (which, as far as I'm aware, is not something most other countries allow). So rather than just having a top-down regulatory process, or a dispute-resolution process worked out in courts, we have both, stacked on top of each other (and the legal system can take forever to get anything done too... see e.g. https://en.wikipedia.org/wiki/Google_LLC_v._Oracle_America,_Inc.).
Presumably they will keep being sued until they produce a decision (and supporting materials) so ironclad that the lawsuit fails. At least some parties will be negatively affected by the decision and will have standing to sue, so only a decision capable of surviving challenge can actually be final and implemented.
> Halving the number of FDA bureaucrats wouldn’t have literally zero effect on this balance. It would mean that approving new drugs would be delayed twice as long. This would be a little more outrageous than the current delay, and might shift an outrage-minimizing FDA director slightly in the direction of cutting rules. But solve for the equilibrium: there would still be more delay than there is now.
Doesn't this imply that the best way to decrease approval times is to alternate between sharply cutting --> increasing headcount? The underlying model for time-to-approval is something like:
$approval_time = approval_work_hours / num_bureaucrats$
You can make this value go down in two ways:
1. Decreasing the number of total work hours required to gain approval
2. Increasing the number of bureaucrats doing the work
You're saying by sharpy decreasing (2) it will force an increase in (1). But you can go back and increase (2) to where it was, and you'll be in a better place than where you started. Rinse, and repeat.
Can someone please tweet a link to this article in a way that will make Ramaswamy read it?
A particularly well documented example of the paperwork processing issue is wait times for the ATF to process NFA forms. Up until last year a 9 month wait for approval was common but after additional clerks were hired approval times dropped to a matter of days.
This of course begs the implicit question that if the vast majority of applications are approved does the manual form need to exist at all? Implementing a self service registry could probably be done without legislative action, but the ATF would need an actual incentive and institutional will to do so.
the good conscientious civil servant who killed or at least tamed her haidtian elephant
"I don’t have a good sense for how well this could work at the federal level."
IIRC the Macron government adopted similar rules, so it's not completely unimaginable at a national level.
Aren't you assuming that all bureaucrats are equally competent and hard-working? If a significant number of them are 10-3 long lunch break types, firing half of them would just require the rest to, like, work a full day.
I sometimes think we should have an automatic sunset period for all laws (or maybe just regulations), combined with a rule that you can't approve or reapprove a law or regulation with a simple majority vote unless the person sponsoring it reads it out loud to the people who want to vote for it.
If it worked, then I imagine the written rules would become a lot shorter, which is probably a win. But you'd need to prohibit rules which include the text of non-rules by reference.
This would presumably shift power to executives and judges, since shorter rules tend to be more open to interpretation.
I know something about the Idaho example. They reduced regulations by restricting state regulations and instead relying on federal regulations. Most states add regulations beyond what the feds require, but Idaho no longer does. And it’s not a disaster for their citizens, because the federal regulations protect them from most bad outcomes.
This is part of it and another example of why using Idaho (or any other state) is not particularly applicable to the federal government. To continue with the FDA theme of this post:
"The U. S. Food and Drug Administration (FDA) publishes the Food Code, a model that assists food control jurisdictions at all levels of government by providing them with a scientifically sound technical and legal basis for regulating the retail and food service segment of the industry (restaurants and grocery stores and institutions such as nursing homes). Local, state, tribal, and federal regulators use the FDA Food Code as a model to develop or update their own food safety rules and to be consistent with national food regulatory policy."
Idaho, like many states, just adopts a version of the Food Code for their jurisdiction. If FDA didn't produce the Food Code, then Idaho would have nothing to copy.
> How much will they get yelled at if they take too long to approve drugs, vs. if they mistakenly approve a bad drug?
An alternative way of framing this would be, "How many people will die if they take too long to approve drugs, vs. if they mistakenly approve a bad drug ?". This is a complex question; for example, in the extreme case where we abolish the FDA and cease all drug regulation, maybe pharma companies will remain conscientious enough to do at least some minimal drug testing, so on average more lives could be saved; however, upon seeing a mass uptick in drug-related deaths, consumers will conclude that drugs are too dangerous now and stop taking them anyway, thus driving up the number of deaths. And maybe we'd get a situation like in early 1990s..2020s Russia, where snake oil vendors dominated the market.
I mean I think it's important here that those *aren't* the same question, and that this can shift the balance from where it ought to be (many people would say it has).
The whole post can be boiled down to: bureaucracy size might not be the only metric for you to look at if you want to improve the overall efficiency or throughput of the system.
There are tons of other levers for you to consider here. Complexity of requirements. Number of hand-offs. Outdated processes taking days instead of minutes due to lack of automation. Staff motivation levels. Staff incentive structures. Bullshit jobs that should have never been created to begin with. Org structures that have not scaled with the needs of the organization. Etc etc.
The size of the bureaucracy is certainly *a* metric to look at, and organizational size is generally correlated with sluggishness, but it would be silly to reduce the whole audit to just this one metric. I'm confident Vivek knows that well, but perhaps the nuance of improving organizational efficiency doesn't fit nicely into a bumper sticker unlike "fire half of everybody".
This was what I was thinking about. If bureaucracy isn't measured in bureaucrats, then what unit is it measured in? That sounds like a question that is actually worth talking about. Obviously it's a political question, but I wonder if there are approaches that could attract reasonably broad acceptance?
Like, The Economist has something like an ease of doing business survey that it conducts annually, I think. I wonder if that methodology could be applied in some way? In industries with large companies that have actual compliance departments, you could look at the budget for compliance as a fraction of turnover, I guess? Another possibility would be to measure corruption in some way. Excessive bureaucracy preventing lucrative business is a recipe for corruption, so if you could quantify that, it might give you a snapshot of where you should be focusing your streamlining efforts.
As someone with experience in defense contracting, I think your overall thesis here is definitely sound. If anything, our government customers are understaffed, because they are overloaded with low-value added but legally required bureaucratic cruft work. Firing half of them will make the problem worse.
For example, we have projects running late and over budget not because we can’t execute them, but because (in what should be an entirely expected occurrence) development is hard and the requirements and circumstances change, but the bureaucracy can’t respond in any kind of agile way. The changes are not a big deal from a technical perspective, except that these contracts are firm fixed price (an outcome of the last round of government efficiency do-gooding) so any tiny change that impacts cost at all requires a contract update. Which incurs a massive amount of bureaucratic overhead and schedule hit to get all the paperwork filed, and our government contracting officers don’t have the bandwidth to keep up.
Literally, on a nominally 24 month contract, an update that changes the bottom line price by 1% has taken 4 months to actually get on contract. And in the mean time the contractor either has to work at risk of never getting paid (which we are increasingly less willing to do because the federal bureaucrats have been told to stop offering us any accommodations in return) or sit on their hands while the delivery slips to the right.
Now, there are definitely some bureaucrats that are worse than others - an annoying stickler in the wrong spot can bring hundreds of people’s productivity to a screeching halt - but you won’t get rid of that roadblock by randomly giving the agency a haircut.
This post seems to be missing the fact that many, perhaps most, bureaucrats that work in bureaucracies do not do the work of processing forms or dealing with the public. Many are middle managers, HR functionaries, and all the hundreds of different make work type positions that don’t really need to exist. You couldn’t just randomly cut half and get just these positions, but cutting half wouldn’t slow things down by 50% either. It might even speed it up, depending on how many stupid meetings could be avoided.
Regarding crypto,
> But the way that the SEC actually makes rules is: [ ... long complicated process ... ]
> The incentives, for the SEC, are bad. If it makes new rules, that is a lot of work, and those rules will be attacked from every angle. Probably they will be struck down, in ways that limit the SEC’s power and create greater confusion about what is allowed.
> Meanwhile if the SEC just sees some crypto project that it doesn’t like, it can sue that project — probably in a court of its choosing — say “this violates longstanding securities law,” and have a decent chance (not a certainty!) of winning. Its chances are better not only because it can pick a more sympathetic court, but also because it can pick a less sympathetic antagonist: It can argue “we need the power to regulate crypto” in a case where investors lost everything and the value of regulation is clear, rather than writing general rules and getting sued by a nice upstanding firm that doesn’t like them.
https://archive.is/aSBWi
This sounds similar to what you're saying but a big difference between your understanding of bureaucracy and Matt Levine's (at least with regards to crypto) is you think the bureaucrat's goal is to make clear bright-line rules so everyone knows the game, and Levine's view is that bureaucrats are playing political games to support unofficial goals.
Oh, and on your section II, it’s important to note that not only are all agencies not the same on the “our job is to approve things” vs. “our job is to deny things” spectrum, often multiple agencies with overlapping domains are in conflict with each other!
So, for SpaceX: on the one hand, they’ve got some environmental bureaucrats to deal with in California that are absolutely BANANAs (build absolutely nothing anywhere near anything). On the other, they’ve got customers in the Space Force that would really like them to get this expensive critical National security asset into the sky as soon as possible, thank you very much. Who are also pissed off that the CA BANANAs are stepping on their turf (literally, in the case of their Space Force launch sites).
So you need the Space Force / DoD bureaucrats to work on your behalf and make the BANANAs shut up. If you fire half the DoD and half the environmental bureaucrats, the balance of power doesn’t change and nothing speeds up. If you fire half the DoD and the BANANAs stay because they are state of CA employees, well now you’ve played yourself and things will take even longer.
Bureaucracies and law share an important quality in common: They calcify knowledge. This comment isn't intended to be a precise description of something, so much as to convey a concept.
When this works the way we want it to, after a fertilizer plant explodes a rule is created which prevents future fertilizer plants from exploding for the reason that this fertilizer plant exploded. Say, all fertilizer plants have to have a dedicated ammonia processing unit that follows these standards and is constructed out of stainless steel and is maintained and checked every month.
When this works the way we don't want it to, the rule prevents a new process from being implemented which is more efficient and/or better and/or has an overall lower risk profile because the rule requires the old process to be in place. So the new fertilizer plant which doesn't use ammonia still has to have the dedicated ammonia processing unit, even though it doesn't use it, so nobody builds it because that extra expense makes it unprofitable.
And in the third situation, the rule doesn't matter - the new fertilizer plants don't produce fertilizer, but rather a precursor to fertilizer using pure nitrogen that weighs less and the new plants don't quite meet the definition of a "fertilizer plant" as defined in the original rule.
Anti-regulatory and anti-bureaucracy efforts tend to focus on the second situation. Pro-regulatory and pro-bureaucracy efforts tend to focus on the first situation. Meanwhile the third situation keeps being the reality of libertarian progress.
The older a rule is, the more likely it is to be in group 2, until at some point it joins group 3. The newer a rule is, the more likely it is to be in group 1.
The older a bureaucracy is, the more it will tend towards group 2 and group 3 rules, both because the rules for creating rules slow down the creation of new rules, and also because older bureaucracies will have more rules in the first place. The faster technology changes, the less time this takes.
At some point in the rate of change in technology and the average age of bureaucracies, it becomes nearly impossible to create group 1 rules; by the time the rule has actually been calcified, it is obsolete. And at some other point, it becomes nearly impossible to create group 2 rules. I think we're currently stuck in the pit of "It's very difficult to create new group 1 rules".
In the ideal situation, we only have group 1 rules; that's basically what we've been trying to do for the past century but it hasn't worked out great for us. In the second-best scenario, maybe firing half the bureaucrats might slow things down enough that we're left only with group 3 rules going forward.
And the only way you can achieve the ideal scenario is to burn down the entire rot-infested system and start anew.
Lol. First of all, Vivek is a very smart guy but he spent about 7 years running his company so saying he's an experienced pharma guy is a bit of a stretch. Seven years is just about enough time in pharma to realize you don't know very much.
Second, I don't think he's talking about the FDA specifically.
My experience with the FDA is that they're very very busy. Most reviewers are swamped, and they don't have a lot of extra time on their hands. On the contrary. I remember showing up at an FDA Advisory Comm Meeting with a phalanx of 30+ support personnel (several people to run the powerpoint slides, several statisticians, people to pull up data real time, etc.) after practicing 8 hours a days for 6 months for our presentation, and the FDA clinical reviewer showing up with no support people and a thumb drive - we had to load her presentation on our laptop for her because she didn't have a laptop. In general they do a good job.
Besides, FDA is funded 50% or sometimes more by User Fees industry pays, so I'm not sure they can save money by firing the reviewers. Where the fees go is negotiated between the FDA and industry.
Now, one area where there might be savings is the scope of FDA's remit. They have extended jurisdiction into a lot of areas that may be beyond what Congress intended, and if half of their headcount disappears, those would be the areas they might cut back on. These are things like regulation of compounding pharmacies, regulation of off-label promotions (they lost a huge case,Amarin Pharma, Inc. v. FDA, on Freedom of Speech grounds, and they didn't even appeal it, probably because they didn't want to lose and make it binding across the whole country), non-interstate commerce, importation of drugs (if you go to another country and you get treated, then FDA can claim that you are importing the drug… because the drug is in your body, see?), lab-developed tests, right-to-try (experimental therapies) etc.
One addendum about the compounding pharmacies. It used to be all drug was compounded by pharmacists (and physicians). FDA was created partly because when manufacturers started selling drugs, the pharmacists complained, "hey, these guys who aren't even always pharmacists are making drug and just selling them."
in other words FDA's original job was partly to protect pharmacists from inferior and unfair competition. If you weren't a pharmacist, you had to show that your product was not terrible.
Now, ironically, the FDA is trying to shut down compounding pharmacies.
The user fees are an interesting example because FDA only started using them in the early 1990s. My understanding is the pharmaceutical industry and others actually worked with Congress to authorize FDA to collect user fees. They wanted to help fund FDA because they wanted FDA to hire more staff and speed up the review process.
Scott- you’re misunderstanding these two nitwits by engaging with them in good faith. They don’t want “efficiency,” they want to cripple the administrative state. Is it any wonder that Elon Musk, who is under very active and very public prosecution for all manner of securities fraud, thinks it would be great if the SEC had half as many employees?
This whole project is “make the drug applications take two years instead of one, so we can then justify getting rid of the FDA,” not “we’re just these nice well-meaning bros who want to help make the trains run on time.”
This may ultimately be true but doesn't undercut the need for good faith engagement, which isn't directed to these two people personally (are they SSC readers?) but towards public opinion.
Part of proving they're operating in bad faith is proving their solutions, which much of the public takes seriously, couldn't possibly work
Respectfully, this is the failure mode of Scott-style “rationalist” thinking and I see it happen all the time. They’ll keep using arguments as soldiers and throwing them at you until your forces are depleted. While you’re busy trying to calmly and rationally rebut points 7,8, and 9, they’ve already made up points 10, 11, and 12.
I don't see a connection between this claim and what you're responding to.
I think also "assuming people one doesn't like don't actually believe the things that they advocate for" is a take that's been heavily influenced by your political position... but undoubtedly, my having the exact opposite impression re: Musk (don't know much about Ramawhatever) is influenced by my politics too.
I'd be fine if the outcome was "completely get rid of the FDA", though. If that's the ultimate plan, more power to 'em!
People dramatically underestimate the extent to which Congress and the courts are responsible for the state of our government institutions. It's of course understandable why some elected officials want to deflect blame to government institutions. Sometimes, they do mess up. But it's pretty lazy to ascribe "laziness" as a monocausal explanation for how we ended up here.
Let's take the current wildfires in LA as an example. Everyone who works in wildland fire management in government knew something like this would happen somewhere due to continual encroachments in the wildland-urban interface. Wildland management officials also understand that clearing out brush (either mechanically or with prescribed fire is helpful). To emphasize this point, people in the field literally refer to prescribed fire as "beneficial fire." They know how to mitigate some of these severe outcomes. It's worth spending some time pondering why there is a gap between what bureaucrats "know" they need to do versus what actually gets done.
As one example, to continue the theme of this post, paperwork exists. Burns on federal lands have to comply with environmental permitting laws or other things like the Clean Air Act (go read about the interaction between the exceptional events rule and prescribed fire). The bureaucracy is not magic; it cannot ignore laws just because they're insane or result in long timelines or a lot of paperwork. The US Forest Service spends a massive chunk of its national budget (~40%) just on the paperwork to do stuff.
This is stupid, but again, it is also not the fault of bureaucrats. We need to either remove the threat of lawsuit or for Congress to pass new laws and fix bugs.
The bureaucrats also know all this stuff is insane! On wildfires specifically, they've asked Congress for comprehensive fixes to improve the way we manage wildfires. Has Congress done anything? No, of course not. This means we get to continue waiting 4+ years for wildland fire managers to complete environmental impact statements.
Here is a comprehensive report the government spent a lot of time on with some reasonable fixes to address all of the things that people are suddenly now armchair experts in:
https://www.usda.gov/sites/default/files/documents/wfmmc-final-report-09-2023.pdf
Yes, and down below you'll find a perfect example of "armchair expert" ranting on the wildfire example from someone labeled Darkside007.
I feel like Scott's model is not capturing an important element of the FDA (or other) bureaucratic process in step 1 of his model. Specifically, when regulators are not totally constrained by statute, and can ask for process or application changes based on their discretion, more regulators or more time per regulator inspecting means more work for compliance.
Said another way, I think Scott's model of drug approval should be reworked from “drug approval is processing 1,000 forms” to "the number of 'forms' required for approving/getting a drug to market is modeled as 1000 + f(bureaucrats)."
This model reflects the fact that there are many statutory requirements the FDA must fulfill at each step of the process (the 1000 forms). It also recognizes that additional bureaucrats add 'forms' or requirements at their own discretion.
I would offer the following as an example of where discretion in enforcement is both necessary and genuinely increases the amount of compliance work/difficulty in getting a pharmaceutical made in proportion to the number of additional bureaucrats.
The FDA inspects pharmaceutical production facilities for compliance with good manufacturing practice (GMP). Strict GMP enforcement ensures only high quality pharmaceutical products are produced, there are no adulterated products/byproducts/etc. During this process, the FDA inspectors must have some discretion to flag process deficiencies that are not exactly anticipated by the guidelines.
However, because the inspectors have discretion, more eyes means more flagged items. I want to share a good example of this from my work. Pharmaceutical manufacturing almost always includes a chromatography step -- the purification of the API/molecule of interest on a column packed with a specific resin. Over time, the column resin discolors due to impurities that are being removed from the pure API (this is the purpose of the process after all). In research settings, the columns are clear glass or plastic because you want to monitor the fouling/color of the column.
The manufacturer of the columns we used (glass covered) told us that they offered stainless steel housed columns so that when we get to GMP production, we can switch over. We asked them why we'd want that? We were perfectly happy with the glass system, it met pressure specs etc. They responded that FDA inspectors would always flag a slightly discolored resin as a hazard (e.g. a contaminant or potential for bacterial/fungal growth). In reality, the resin being somewhat discolored is totally normal and doesn't effect the quality of the chromatography or separation (you have checks for this). But, because FDA inspectors would see that, not like it, and require a process change or an expensive swap, companies responded by enclosing the columns in stainless steel and just hiding it from inspector eyes.
Now, imagine that this is happening across an extremely complex pharma production pipeline with many inspected processes, machines, inputs, etc. This is a situation where 10 inspectors in the building are going to find 100 "problems", and 50 inspectors are going to find 500 "problems" with your process. Note that those 50 inspectors are also more likely to find real deficiencies or catch problems you don't see.
The IND and drug approval processes share many similar features to what I've described here. Because they are not totally statutorily constrained, and they have essentially no political cost to finding fault/deficiency, the more bureaucrats you have, the more compliance items you will have to resolve.
Ultimately, I think Scott's model needs to adjust for this fact. Clearly cutting bureaucrats will reduce the throughput on the statutorily required and generally unavoidable forms. However, it may also substantially reduce the burden the companies face from discretionary enforcement.
Many people that function as federal bureaucrats aren't even employed by the federal government. Many roles in HR and Compliance departments, most especially in extremely regulated industries like banking, are effectively monitoring, enforcing, and recommending new federal regulations as their full time job.
The problem is the scope of the federal government, not strictly the size. Fix the scope and everything else falls into place.
Admittedly dumb, I smell barbarians at the gates again. They won't have much time to enjoy their gains this time, though, if they gain anything at all. The nuclear power that's most efficiently organized will probably call the shots after some trouble.
Russia and US combined have 90% of the world's nukes, though. If they joined forces (which seems increasingly likely), they would easily be able to reduce the entire rest of the world to ash if they wanted to.
Thanks for your thoughts. I consider Russia a shitshow and would rather expect Putin to sell to China than to the US, but I may well be wrong.
They're not selling. This is just basic game theory. If World War III was a complete free for all, there's a much lower chance that either the US or Russia comes out on top. But if they work together to deal with the rest of the countries first, now both of them have a much higher chance of winning than before, since now there's only two possible victors. And they wouldn't have to worry about nukes until later, since they basically have a monopoly on them.
A quick google of “afghanistan translators us citizenship” tells me you are wrong. But I’ll be honest, I knew that from the tone.
I understand that many people such as yourself have absolutely no inkling of the difference between a United States 'green card' and U.S. citizenship but, you know, you really should try harder. FACT is that both 'quick' and belated google searches show no such thing as you claim.
It doesn't pay to be ignorant, really it does not. Learning is Good.
So the difference is largely a difference of a few years, since the green card gives people the right to petition for citizenship after 5 years, and so it’s a matter of waiting. Seems like this is a debate ok semantics.
John 11:35 KJV.
As a green-carded naturalised U.S. citizen, I can personally assure you that the wait, the paperwork and the (meanwhile) duty of serving the USA IN a war isn't semantical at all. It's a teensy weentsy bit more dangerous than your classroom semantics.
But I've got a DD214. Do you?
I have no idea what a DD214 is. And don’t care.
ALSO I don't think this is the place for actual debate or flaming, but if you feel any urge to continue, please take it to my actual real email: dawson55@gmail.com.
Cheers
Banned for this comment.
I highly recommend Michael Lewis’s book *The Premonition* to get a heartening glimpse of how quickly it is possible to address a knotty problem outside the bureaucracy. It’s about a few scattered professionals who took it upon themselves to figure out the covid situation & smart, prompt ways to address it. Most tried but failed to implement their ideas & inventions. Some of what they came up with turned out to be flawed, but you get the strong impression that these people would not have pushed back at all against good evidence of flaws, but instead thrown themselves into developing an improved plan.
For example, one guy was an expert on flow structure within fractures and soil. He helped his high school daughter do a science project on contagion, using approaches he had developed to model collisions of particles. They ended up concluding that schools were the setting with by far the most close & prolonged contact between people, and based on that thought that closing the schools would be the best step to slow the spread of the virus. It later turned out that children were not powerful vectors, though. I forget the details of the reason. It had something to do with ways kids respiratory tract is different from adults. So what we really needed to identify was the settings in which there was the most contact between adults.
I don’t know what we would have ended up doing if you could have gathered these 10 or so people, funded them and let them run the show, but I’m pretty sure we would have navigated covid far better than we did. Oh, and we should have added a good, honest communicator to the batch to be spokesman (maybe Scott, if he could be a bit folksier. Though he might demand to be spared TV appearances).
It wouldn’t surprise me to hear that some of the people Lewis’s book features are not as admirable as they sound in the book, but even if you shrink all of these people’s smarts and integrity by 50% I think they still would have made a decent team. Or if one is in fact a genuinely bad apple, they could be replaced. You don’t run across people like the ones Lewis talks about every day, but they aren’t as rare as hen’s teeth. I know a couple. You probably do too.
And of course I don’t know how to set up the government so that groups like that are the ones who take care of business.
Regarding crypto regulation, my sense is that actually the rules are quite clear, it's just that crypto companies don't like those rules and would prefer to be subject to different, laxer rules. But just saying that would sound bad, so instead they falsely claim that there is no regulation or that the regulations are ambiguous.
> They did it through the power of nominative determinism. In that year, they elected a governor named Brad Little. His administration is called the Little Administration.
For those who showed interest in English prosody on an earlier open thread, I'll note that the government-diminishing phrase "little administration" has primary stress on the penultimate syllable of "administration", whereas the government-assigning phrase "Little administration" has primary stress on the penultimate syllable of "Little".
Neat!
Can you confirm or deny the claim that stress placement in English is nearly random? I read this in some book (that there's no handy rules that work nearly all the time, as in e.g. Spanish), but I'm too lazy to think about it.
There are two levels of stress you'd want to consider.
Every word has its own stress which is just part of the dictionary entry. (Or, in other words, word-level stress is an arbitrary fact and the only approach you can take to it is memorization.) Your comparison to Spanish makes me feel you're probably thinking about this kind of stress. At this level, your claim is correct.
Stress (and other prosodic phenomena) are also applied at a higher level, to highlight the grammatical structure of a clause or to focus emphasis on a certain point. This is what I'm talking about in the comment above. This kind of stress is governed by rules. It might be hard to find good literature on those rules, though.
At this point, I'm fine if they just close down the fda entirely.
Close them utterly and then let them go before congress with a rationale for why they do what they do better than the private sector could do it. Let them beg to be allowed to continue their "work of protecting American citizens". Let them explain why so many of their ex employees work for big pharmaceutical companies. Let them justify why they have surpassed information.
And no more lies about being objective and not under the pay/thumb of pharma. We know it isn't true.
Let them squirm.
And some shady pharma operator, probably not one of the big ones but a marginal one trying to make it big, will start selling a drug that, oops, causes ten thousand horribly malformed flipper-babies to be born, and the public will overwhelmingly vote for whichever political party most convincingly promises to stand up the UberFDA, just like the old one but a thousand percent moreso.
Do you have a more serious proposal to offer?
You say that speeding things up requires either preventing lawsuits or changing laws, but of course, one can also just *change the regulations* rather than go through Congress. This probably requires a change of leadership at the regulatory agency, to someone who actually wants to make those changes, but an incoming administration obviously has the power to do that.
Of course, it remains true, as you say, that you don't do this by randomly firing people! If you want to change the regulations, you need spare people to make that change; if you want to automate things, as Charlotte Dune discusses in her comment, you need to hire people to perform that automation (and then afterward you can lay off a bunch of people, but the people doing the automation will probably need to be new hires).
By the way, you actually *can* just reduce what you can sue (say) the EPA for. The US government has sovereign immunity; you can only sue it when it's made a law saying you can sue it. I don't really approve of this principle, but given that it exists, you absolutely can just find what law lets people sue the EPA and cut down what it allows suing over. (Although as other commenters have pointed out, the recent overturning of Chevron deference sure makes a real mess of things. Not that those are directly related, but they're on the same general point...)
See also Matt A's comment about how much bureaucratic process was put in place to remove bureaucrats' decision-making power; people will have to be OK with that! Also see the comments people made about the bloat coming from agencies' *contractors*, rather than the agencies themselves -- in some cases, reducing bloat and improving efficiency may require *growing* the agency so the *contractors* can be fired (Alon Levy talks about this all the time wrt the MTA and other American train agencies).
Let me tell you about John Nestor.
In the 70s, the government did a traffic analysis to figure out why traffic was coming to a standstill at roughly 16:30 every day. What they found was that 1 guy would get on the highway, merge into the passing lane, set his cruise control to 55mph, and jam up traffic for the next 4 hours. This made local news.
This man was John Nestor. We know this, because, in a massive display of entitlement, he wrote a letter to the editor in the Washington Post identifying himself as the man in question, and justifying his actions by, in brief, asking why he should bother speeding or dealing with merges and the traffic on the right for the benefit of other people?
Now, why does he matter? Because he worked for the FDA. In drug approvals. He applied that same ethic to the drug approvals he oversaw - why should he put himself at risk for benefit of a billion-dollar multinational drug company? The number of approvals he granted over his career was a nice round number. In fact, it was the *roundest* number, 0. If your drug ended up on Nestor's desk, you've lost your entire investment and there's nothing you can do about it.
Despite actively sabotaging the approval process, Nestor couldn't be fired. Eventually, he was shipped out to Iowa and removed from the approval process, so at least there was some improvement to the process and the traffic.
Did I get you? Of course this story doesn't end on a positive note. Ralph Nader launched a campaign to restore Nestor to his position and eventually won, because success must be punished as often and as hard as possible, and failure must be rewarded.
Cutting the bureaucracy in half would remove a lot of Nestors and other bureaucrats on that scale - maybe not everyone will kill everything, but only allowing no side-effect, perfectly safe, no-risk products can also do immense harm to the public. Getting rid of half of the bureaucrats will eliminate half the School of Nestor.
It will also expose the rest. Since fewer people are getting more work, the harmful people will be more obvious, and addressed directly. Eventually, there will not be 1000 bureaucrats processing 10 forms a year, but 50 bureaucrats processing 200.
Or, you know, reducing the number of forms needed for anything to begin with. 1000 bureaucrats with 10 forms apiece for the year can drag out each form and crush the private sector by design or indifferent malice. Musk complains that designing and building a rocket takes less time than getting the approval to launch from 1 desk to another. That is possible because there are too many bureaucrats, so they have time to stall out on processes and approvals.
Most salient example: Bureaucrats demand environmental impact studies to clear out deadwood and underbrush for fire prevention. This process adds several *years* to the process, making the process longer than the fire cycle. So many agencies don't bother trying, because it's just a waste of money that will never actually get to clear out the deadwood. And now LA is burning down, and if they had just let a couple guys with heavy equipment go in and work continually, the fire wouldn't've happened.
The fires in LA are not a Climate Change tragedy or the result of extensive arson. The only factor that mattered was the deadwood, and it was not cleared because the bureaucracy said "Wait. (And by wait we mean no)"
Ah, see, the issue was they reassigned them to Iowa. Iowa ain't DC, but it ain't bad.
That said, I hear Nome is positively lovely this time of year (as is Greenland, but one thing at a time)
He is the undertaker in the "invisible graveyard" in which Jake Seliger lies
I think your story must have grown in the telling before it got to you. The 1984 Washington Post article I found mentions nothing about a government traffic analysis: https://www.washingtonpost.com/archive/lifestyle/1984/11/21/john-nestor-strife-in-the-fast-lane/177dbb31-aeed-499e-8be4-9de519efd37a/
> A woman had written a letter to this newspaper complaining of being tailgated on the Beltway by a large truck whose speed-hungry driver wanted her out of the way. Truck lovers fired back, accusing her of wanton lane-hoggery, and a prolonged epistolary squabble ensued. Still, the whole affair might have expired of inanity. But on Oct. 1, in a quiet corner of the letters section, appeared Nestor's modest manifesto:
"On divided highways I drive in the left lane with my cruise control set at the speed limit of 55 miles per hour because it is usually the smoothest lane. I avoid slower traffic coming in and out from the right, and I avoid resetting the cruise control with every lane change.
Why should I inconvenience myself for someone who wants to speed?"
The public exploded in a partisan frenzy. Nestor's name ("It's as Irish as Paddy's pig -- comes from County Clare") entered the local lexicon: His supporters called themselves "Nestorians," and an editorial minted the term "Nestoring." But for the reluctant eponym, the celebrity rankled. "After all," he protests, "I have really accomplished some important things." Chiefly during his 21 years as a medical officer in the Food and Drug Administration, where he developed a reputation for whistleblowing that makes the Mormon Tabernacle organ sound like a kazoo. Dr. Sidney Wolfe of the Naderite Health Research Group calls him "sort of the ideal public servant."
(Quotation marks around Nestor's manifesto added for clarity.)
I also don't think it's the case that he didn't approve any drugs in his career - just that he didn't approve any between the years of 1968 to 1972.
This doesn't change the takeaway, though: Nestor's still also the ideal example of a short-sighted, selfish, and obstructive civil servant.
I'm certainly no partisan of the FDA or of John Nestor.
As for the takeaway of the parent comment, I think it hinges on this claim: "Cutting the bureaucracy in half would remove a lot of Nestors and other bureaucrats on that scale - maybe not everyone will kill everything, but only allowing no side-effect, perfectly safe, no-risk products can also do immense harm to the public. Getting rid of half of the bureaucrats will eliminate half the School of Nestor. It will also expose the rest. Since fewer people are getting more work, the harmful people will be more obvious, and addressed directly."
Getting rid of half of the Nestors obviously does no good if you're also rid of your competent employees - the ratio hasn't changed at all - so you have to believe that firing half of everyone will make the system more transparent to management and make management more inclined and able to remove Nestors. Since I don't see any reason to expect this, I'm going to side with Scott Alexander.
That's a good point. I misremembered the takeaway as "we hate this sort of thing" rather than "yeah let's fire half the people on a random basis", heh.
"so you have to believe that firing half of everyone will make the system more transparent to management and make management more inclined and able to remove Nestors."
From my OP:
"It will also expose the rest. Since fewer people are getting more work, the harmful people will be more obvious, and addressed directly. Eventually, there will not be 1000 bureaucrats processing 10 forms a year, but 50 bureaucrats processing 200."
One person blowing up the drug-approval process can get Congressional hearings and might actually be an election issue, like Fauci. 1000 people blowing up some of the drug-approval will not.
The core ideas behind this are:
- Red tape amount is related to the number of bureaucrats: If there are far less, this incentivizes moving away from 800 page documents. There are certainly various cases where these are required by law/regulation, but I would not be surprised if there are many that are not actually needed.
- Existing bureaucracies are inefficient. Just because someone said they had a good time with those people doesn't mean it wasn't inefficient! They're just decently sociable. (You can enjoy, for example, talking to a college advisor even though the college is gouging you and the advisor isn't spending more than twenty minutes on you)
- This is an empirical question of how much work is actually done in various parts of government. How much slacking off occurs?
- This will pressure other parts of the government and companies to downsize on redtape. They can focus on the departments that don't keep up, and hopefully actually talk about reducing X,Y,Z requirements.
- It would be better to select who stays based on competency, but I think part of the idea is to avoid goodharting of metrics (and just straight up fraud/exaggeration that ingroups which have existed for decades can have)
- Internal delays. If you have to talk with five different people about five different things, that is going to add delays in how things are handled. If, instead, those five different things are handled by two people? Less delay. Presuming enough slack, of course.
(One issue with some examples of good bureaucrats is they're often top 1%, but this is like saying that because there's top programmers at twitter then cutting half of all programmers won't help any. It does increase workload, but it also avoids a massive amount of internal inertia. Government has more issues due to required inertia via regulations/compliance/forms, however.)
I think the idea behind firing the bureaucrats is to put pressure on Congress to reduce the FDA's mandates. There's a reason why the FDA has to many things to do -- because of pressure from special interests. Well, if the special interests can't get anything done anymore, then they're going to have to give up on some of those things.
Agree with the article in general, but to me raises the question: What would it take to reduce the risk of lawsuits more generally? How to make the country less litigious?
I don't think it's likely to happen. In addition to individual motivations - the 'right to one's day in court' which, though originally from Britain, is stronger in the US than anywhere else - there are collective efforts - activism; green tape; 'lawfare'; lobbying; ever-expanding rights...all building into a massive law/academia/media-admin landscape. And huge numbers of people, at different times, and often with good motives, contribute to the general gridlock.
I think being litigous is actually good. In non-litigous country, seems like problems are still there but solved in extralegal ways. It may be good if the problems are better solved that way, but most of the time it's because the legal officers can't be trusted. A litigous country means that people at least trust the legal system, and you don't really want to reduce that trust.
Loser Pays would be a very good start. Yes, the poor would still be able to sue the rich, presuming the suits had merit; there's several ways that could be arranged depending on how exactly you structure the system.
It would change the incentives that inform settlement negotiations. It's fairly common for defendants to settle cases they're very likely to win at trial because it's cheaper to settle than to go to trial. Plaintiff attorneys know this and it factors into their decisions on whether or not to advise their clients to file cases.
Please. I just want to go climb trees in the park without anyone worrying that I'll fall and break my arm and sue them.
Former (Canadian) bureaucrat here, and I think you're making a prior assumption that may not be true (and was certainly not true in the depart of the Canadian government I worked at back in the day). The number of bureaucrats may not be optimized to the amount of work they have to do. Some of them may not be working very hard, or really very much at all. At least in Canada it is almost impossible to get fired as a unionized government employee. If you don't commit a criminal offence or cause your department to get onto the national news in an embarrassing scandal, you are probably fine even if you don't do any actual work. My experience was that about 5% of the people there did all the work and the other 95% averaged about 2 to 3 hours a week of actual work. The primary reason I left was that the temptation to coast like that was ever-present and I found it to be somewhat soul-destroying.
Whelp, looks like we should get rid of all of them then.
Actually...what if they just shutdown the government and just...didn't open it back up? Veto every budget bill, force congress to veto override or impeach/remove?
Has Musk thought of this yet? Military still gets paid, essential services would still need to be provided. I'm sure 90%+ of the people affected would be democrats...could someone steelman the counter-position here?
Do you honestly think that Democrats make up 90% of Medicare recipients? Not to mention Social Security..?
Medicare and Social Security recipients still get checks during a shutdown...they're considered mandatory spending
Ahh, thank you.
The counter-position is that a bunch of good things don't happen that need government approval - for example, no new power plants are approved and no oil wells are approved. If people build the power plants and oil wells anyway, they'll get sued, and then the courts have to deal with it (and they'll shut it down, because it's blatantly illegal and the courts aren't that biased).
I suppose the next version is that you first repeal all the laws banning anything, but that's not happening. Even if you magically did though, you would then get a ton of people mad about people doing stuff they don't like (like dumping chemicals in rivers, building coal power plants in their backyards), and then you would be voted out and everything would be put back in place.
The final version is that you just abolish the government entirely and get rid of the constitution, but again that's not happening, and then you lose the military and immigration control, and presumably you like those. I mean, maybe you're a hardcore libertarian and you at least have a consistent policy, but even Trump and Musk don't want that.
The core issue is that Americans (and people in general) *want* many things to be illegal. The FDA exists because there was a huge public outcry about drugs that were causing harm, and they elected politicians that promised to do something about it, and they created the FDA.
Also, my interaction with the federal government is in the energy industry. The main thing that the bureaucrats are doing that's relevant for us is making everything simpler and faster. Currently every electric company and state has processes that are different from each other in meaningless ways, but building anything across state lines or between different companies is extremely difficult because all these processes differ. The bureaucrats at the Department of Energy are coming up with simpler processes that unify all of the different ones. All of these companies and states can't agree on a single process, and someone at the federal level needs to do it.
So if we get rid of those bureaucrats, the result is that fewer powerlines and power plants are built - not because the federal government needs to approve them, but because the federal government is forcing industry to get on the same table and work together.
The federal government used to train bureaucrats to reduce red tape. They came up with a (forgotten) training course during World War II, to teach bureaucrats to reduce paperwork delays and thereby aid the war effort. I rediscovered this initiative and wrote about it here:
https://www.statecapacitance.pub/p/eisenhowers-bureaucrats
It's a very concrete look at what it takes to cut red tape.
While I can see your point for this particular FDA case, I think it's bad to dismiss the "too much bureaucrat" problem entirely. I still believe that "the bureaucracy is expanding to meet the needs of the expanding bureaucracy" because it still occurs even in pure capitalistic corporations who have to red tape one of their department talking to another, even when no one would really sue them if they don't do it.
It's just a fact that more people is harder to coordinate than less people, and some companies have tried to be leaner to account for it (either by mass layoff, or spinning a new taskforce solely for specific task so they can run faster). It's harder for government to do it because civil servant can't really be fired or demoted, hence the nonsensical tweet. But the problem it tries to solve is a real one.
Can anyone point to good research on how to effectively create and align incentives to get an effective bureaucracy? It's been an issue for...~3000 years? Surely there's been some systematic work by now.
Honestly if we can figure out the bureaucratic alignment problem, the AI alignment problem ought to be easy :P
It being unsolvable for 3000 years may mean AI alignment problem is unsolvable :D
But optimistically it may have been solved! For all I know Carthage had a perfect bureaucracy but because they couldn't mobilize like the Romans their solution was wiped from the earth forever. Wait, is that optimistic or pessimistic?
> Can anyone point to good research on how to effectively create and align incentives to get an effective bureaucracy? It's been an issue for...~3000 years? Surely there's been some systematic work by now.
Singapore has a famously effective bureaucracy, and LKY talks about some of the bureacratic practices and measures he instilled in his book Third World to First.
For that matter, China has a much better state capacity than the US, as everyone will say SIngapore isn't a good example because of its size. But anyone who's traveled to China or done business there over any appreciable time is amazed at the sheer pace they're able to build roads, power plants, skyscrapers, and everything else.
This seems to be mostly achieved by giving below-head-guy levels more autonomy and pretty clear KPI's to execute towards.
I'm not sure what the level of "legal recourse the average citiizen has against the government" is, but I'd certainly bet on it being "significantly less than in the US or Europe."
So the solution is probably more autonomy, more autocracy, and less legal remedy.
We had that, and then Old Penn Station was torn down, and Robert Moses happened, and then Overton Park v. Volpe happened, and now we don't trust autocratic public planners.
It turned out that there are costs to giving planners autocratic power and the backlash was significant enough that government in the US at national state and local level added emergency brakes all over the process of getting stuff built.
It also turns out that if you add in brakes everywhere, people will use them, whether that's for an intended purpose or not. So the pendulum will swing back and we'll probably spend the next decade or two tearing out the brakes until something suitably popular is destroyed or a vogon constructor fleet shows up.
This is probably right, but I'd like to cut the USA some slack on this problem. The thing is, when you're not the world leader, you have a model that you can follow. When your economy is already at the frontier and working better than anyone else's, the vast majority of things to be done will just make the country worse. And those few things that will make things better are, by definition, untested (in the conditions that exist at the frontier).
There's certainly something to be learned from China, but they mess things up all the time.
Perhaps someone wants to read and review this classic:
https://www.amazon.com/Bureaucracy-Government-Agencies-Basic-Classics/dp/0465007856
I agree with the though behind
"
— How likely is it that their decision will get challenged in court?
And if it gets challenged in court, what amount of paperwork do they have to show the judge to prove that they made the decision on a “reasonable basis”?
"
But is there any reason to believe the bureaucrats are well calibrated in this regards? Is a 5,000 page 6 year Environmental or FDA review any less likely to trigger a lawsuit then a 50 page 6 month one? Does it make it any more likely for the agency to win the lawsuit? I could easily believe it doesn't matter or maybe even the reverse (I'm sure there's a lot more to nitpick in the 5,000 page version)! Has anyone rigorously looked at this? Are the agencies constantly experimenting (ha! no!) or do all the incentives run towards being overly cautious, to 'doing more' even if that's actively detrimental, or even Molochian standstill (something like CA HSR where they'd get more money and more bureaucrats if they could actually get moving but the whole process is so regulatory borked that the whole thing is likely to get shelved).
I have some experience adjacent to the adversarial end of the bureaucracy (general contractor stuff on military bases but no clearance past the gate), and was always struck that the actual factual federal bureaucratese were all vastly more competent than their private sector and contracted equivalents.
If there were two or three times more government paper pushers at camp pen lets say, they could easily make back their salaries multiple times in efficiencies and cutting out stupid privatization.
Here's a meta-factor that I think you're not taking into account: If you yell at people loudly enough, make it abundantly clear that you can and will swiftly fire them, and tell them to figure out a way to achieve goal X or else, they often change their mindset and figure out a new way to achieve goal X. Because you have significantly changed their incentive structure.
This is not hypothetical. Musk does in fact use this technique. He even got rockets to land on their tails this way, you know?
I'm not making a strong prediction about whether he or Vivek will succeed at DOGE. But I do think your account of the present incentive structure at FDA is not telling the entire story.
"they often change their mindset and figure out a new way to achieve goal X."
If they have the power to do so. There's plenty of examples of people lower down the chain of command advising "this is a bad idea, it will cost too much, it's against regulations" and getting over-ridden by the higher-ups.
It's all very well for me, Acting Grade III, to figure out a new way to achieve goal X. But if the person who can sign off on that is the Grade VIII who doesn't want to do so, then whether or not Vivek fires me won't get that done.
And that is the problem we foresee here: it's not the Grade VIIIs and above who are in the firing line, it's the rank-and-file who are the ones dealing with the public. You can't fire the Principal Officer because they have an iron-clad contract and will take you to court and the case could run for years, but you can fire twelve front line staff because they're disposable. Now you can issue a press release about "number cutting achieved!" but you still won't have improved the services.
I think an overlooked aspect of the excess burden of bureaucracy is an excess number of bureaucracies. Many agencies have authority/responsibilities that overlap with another's; for instance, both the DoJ & the FTC can bring antitrust actions, so private actors have to jump through two similar-but-not-identical sets of hoops.
The crypto industry's complaints are at least partially driven by this: nobody's the exclusive regulator for them so everybody's pestering them (that the SEC and its we'll-tell-you-what's-prohibited-by-suing-you-for-doing-it approach is in the mix definitely exacerbates the situation).
There are places where refactoring the agencies such that at most one had authority over a certain thing could have salutary effect, e.g., carve up the FDA and combine the "food" mission with the USDA and the "drug" mission with the DEA.
I work in private. A oversight group was making various paper work requests of our group that got so burdensome we hired a team with the sole task of doing the paper work. Overtime the team managed to increase the required paper work. The requirements didnt come from them but they would always propose solutions to the oversight groups concerns that were maximumly bureaucratic. Now our group wants to interact with oversight group more directly. So there are some cases where the paperwork does in fact adjust to the number of bureaucrats.
There’s deregulation for its own sake, and then there’s maiming the system & documentation to make space for sycophantic new hires. Who will be “the only ones who can fix it” and will add myriad well hidden regulations that benefit themselves, or whoever they’re hired to benefit. Firing half the workers is probably an effective prelude to this. The code will end up longer, processing times will be longer, but whatever benefits are available will flow reliably toward the leader’s allies. And the old guard is no longer numerous enough to stop them.
I do think they're genuine about their desire to reduce bureaucracy. The party doesn't care what you do as long as you stay loyal to them. Dissenters can be dealt with on a case-by-case basis; no need to establish concrete rules for that.
I agree to a point - I think some of them are genuine in their desire to reduce the size of the bureaucracy. I think of “reduce the size of the bureaucracy” as an abstraction that can obviously be carried out in multiple ways. People who really like that abstraction clearly may or may not have thought about how to actually make it happen in a certain instance. Also if it were an easy problem it would already be done. So I think the people with the genuine desires for smaller government are the tip of the spear, and their earnestness will open up a whole bunch of gaps which will be filled by people with slightly different goals who want certain outcomes more than they want small government. The Idaho example is interesting because like many other commenters I think it doesn’t scale.
Ah... it's impossible, then, and counterexamples be damned. We need MORE government to prevent this from happening, clearly!
The harder we pull the string, the tighter the knot gets!
But seriously. Something Idaho did right was having each department cut their own code. They’re more likely to know which parts actually matter. But over a certain threshold - whatever it is, the feds probably passed it long ago - the code is so large you’d have to hire more people to revise it (removing that advantage of familiarity). Even county land use codes - just one little area of code - this is the part I’ve seen - a hundred pages of regulations and one wants to run screaming from the building “it’s alive!” The two people who wrote most of it last time know how it hangs together and for everyone else it’s …. Time consuming.
Also complexity is part of community life and some rule systems are legitimately complex.
If Ramaswamy was serious about smallifying bureaucracy he would set up a commission (I know) to find all Those Two Guys who wrote the code, bring them all back from whatever they’re doing now, sit them down, pay them handsomely and task them with making it elegant. That’s a plan I could get behind. The people who know what’s actually in it are the only ones whose edits will not make everything worse.
Cutting chunks off the document and firing half the staff doesn’t actually help. It’s not the first, second or third step in any process that will actually smallify effectively. It’s not an exact parallel to computer programming but one useful part of the analogy is that these regulation writers don’t comment their code and it’s not always self-explanatory even though it’s in the common language. (Yay lawsuits for clarification purposes).
For social media purposes, though, saying “bureaucrats first against the wall when the revolution comes” sounds compelling to some audiences. I appreciate Scott forever for his willingness to gently and thoroughly explore the actual results of this type of thing. I’d put him on the editing commission (he wouldn’t do it though!) because he has the ability to consider how to move language around to create structures in society. Not everyone has that. The incoming folks will try to streamline the government and they will prove things about systems and unintended consequences … probably often.
That all sounds pretty plausible to me---I just have a knee-jerk reaction to things that sound like "well, we can't ever go in the 'make it smaller' direction" (because I often hear that from people who profess to /agree/ that that would be better, and it's like "okay, so we're doomed forever: if any attempt to fix it makes it worse, might as well just let it balloon insanely for eternity, I guess?!", heh).
This is why I don't favour libertarianism, it replaces rule by elected representatives with rule by lawsuit—a clearly less efficient system. Reduce the scope of civil courts and increase the scope* of elected representatives (e.g. Governor Little).
*In practice they already have the scope they just need to *want* to do it.
I think this presupposes a degree of benevolence from the bureaucrats. I work for a Department of Energy contractor doing environmental remediation. The bureaucrats I deal with speak in terms of benevolence (“save tax payer money,” “achieve site cleanup”) but act in ways detrimental to those stated goals. More and more resources get tied up in talking about something than actually accomplishing anything. And when you act to achieve an objective, the bureaucrats second and third guess your decisions. Prolonging the endeavor is the bureaucrat’s objective, regardless of their stated intent. This is further exacerbated by decision by committee to avoid any responsibility or accountability.
Chiming in with first hand experience that generally aligns with this analysis.
I am a former bureaucrat who was part of approving international aid funding. I sat between the bureaucrats doing programming and the Ministers office (political staff) signing off on the funding.
Red tape is part of the democratic process. If something stupid gets funded, the Minister or Prime Minister wears it and could take a beating at the ballot box. What if the FDA approves a bad drug or former Taliban was allowed to immigrate? The court of public opinion and the news media create incentives for more red tape to protect against backlash. It’s easier to accept backlash for slow speed vs bad outcome.
We had a specific file that the Minister rejected for political reasons and tried to push onto the bureaucrats. When the political decision finally came to light, all those forms and double checks ensured the person who made the decision was held to account.
Secondly, I was tasked with simplifying the overall approval process from the top. Cutting red tape requires someone to deeply understand why something was added in the first place then efficiently condensing things. Chesterton’s Fence is a useful framework here. People like Vik appear to think everyone but them is a dummy. To me, he is just another form of elitist arrogance.
Absolutely agree. My experience with people doing programming/grants/etc was a desire to move as quickly as possible. The political level are the ones who often get squeezed for a mistake then run it downhill. Ironically, leading to a new piece of red tape somewhere.
I think the core conflict in the US is actually between Yale Law School grads and engineers from Cal. One group has been endlessly creative interpreting laws and procedures in novel ways to advance their goals while the other tries to create new things.
Ramaswamy is a Yale Law School grad, and not an engineer from California, so I'm not sure what your point is.
Probably that he doesn't actually want to solve the theoretical problem, he's just throwing red meat to the hogs.
Dude made his money by hyping drugs that had already failed in trials then dipping when it came time to carry the bag, why expect him to change his ways now?
The feeling most people have about civil service workers is that they don’t work very hard, because they don’t have to — they are almost impossible to fire if they just make sure they cover their asses, so that’s what they spend their energy on.
I honestly have no idea how true that perception is, but I know it’s pretty widespread. If you start with that, it’s easy to believe that laying off half the staff would encourage those remaining to do their job like it mattered to their own personal future, rather than act like time-servers and paper-pushers.
I think there’s also a widespread sense that government agencies spend an unreasonable fraction of their time doing things that are not actually in their mandate (DEI, for instance, or the CDC deciding that racism is a disease) and even things that are arguably not properly authorized by the Constitution in the first place. It *might* be that cutting departments in half would force them to scale back what they do to include only their legitimate duties, but I fear there is a step missing in that argument — more likely what you have to do is explicitly and thoughtfully cut back on their missions and then set staffing to match.
I have a vague intuition that the biggest problem with bureaucracy is about who will eventually have to take the blame if a decision is incorrect? Often, when you deal with the “bureaucrati in front of you, their hands are tied by a higher authority, and they’re probably grateful for that. But the same idea persists up the layers; a perfect paper trail, a diligent assigning of responsibilities, and the whole process completely infected with CYA, which I think most of us can understand to some extent. Ultimately the buck stops at some big shots’ desk, and big shots come and go; they resign, they reappear…
I don’t know. It’s a puzzle.
> who will eventually have to take the blame if a decision is incorrect?
That's what scapegoats are for. There's always someone that everyone hates that you can blame everything on. Or you can plan ahead and hire dedicated scapegoats that can be sacrificed when things go wrong!
Why would it encourage them to work harder, when the criteria for being fired is not "least hardworking" but "wrong social security number"?
Firing civil servants for not doing their job, is not actually impossible but it is very very hard and it will almost certainly remain so no matter what Trump et al do - that's baked deep into the system. Laying off civil servants because you've decided their organizations are oversized, or changing the rules re mandatory retirement, or firing them for some weird new reason nobody has bothered to set rules and precedents against, will be much easier.
So that's what will be done, and what is done will not serve as an incentive for the remaining civil servants to work harder. Except perhaps on polishing their resumes and networks.
Well, that last might work out. :-)
But I generally agree. We might hope that a growing sense that their jobs are not nearly as safe as they thought they were might spur them to work harder, on the grounds that if mere bad luck can strike them where they thought they were untouchable, maybe slacking would also be a bad idea. But I agree the model is simplistic.
I suspect you might get better results by firing everybody in a target organization and then staff it back up to 50% levels; layoffs who could point to solid accomplishments would be more likely to get rehired. Maybe?
But I'm afraid there are really only two ways to reduce government to half its current levels. One is to have the consensus will to spend a century doing it, bit by bot, as we have spent a century getting into our current fix. The other involves M. Guillotine. Neither is very likely.
I think it's definitely a stupid rule/worried about getting in trouble problem and not an intrinsically obstructionist bureaucrat problem. I have worked for county and municipal governments in Texas for the last 10 years (as a sysadmin and now cybersecurity) and the descriptions of how the federal government works or how the government of San Francisco works have always baffled me. My experience has been that when we needed to do something we sort of just did it. It didn't take 5 years. We were not worried some strange activist group would sue us. At most we did have certain state or federal rules we had to follow (CJIS - criminal justice information systems standards, say) and an audit every X years. Or if something was really expensive, you would have to wait until October to put in on the next years budget. But it did not take us 10 years of review to build a facility or 1 million dollars to build a public toilet in a park.
I think population density and the density of existing infrastructure would have an impact on that, although to what extent I would not know. Building a tunnel under the east river of Manhattan is a different proposition than doing something of that scale where you are working, I would imagine.
I think so but Europe builds subways way faster and for way cheaper than Manhattan, and I don't know why it should cost 1 million dollars to build one toilet even in San Francisco.
https://www.nbcbayarea.com/news/local/san-francisco/san-francisco-noe-valley-expensive-toilet/3511302/
One of the cities I worked for had an on paper population of 13000, but a day-time population of over 100,000 because it was in a very popular shopping area and also had several major employers (a couple of hospitals). There was a big stretch of city road that was having a lot of wrecks from people making risky left turns. They decided to build a median down the whole length. The whole process of doing that from planing to completion was like a couple years at most.
Another place was an exurb with a population of about 50,000. I worked there for about 6 years. During that time, the city built 1. a brand new fire training facility, 2. a huge new fire station and police sub station on the West side of town, 3. an incubator facility for local small businesses to rent office space. They also remodeled the public library, bought and remodeled a building for the growing planning and engineering department, and then remodeled the old planning building and gave it to us (IT) because we had run out of space for storage where we were formerly squished in the server room at the old PD.
This is all within the Houston area. Not as dense as New York or LA but hardly the country either.
I agree with you about the insolence of office and the laws delay. The reasons for it are often perverse and difficult to understand. I can almost guarantee that it’s working for someone somewhere.
I sometimes wonder about the amount of damage Western society today faces because of its own wealth and generations of capital accumulation at the family level. When your family is rich enough for you to be genuinely comfortable your entire life without doing anything productive, but you aren't the brightest bulb, so you end up joining these 'advocacy' organizations. And because lots of rich people know people like this in their circles, this is going to continue to go on, despite how parasitic it is.
I am very skeptical of the claim that the "advocacy organizations" are staffed primarily by trust-fund babies. Do you have any evidence?
PMC, yes, but the PMC still needs a reasonably high-paying job to maintain their status.
Staff matters. Direction matters.
It is easier for staff to repeal regulations than it is to implement them. It is also easier to ignore them than it is to enforce them.
The problem with work exceeding the capacity of bureaucrats is both imaginary (it doesnt actually exist anywhere) and false. The real problem is they are sitting around trying to think of things to pile a regulation on. Thinking of a person to file a lawsuit against.
Javier Milei fired a bunch of federal bureaucrats and slashed regulations and size of government. Now Argentina's inflation has dropped from 300% annually to 20% while the economy expanded 4.5%. This seems like proof that getting rid of unnecessary bureaucracy can be a good thing, no?
Unnecessary is the key part. Selecting by SSN isn't correct, though I'm sure that process was proposed sarcastically.
First off, the Argentinian system was burdened by an outrageously bloated state bureaucracy and an irresponsible fiscal policy that resulted in several national bankruptcies and caused the absurdly high inflation rates. So in that case, yes, dramatic measures are probably needed. AFAICT, the US is nowhere near that situation. Yes, there was some inflation, but that happened in Europe as well, probably due to similar issues - Covid, the Ukraine war and other factors disrupting supply cains and economic activity. No, a national bankruptcy is not in the cards unless Congress is extremely foolish. An unemployment rate of 4% is also not particularly worrisome. So, to take an analogy, while there's good evidence that chemotherapy can save lives, you probably shouldn't prescribe it if the patient doesn't actually have cancer.
Second, where do you get the number of +4.5% from? What I read is that they have barely managed to stop the bleeding, and are hoping for a recovery this year. Also, unemployment and poverty rates did go up, as expected. This may be necessary hardships on the way to recovery, but that recovery hasn't fully manifested yet.
All I know is that I bought into some Argentinian companies and my buddy got into Argentina-heavy ETFs on my advice when Milei was elected, and they've made us fantastic returns—better than nearly any other positions in the portfolios.
Does this indicate Argentinian economic strength? I dunno; that was my impression, and why I bought in, but if not I can live with that too (the profit helps).
Milei has a lot more authority to get rid of rules and regulations than the US President. Most things the federal government does, it is mandated to do by Congress (and that includes the writing of regulations).
In practice, I doubt that the number of forms provided has any impact on the outcome of a case, and if it does it will surely be at least logarithmic to the number of forms. But there's no incentive NOT to have one more form, so up it goes.
Scott, I'd love you to do an article looking at this on the UK side. We have the Lower Thames Crossing (linked below) - a tunnel whose application document alone has broken records, at over 350,000 pages and the application process costing over £300m.
Are some bureaucrats needed? Of course. But any system that creates this outcome is utterly broken beyond doubt.
https://www.cityam.com/lower-thames-crossing-planning-application-becomes-uks-longest-ever-at-more-than-350000-pages-and-costing-almost-300m/
Scott is not arguing that the system is not broken; just that the fix is something other than firing bureaucrats at random.
Plainly you can save money by employing fewer bureaucrats, but if you want to make government more efficient you have to reduce the number of forms and regulations in parallel with those layoffs.
There is a new problem in all this: no one has ever had to prune government employed bureaucrats to a significant degree. Let me give an example: when I moved to Canada in 1985, hospitals managed their own budget and were run by volunteer community boards. Physicians gave their time to sit upon committees, unpaid (and had a vested interest in keeping them to a minimum). My province's government decided to save money by employing administrators to 'professionalize' the running of hospitals, and have several levels of management to co-ordinate services so that there was no duplication. What could possibly go wrong?
My community hospital went from 29 beds, 24/7 ER,X-ray,and lab run by 2-3 physicians and one administrator to 8 beds, ER 8am-4pm on weekdays, same for x-ray and no lab. But has five physicians who are salaried and see very few patients, and 14 administrators.
The regional hospital went from 100 beds and four administrators to 50 beds and 74 administrators.
Attempts to streamline this result in new vice-presidents being appointed (along with support staff) to "oversee change." And patient care spaces get taken over for offices for them. Administrators never reduce their empires, but always expand. It is like watching an old oak tree being smothered with ivy, parasites sucking the life out of the body. We haven't faced this before, and have no good idea of how to reverse the process. But we have to figure it out and do it, or else we shall see the old joke come to life that "this hospital would run better if there were no patients in it."
What province were you in?
NS
Makes me wonder what things were like in NFLD
I'm sure it's much the same everywhere, and the same kind of thing is happening in the NHS in the UK. Bureaucracy grows like a cancer, with much the same effects on the host.
My only experience with the healthcare system in Canada was living in Ontario. In Toronto, particularly I’m sure the regional disparities are extreme.
Sorry if this has already been said in the comments, but as a lawyer, I found the "what if they get sued" section to be disappointingly beneath your usual standards of rigor, Scott. The case you cited for the "reasonableness" standard is not an employment case, but a claim under the Administrative Procedure Act, regarding the standards a federal agency must meet when issuing regulations and responding to citizen petitions about issuing or amending regulations. (Google "CRS petitions for rulemaking"). Instead, reductions in force of federal employees are governed by an extremely complex set of rules that probably only a government lawyer or union rep could explain (Google "RIF rules"). Except to the extent a such a DOGE-esque layoff relied on an agency interpretation of the rules governing reductions in force, the APA would be irrelevant. If anything, Vivek's tweet is another example of how he and Elon seem blithely unaware of how much law they'll need to tangle with in order to achieve their goals.
"Fire half of the bureaucrats" is obviously a simple solution to a complex problem. That proclivity is not a conservative trait nor a liberal trait. It's a human trait. It was incredibly effective for cavemen.
The problem is that our subconscious mind evolved for the animal to survive and thrive and its environment. It took millions of years for it to evolve to the point where caveman was well served. It hasn't evolved much past. However, our rational capacities have skyrocketed in the last 100,000 years. What we have today is it chimpanzee attempting to operate a supercomputer.
The brain essentially operates just like a computer - garbage in, garbage out. Every computer programmer thinks that what he's written is solid until he tries to run it. The programmer attempts to blame someone or anyone other than themselves for the multiple crashes. Then comes the arduous process of trying to locate and fix all of the bugs. But the computer is built to be an incredibly rational entity. The human brain is built for every dog to snatch his big a piece of meat as he can and bite anybody who tries to interfere. So the problem is that our cognition makes decisions based on far too little information and grossly lacks the predilection to do sufficient research to do any better.
In reality, the problem isn't due to Congress, the federal bureaucracies, the massive number of regulations, or the selfish idiots voting these selfish idiots into office. The problem is human nature. Very few people possess the cognitive capacity and skills necessary to do any better than what they're doing. The result is a shit show that no one cares to take responsibility for.
I challenge every reader to present a solution to that problem. I guarantee it's deeply errant if the proposed answer is less than 57 pages.
Einstein found the problem that resulted in general and special relativity sitting in a chair imagining what light produced in a moving train looked like to both someone in the train and someone on the platform. That probably took 10 minutes at most. But it required thousands of pages of complicated mathematics to prove it.
Great article, but my only gripe is this: Everyone wants a simplified process for getting drugs approved, but criticizing the number of “processes” and “forms” without specifying exactly which processes should be eliminated or simplified isn’t that helpful
People are very good at identifying problems with the status quo but bad at thinking about trade offs. The FDA that Scott seems to want would be more opaque, arbitrary, have more lenient quality standards. There would be some upsides to that, but I’m very skeptical that the public actually wants that.
It’s also much easier said than done from a regulatory and statutory perspective.
I don't think "more opaque and arbitrary" is something Scott wants for the FDA.
More generally, I think it's meaningful to say "we err on the side of too many regulations and not enough approvals" without reviewing every extant regulation and process; in past posts, Scott's made a case for why the FDA ought be more lenient, which involved things like calculating cost vs. benefit as it stands now vs. hypothetically approving more drugs (e.g., what has the cost been from delaying good drugs vs. approving bad ones, and what is this likely to look like with X% more lenient standards).
In other words, this is the sort of big-picture question which could go either way, and which posts of Scott's were spurred—IIRC—by a few people writing articles saying it does go the other way ("FDA needs to be stricter")—so it's not a meaningless question, even without saying exactly which pages of the 20,000 page* FDA Guidelines for Drug Approval need to go. Building consensus that yeah, this probably needs to be fixed in a certain direction /is/ helpful.
>but I’m very skeptical that the public actually wants that. [...] It’s also much easier said than done from a regulatory and statutory perspective
Yeah, Scott's aware—cf. the bit about people "yelling at" the FDA, which is treated in greater detail in the other posts on this question.
-------------
*or whatever
I think it's important to remember why the FDA (and similar bodies internationally) were established; it's because in the halcyon pre-regulation days, private enterprise was putting literal poison into products sold to the general public as food, drink and medicine. And literally killing their customers:
https://en.wikipedia.org/wiki/Swill_milk_scandal
"The swill milk scandal was a major adulterated food scandal in the state of New York in the 1850s. The New York Times reported an estimate that in one year, 8,000 infants died from swill milk.
Swill milk referred to milk from cows fed swill which was residual mash from nearby distilleries. The milk was whitened with plaster of Paris, thickened with starch and eggs, and hued with molasses.
After the extraction of alcohol from the macerated grain, the residual mash still contains nutrients. Therefore, keeping cows stabled near distilleries and feeding them with swill was an economic advantage.
...In May 1858, Frank Leslie's Illustrated Newspaper did a landmark exposé of the distillery-dairies of Manhattan and Brooklyn that marketed so-called swill milk that came from cows fed on distillery waste and then adulterated with water, eggs, flour, and other ingredients that increased the volume and masked the adulteration. Swill milk dairies were noted for their filthy conditions and overpowering stench both caused by the close confinement of hundreds (sometimes thousands) of cows in narrow stalls where, once farmers tied them, they would stay for the rest of their lives, often standing in their own manure, covered with flies and sores, and suffering from a range of virulent diseases. These cows were fed boiling distillery waste, often leaving the cows with rotting teeth and other maladies. The milk drawn from the cows was routinely adulterated with water, rotten eggs, flour, burnt sugar, and other adulterants with the finished product then marketed falsely as "pure country milk" or "Orange County Milk"."
That background, and high profile cases like thalidomide, selects for a high degree of cautiousness in any government entity - see the quote from the Department of Labor about "out of an abundance of caution and to avoid any doubt as to its validity" above. So you triple-check everything and make sure all possible 'i's are dotted and 't's are crossed, even if that takes longer.
Wow.
It seems like the problem here was one of "information liquity": apparently, no one made the connection between the milk and infant deaths (one wonders where the "8,000" number came from---controlled for general conditions among poor New Yorkers?) because "germ theory was in its infancy" and (presumably) because people were generally ill-formed at the time.
I bet such a strategy as "selling diseased, rotten milk that kills thousands of infants" wouldn't be very viable in today's America... although---actually, conditions in factory farms even these days make me question that, come to think of it---
Lord Rama *is* associated with cutting a lot of people in half, though. Probably some of them were bureaucrats!
Less crucially, but maybe it'll be relevant on some object level or whatever, I think you're modelling Ramaswamy's motivations wrong. I think he just wants government to not have to pay half of those people's salaries. Possibly he overestimates what proportion of the federal government consists of bureaucrats, or maybe he's just like everyone else on this point and hates bureaucrats more than he hates soldiers or medicine.
As for the issue itself, I think the root problem is that America chose litigiousness as its solution for the tension between liberty and responsibility. This probably made a lot of sense at the time! However, as time passes and the monetary stakes go through the roof, the value of optimizing for gaming these systems also shoots up, sometimes far past the value of e.g. creating new products. Then the same ends up true of the government, developing new bullshit counters becomes legitimately more productive than solving people's remaining, often intractable problems – and now we have huge bureaucracies, cost disease and medical insurance so good it makes people homicidal.
Other countries don't seem to have this problem or not to the same degree. Admittedly part of it is probably because they just don't have the wealth and industry – Belgium isn't making all the new drugs, so their FDA only needs to regulate the importation of drugs somebody else already proved safe at huge bureaucratic expense – but a lot of the rest seems to consist of just being perfectly content banning things that Americans would chafe at. No bureaucracy, the politicians just pass a law saying you can't own scissors under the age of 30 or some stupid shit like that, and people just *accept that*, and in exchange, if you decapitate yourself with your scissors you can't sue the company because you were either of legal scissor-owning age and should have known better, or you had those scissors illegally and it's not their problem.
I work in the electrical industry, working with a bunch of different groups including electrical utilities, transmission operators (the people who actually run the cross-state electrical grids), state regulatory bodies, and the Department of Energy. My experience is that a lot of what the Department of Energy has been doing, particularly in the last four years, has been making it easier to build.
Each member of these groups has their own idiosyncrasies and different rules and procedures for developers who want to build things like power plants or power lines or data centers. If I want to build a power line from California to New Mexico, through Arizona, I have to worry about the processes of three states' regulatory bodies, plus the processes of the people I'm connecting to on both ends. The bureaucrats at the Department of Energy have been doing a lot of really important work in simplifying these processes and forcing the different groups to all work in the same way.
Getting rid of the federal bureaucrats doesn't mean that there's less red tape, it means that there's *more* red tape, it's just from private entities and state bureaucrats instead. It doesn't make any sense to get rid of a federal bureaucrat if it means that every state in the nation has to hire one to replace it, and then every business has to learn fifty different processes instead of just one.
This is really interesting. Thanks.
"If I want to build a power line from California to New Mexico"
then God help you, given that California seems to want to slap warning labels about "this is a carcinogen" on everything including air and water? 😁
I think a lot of us who have worked in/adjacent to the public service have similar war stories about how it's much different on the inside than the commonplace image associated with it.
>An old tweet from Vivek Ramaswamy, now co-head of the Department of Government Efficacy:
The E in Doge is 'Efficiency'
"red tape expands to consume the resources available to it"
There is a HUGE amount of truth to this.
And in the opposite direction, the absence of resources does indeed drive process simplification (and shortcuts). This shouldn't be dismissed out of hand.
I would argue that well considered process streamlining has been historically rare in government; especially when compared to changes driven by resource limitations.
Could another thing be labeling each regulation with an individual person responsible, like a git repository? If the specific element is challenged successfully, then the person who put it there is blamed, with suitably severe punishment to further disincentivize people to expand regulation?
Mostly when you follow the chain of "who is responsible?" to the end, you end up with Congress.
I more meant the person who actually wrote the regulation. Not the “make a regulation” guy (which is congress), but the guy who hammered out the details.
Here you go, decide from this who exactly is "the guy" who hammers out the details of the relevant regulation:
https://www.reginfo.gov/public/reginfo/Regmap/REG_MAP_2020.pdf
The person who physically signs off on it, or clicks OK.
That may not be as simple as you think; just looking at some new rules on the Federal Register, one is from the IRS, one is from the fisheries service to do with tuna catch allocations, and I'm not terribly sure who gets the blame or credit for signing off on that one.
https://www.federalregister.gov/documents/search?conditions%5Bpublication_date%5D%5Bis%5D=2025-01-13&conditions%5Btype%5D%5B%5D=RULE
Is it:
"Karen H. Abrams, Acting Director, Office of Sustainable Fisheries, National Marine Fisheries Service."
But the allocations are the result of a series of agreements:
"Atlantic BFT fisheries are managed under the 2006 Consolidated HMS Fishery Management Plan (FMP) and its amendments, pursuant to the Magnuson-Stevens Fishery Conservation and Management Act (Magnuson-Stevens Act; 16 U.S.C. 1801 et seq.) and consistent with the Atlantic Tunas Convention Act (ATCA; 16 U.S.C. 971 et seq.). ATCA is the implementing statute for binding recommendations of the International Commission for the Conservation of Atlantic Tunas. HMS implementing regulations are at 50 CFR part 635. Section 635.27(a) divides the U.S. BFT quota, established by the International Commission for the Conservation of Atlantic Tunas (ICCAT) and as implemented by the United States among the various domestic fishing categories, per the allocations established in the 2006 Consolidated HMS FMP and its amendments. NMFS is required under the Magnuson-Stevens Act at 16 U.S.C. 1854(g)(1)(D) to provide U.S. fishing vessels with a reasonable opportunity to harvest quotas under relevant international fishery agreements such as the ICCAT Convention, which is implemented domestically pursuant to ATCA."
So if I disagree with transferring the tuna quota from December to January, who do I go after as responsible in the end?
Sometimes there are named individuals - as here. But as here also, those people seem to have moved on or been transferred:
"...(W)e are affirming and ratifying a prior action by Jane Oates, Assistant Secretary for Employment and Training, and Nancy Leppink, Deputy Administrator of the Wage and Hour Division.
...The Final Rule was signed by Assistant Secretary Oates and Deputy Administrator Leppink. We have full and complete knowledge of the Final Rule action taken by former Assistant Secretary Oates and former Deputy Administrator Leppink. Subsequent to the Secretary of Homeland Security's documented approval of the Final Rule dated November 25, 2024, in consultation with the Secretary of Labor and Secretary of Agriculture, and out of an abundance of caution and to avoid any doubt as to its validity, we have independently evaluated the Final Rule and the basis for adopting it."
So even if you're angry about letting migrant workers in to work on the harvest, Oates and Leppink are now former officials, so how do you punish them?
Probably whoever “we” is. Current person who ratified it. So there always needs to be a person standing behind it, who is willing to stake their personal reputation on the rule being good. It would require an ID based commit system, like git. Whoever committed the bad code is the person responsible.
Perhaps the only punishment needed would be public shaming?
You don't want to be the engineer that committed atrocious code into a repo.
I certainly hope so. But there needs to be some mechanism for firing a bureaucrat who consistently pushes bad code.
I think that would come naturally! Ie. whatever group they work under would eventually stop assigning work to them because they'd be afraid of institutional shaming.
Shame is transitive (I hope!)
>possibly illegal (does the administrative branch even control how judicial procedure works?)
Not the Executive, but Congress could strip lower court jurisdiction over cases where someone sues the Federal government via the Exceptions Clause (Article III, Section 2), allowing only the Supreme Court to take such cases. The Supreme Court hears <100 cases/year, so this will reduce the burden on the government by ~99%. There's also tricks involving International treaties (Missouri v. Holland), but again this requires Congress to cooperate.
That being said I'm sure the Executive could still do lots of good if they applied every trick in the book to overcome the bureaucracy, even if it won't last forever.
On the one hand the argument is strong, on the other hand having incentive system of adding bureaucrats until you get the results you want has some obvious failure modes and cutting the staff by half with the message that you expect them to streamline things creates credible incentives to actually streamline.
Ah, the ol' trick of murdering one of your lackeys at random just to send a message to the others.
"the message that you expect them to streamline things creates credible incentives to actually streamline."
Yeah, we had a guy who streamlined like heck in the old job in the social housing department. Turns out he just promised everybody who applied for a house in the projected new development "yes". Naturally, this ended up with 100 people all expecting to get their promised houses in the 25 accommodation development. Not his problem, because he had quit to go work elsewhere before the chickens came home to roost.
That sure streamlined things, but it didn't in fact solve any problems. What it ended up doing was *cause* problems, because he slipped up and made the mistake of sending a letter to one of these applicants in effect telling them "yes" (generally it had been verbal assurances) and so the applicant was able to go to court with this as evidence and force the council to give them the house.
Very popular decision with the rest of the people on the waiting list, as you might imagine.
I could have streamlined application processing for the grants by telling everyone "yes". Problem is, there wasn't the money for that. Streamlining is great *IF* the *entire* process gets streamlined (e.g. yes we will be building 200 new houses and apartments to come onstream by this time next year) but not so much if it doesn't, and the public-facing level of "I process your application" civil/public servant does not have the final say on "sure guys, send the diggers in to start on the foundations!" Not even the head of the department has, that decision rests finally with the holder of the purse strings, which is the national government in the capital.
It also creates credible incentives to look for a job in the private industry, since they pay better and may be less arbitrary in who they fire. Guess who'll be the first ones to leave?
Also, the number of public service employees who have any competencies to streamline anything is probably pretty small. "Incentivising" (or actually, threatening) THEM might make a bit of a difference - for the rest, all that remains is the demoralization.
As an urban planner who has worked on trying to streamline approvals, I really appreciate this post. People too often argue about small or large government, and there’s not nearly enough discussion on how to set up regulations so they’re less inefficient.
For three years I’ve been working on getting a housing development approved that has the support of politicians, the public, and city staff. It has even been approved by council! But still individual staff members who get the file have found reasons to delay the project by years. There is just no incentive for them to do otherwise.
One day I will run for city council and will attempt to set up better incentives.
Reminds me of David Graeber's book Utopia of Rules. In it he says something like: Once a bureaucracy is established, it's almost impossible to get rid of it. Looking at hundreds of case studies didn't reveal a single instance of red tape getting removed; it only ever gets added. To paraphrase Max Weber, the only effective way to get rid of a bureaucracy is, really, to kill them all, which is what Genghis Khan did.
- I want a job where there's achievement rather than merely activity. I'm tired of pushing paper. I want to point to something and say "I did that."
- I don't understand.
- I know. That's why I'm leaving.
- You're not saying the government is unimportant?
- No, it's very important. It's just that I haven't met anyone who's doing it.
Yes Minister, 1980.
Vivek's proposal reminds me of the joke ascribed to Pope John XXIII; when asked "How many people work in the Vatican?" he allegedly replied "About half".
The problem, of course, is to identify *which* half. Cost-cutting and headcount reduction proposals are nothing new! But the catch is, it's never the Assistant Deputy Vice Junior Secretary levels of management which get cut, it's the low-ranking grades that are public facing.
And then people complain about "I can't get through to a live human on the phone; my application has been in limbo for weeks; why does it take so long to process a simple request?"
Well, because the staff were cut in half and not replaced, but the volume of work remained the same and may even have increased. I saw this with my own two lying eyes in the public service in the wake of the Croke Park Agreement:
https://en.wikipedia.org/wiki/Croke_Park_Agreement
"According to the implementation body, as of March 2012, headcount had been reduced by 28,000 and the annual pay bill had been reduced by €3.1 billion. Over 7,000 public servants were reportedly redeployed."
Yay hurrah, efficiency increased and bloat reduced, right? But what does that look like on the ground? On the ground, in the office were I was, it was a colleague reduced to the point of nearly having a breakdown and needing to go out on long-term medical leave due to stress. You see, a new scheme for rent assistance had been spun up from the pilot programme and was now live nationwide, and she was doing it for our area. On top of all her other work. With no assistance, dropped into "this is starting now and you are handling all the applications". And getting all the hassle about "why aren't people getting their payments?" when it wasn't her fault that the information was not being processed, because it was all piling up on someone *else's* desk in our city office and not being put through the system.
And had she gone out on (deserved) sick leave, the files would have just piled up on her desk and nobody else would have been able to deal with them, since there was no slack or redundancy in the new, lean and mean, cut costs and cut headcount, no new recruitment when people leave or retire, regime. A conscientious, hard-working woman literally reduced to tears because there was too much to do, not enough time in the day to do it, and all the shit was rolling downhill to her, who had little to no control over what was going on.
And then the public continue with the idea of 'lazy civil servants who don't care, are just clock-watching, and are inefficient' because their application still hasn't been dealt with after six months. It's not because the person responsible is lazy and inefficient, it's that they are over-burdened and driven to the end of their rope.
That's the problem with all these top-down schemes that come from Big Bright Ideas - nobody ever asks the people on the ground, the customer and public-facing staff who are the interface between the bureaucracy and the clients, about what they think or how it will work out. What are the obvious points of failure, what won't work as intended (I still fondly remember "new software package to implement online housing applications that won't accept names with apostrophes - in a country where people are called O'Brien and O'Mahony and other such names").
I'm not denying there is waste, inefficiency, and featherbedding in the public service as in any other industry. I am asking "Sure, fire the half who don't do the job - but make sure you know which half that is, by a better metric than 'SSN ends in odd number'".
By the bye, Vivek, what does *your* SSN end in? 😀
When there's a cutback, the things cut back are always the public-facing stuff. The managers never get cut, just the people who actually do stuff the public cares about.
So your application takes longer to be approved, the park is closed, the welfare office has reduced hours. But the number of managers is still the same; they've moved them to a different office, or they're presiding over a reduced number of reports.
Will the Cabinet get reduced in number? Will the number of agencies get reduced, or their overlap in function? Probably not.
"Will the Cabinet get reduced in number? Will the number of agencies get reduced, or their overlap in function? Probably not."
Oh yeah, we all know that. The irony here being particularly thick, since forming a new government agency (DOGE) means carving out its own territory and hiring new staff to run it 😁 Well, I suppose they could transfer some of the "your SSN ends in an odd number" laid off staff to work there!
I get the impression that DOGE is more of a committee than an agency, with the acronym being chosen to annoy the usual suspects.
I could be wrong though.
"But the article suggests that ordinary people’s lives were made easier, and that the move has brought businesses to Idaho. It gives the example of repealing a regulation about what kind of doors pharmacies can have."
Out of curiosity I looked that up, and the old regulations were about security when dispensing filled prescriptions to be picked up outside opening hours - so basically "make sure junkies or thieves can't break in to steal the drugs":
"27.01.01 State Board of Pharmacy Rules of the Idaho State Board of Pharmacy Section 605 Page 72
03. Storage for Delivery.
Filled prescriptions may be picked up for delivery from a pharmacy when the pharmacy is closed for business if:
b. The secured delivery area has walls that extend to the roof and solid core or metal doors, and all doors and other access points must be equipped with locking devices and be constructed in a manner so that the hinge hardware is tamper-proof when closed;"
The new regulations are much more stripped-down, but I imagine Idaho pharmacies still adhere to "make sure doors and locks are secure" practices from the old regulations, because imagine if the pharmacy is broken into, and the insurer refuses to pay on the basis of "a five year old could have picked that lock"?
The regulation might have good intentions, but there are externalities. Where I live, it’s okay for a public facing business to have an entrance that is wheelchair accessible with a little help (eg, a small step). Same with bathrooms. But if that same business asks for ANY permitting to make even a minor change, they are required to build a ramp and re-do their bathrooms. So businesses either don’t make improvements, or do it without permitting, or have to spend a lot more money than they might be able to afford. This happened a particular business which I frequent, which is an old school gym. I’ve spent time there and I’ve never seen someone in a wheelchair. Also I would imagine there are better alternatives for someone in that situation than an Olympic weightlifting gym. And I’m sure if there was a particular person with physical limitations who really wanted to pursue their dream of Olympic weightlifting, the staff would be welcoming and assist as necessary.
I think the ADA does many great things, and I’m totally on board with new building construction rules. Our cities are becoming more ADA compatible every year, and that’s good. And because we are on that trend line, maybe we need fewer rules.
Likewise, there’s a difference between having a practical recommendation for pharmacies which they might be eager to adopt, and a regulation, which leads to profiteering by the one steel door maker in town.
You can also incentivize more directly, for example by punishing pharmacies that get broken into due to poor security, but not if they found a way to manage. Similarly to how the worst penalties for drunk driving actually occur if there’s an incident.
To understand bureaucracy, everyone needs to read https://www.eatingpolicy.com/ (and Recoding America) by Jennifer Pahlka. It is absolutely brilliant and, saying this as a civil servant, perfectly accurate.
From all I hear about the federal bureaucracy, it is practically impossible to fire people, so actually doing work is voluntary. About half the employees choose to do so, at varying levels of effort.
So firing the half that doesn't work would be a big improvement. I think that is where the DOGE guys are coming from.
Of course, before you do that, you need to somehow make working mandatory.
They also plan to remove tons of regulations, which in either case would require far fewer bureaucrats.
You could ask; that works.
A lack of conceptual clarity in Vivek's view is the ultimate goal.
Is it to make things which the government wants to happen, happen more quickly? Then in some cases, you might need to hire /more/ civil servants. (This is Scott's point about # of drugs approved per FDA employee.) This applies to whatever it is you might want to do. E.g., nobody seriously thinks that if Trump wants to deport large numbers of immigrants, the best way to do that would be to first fire half of the ICE agents.
Or is it for the government to spend less money? At the outset, it's worth noting that all civilian federal employee salaries & benefits together are a very small portion of the federal budget. If you want real cost savings, you'd have to make major cuts in entitlement programs (Social Security & Medicare, mainly) and the Department of Defense procurement budget. Also worth noting is that reducing federal employees often /increases/ the amount of money the government spends, because their work is still legally required to be done, and so they get replaced by outside contractors (Booz Allen, McKinsey, etc.), who (1) charge more, and (2) whose spending should be reviewed by federal employees to make sure they're not overcharging, except if you cut the federal employees whose job is to make sure that the contractors aren't overcharging, then you'll /definitely/ get more contractors overcharging, because unlike federal employees who receive a fixed salary, outside contractors have a profit incentive to overbill..
For almost any other goal you might think of, "fire half the employees, at random" is VERY unlikely to lead you there. (Unless you define the goal tautologically as reducing the number of employees, no matter what the downsides and without reference to any purported benefits.)
This is not to say that there aren't problems with many (perhaps most federal, state, and local government agencies. But these critiques are already well-known and understood within those who actually look at the problem. (Cf. Scott's old post on "Yes, We Have Noticed the Skulls.") Serious people who have examined government operations (Don Moynihan, Jennifer Pahlka, the Partnership for Public Service, etc.) have suggested important reforms to improve both effectiveness and efficiency. (For example, modernizing procurement regulations for software to reflect how software is developed nowadays.) But Vivek's tweets are more like the ramblings of a drunk at a bar.
Many people are not aware of this fact, but comparing to Soviet Union, number of government bureaucrats in modern Russia increased, and it's still less bureaucracy per capita than in EU. Because governance is hard and requires human resources.
I am occasionally in the position where I need to claim expenses from finance offices in various public bodies. Most of the time this works reasonably well: there are electronic forms to fill with my bank details, I submit receipts and eventually I'm reimbursed. That's how the system should work.
There's one body that I deal with which:
- uses a paper based system, is spread over multiple locations and allows some of the workers to process paperwork from home offices
- hasn't updated their processes to allow for expenses to non-employees (e.g. one person asks for my employee number every time, and insists I need permission from their CEO to travel anywhere)
- has allowed a low-level employee to email my employer asking blankly whether I'm committing fraud
- regularly re-litigates expenses claims paid 6-12 months ago
- takes up to a year to pay out claims, never pays an amount the equals my claims, doesn't notify me when I'm paid, and seems to have no record of what is and isn't paid
Most people have experienced the second type of process. To the people on the inside it's probably largely indistinguishable from the first kind. I think the post overlooks this type of poorly performing process. Fixing bottlenecks like this absolutely could speed up all sorts of things. (I don't think Elon and Vivek have the discernment to identify these types of issues and reform them without damaging all sorts of well-functioning processes, however.)
You make interesting points from first principle thinking, but I think you’re missing the tacit knowledge of having managed a large organization.
People in organizations want to feel important. You might think that delivering results makes individuals feel important but this is the exception for a typical organization. That’s because delivering results feels like everyone collaborating together towards a single goal and that requires an outstanding leader that squashes distractions. This involves both being careful with planning but also firing a lot of people who naturally impede collaborative progress.
What happens most often at most organizations, is that employees who don’t have status try to feel important by creating new layers of complexity. The canonical example these days is people trying to turn woke viewpoints into company policy, like going through codebases and removing the term “black list”. This makes them feel important. Because it’s much easier to come up with reasons to tell people not to do something, this sort of effort is almost always in the form of blocking others from acting.
The DEI example is salient today, but large organizations are full of these distracting or blocking efforts. People trying to unionize, employees preventing Google from working with pentagon, organized outrage of some form or another, and so on.
Having fewer people, and importantly, having the possibility that there will be even fewer people in the near term, makes those that remain focus more on results and keeping their jobs. And importantly, it also makes the kind of people who seek status in unproductive ways to leave, because they can tell their antics are no good there.
Regardless of Rama’s explicit argument and his experience to justify it, I’m almost certain fewer people will within a medium term speed up for government.
Every $1 of spending on IRA enforcement activities results in $5 - $9 of return. There exist situations where MORE government is more efficient. General rules are hard to infer.
You're defining "efficiency" in terms of maximizing the government's net income. That is not the goal of any reasonable person who is not running a government. The goal is to maximize the prosperity of the nation.
If that $1 if spending results in $9 of "return" but also $20 of reduced private-sector productivity, e.g. because people who should be doing useful stuff are instead chasing down the paperwork for the latest audit, then the nation is less prosperous. If because of that lost productivity, people are laid off and the government is on the hook for $9 more in unemployment benefits or whatnot, then even the government is losing on the deal.
Or maybe those aren't the actual numbers. But you're leaving some of the most important terms out of the equation altogether, and in so do revealing something ugly about your true intentions.
You can disagree about what taxes should even exist, sure, but this is just mitigating tax evasion. There is no extra work to be done by accountants than just normally filing tax returns and paying the bills. If you want to make private companies more efficient by reducing the paperwork you should counter-lobhy turbo tax so that the government is allowed to tell you how much you owe in taxes (which they know but are not allowed to tell you!!)
Maybe so re the FDA, but as to the corporate sector and vast number of years involved in corporate restructuring elimination of people generally meant elimination of layers of unnecessary bureaucracy.
The bonfire of "red tape" (British-speak for the 1000-forms) here in the UK during the Cameron government of 2010- was a significant contributory cause of a fatal fire for 72 people in West London in 2017
Here's a summary from one of the UK news magazines:
https://www.newstatesman.com/politics/uk-politics/2024/09/the-grenfell-report-is-damning-for-david-cameron
The point is that improvements in fire safety regulation was delayed precisely because of a "Little Administration-esque" attempt to prevent new regulation from being drawn-up without other regulations being 'burned'.
Lord Rama should careful what he wishes for - he may accidentally invoke Lord Shiva.
I like Scott's articles, including this one, but he is being quite naive here. Ask yourself who gains if the FDA is cut in half? Politics is about the distribution of power--there is no constituency for across the board objective efficiency.
Of course, you can ask the same question but going the other way. Who gains if the FDA is doubled in size?
I think you are maybe missing the point of why ramaswamey said something idiotic (but i can't tell): he's posting a tweet and getting responses. The republican party has a major ideological tenent against "bureaucracy", and he is only appealing to this.
As far as "D.O.G.E" is concerned, it appears that it probably won't exist, as elan musk can't create a government agency out of thin air, only a private bureaucracy.
I think you are correct to be skeptical of how all this works. All this nonsense with the national debt and the budget is just nonsense, and politicians who talk about how bad it is generally have some hidden motive for doing it.
Sorry if I am a buzzkill, but this whole thing seems to be a lot simpler than you all are making it out to be...
I work with government bureaucrats most federal, but also a lot of state/city/county. Daily.
>My model goes:
...“processing 1,000 forms”.
Suppose they have 100 bureaucrats, and each bureaucrat can process 10 forms per year.
...
If you fire half the bureaucrats, now they can only approve one drug every 2 years.
That’s worse!
Scott,
This is a poor model. If you have 100 bureaucrats, a good half of them are doing next to no work period. They are effectively processing 0 forms per year, or even a negative amount by bogging down meetings with stupid questions, not completing tasks they were assigned, or socializing with productive people.
They also have the corrosive effect of otherwise productive people seeing them doing next to nothing, not getting punished, so now they too sit around all day doing fantasy football and personal email.
I had a city once hire me to "office space" a department. It had ~80 employees. I interviewed everyone and made notes about their tasks and how those tasks went based on objective measures. They ended up doing some early retirements and a lot of sending people to other departments, and afterwards the 45 remaining people got much MORE work done. From 80 to 45 and it worked better, much better actually.
Less circular finger pointing about why nothing was happening on X or Y. Less items stalled "waiting for Sally to respond to an email for 3 months". Frankly less socializing and standing around the water cooler.
Now some of this is of course layoffs let everyone know you actually mean business. But a lot of it is certain people in government are literally a net negative to the functioning of the office. And while the private sector is not some paragon of efficiency, in the private sector these people are generally found and culled a lot faster. A city can just raise you taxes and hire another person. Or more often, not hire another person and city function X just doesn't happen, or takes 6 months instead of 2 weeks.
Not all the government staff are like this, there are also a lot of overworked super diligent people. And an issue honestly with unfilled positions too, though that again can come back to the same thing (with the HR bureaucracy taking forever).
That said Ramaswamy's plan is stupid. You need to sort the wheat from the chaff. And there is so so much chaff.
I would probably fire the bottom 60%, increase the payscale dramatically, and replace a lot, though not all of them with much better people. Right now important positions managing like multi-billion dollar programs are being managed by government staff who make like $140k/year in DC. You don't see that in the private sector so much.
That isn't even getting into the whole issue with the legislature not needing to pass things that actually work or make a lick of sense. They can just legislate nonsense if they want, and often do. And then the staff need to figure out how to make nonsense work. Once again less common in the private sector because the people at the top are accountable to the same thing as everyone else (dollars and cents) unlike in government where the people at the top have an incentive system that is totally different from actually doing anything effectively.
LOL 😅
Can someone explain why bureaucracy is a bad thing? It seems to just be assumed.
I, for one, don't want xenoestrogens in my food. And I like the idea of cutting down on animal testing.
Man, why ya gotta be hatin' on pharmacy
(also if you happen to read this and also remember that retail pharmacy explainer I pitched in 2018, I published the most recent version of it in 2023 if you want to take a look)
In old movies they liked to show a busy office was filled with typists...to the vanishing point!
I highly recommend the book Recoding America for a perspective of government bureaucracy and infrastructure from a software engineering adjacent lens.
"Upon being read, many of the [Idaho] regulations were not justifiable, for example “rules for a lottery game show that never aired”"
Removing such inoperative rule looks nice of the graph, but decreases the bureaucracy by exactly 0 (as it was inoperative and applied to noone).
Also to be noted that those regular reviews of all regulations have to be done by someone, meaning the same bureaucrats searching for something to cancelled. Which inevitably leads to creation of "department for decreasing of administrative burden" with KPIs, yearly reports, workplans, etc
Re: Idaho, one dynamic that gets missed in these sorts of clear statistics (i.e. number of individual regulations or pages of regulation) are the messy ways the bureaucracies themselves adapt to maintain their status quo.
At a prior job, I had to interact with Arizona Dept of Health during Gov. Ducey’s administration. He’d put into effect many of the same rules you mentioned that Idaho has - re-justifying every regulation every n years, removing x regulations for every new regulation, etc.. The end result, however, was that AZ DHS just made up a bunch of unwritten, illegal “non-rules” that they held you to anyway. In one case, it meant them publishing an “interpretive and explanatory guide” that was twice as long as the actual regulations and included a significant number of completely made up new requirements.
When arguing with that division’s chief about it, she flat admitted that they had to do so because they were restricted from passing new rules. In the end, it made what sounded on paper like some kind of libertarian dream, in fact so dysfunctional that I told Ducey’s policy chief I’d rather open a new site in Chicago or Los Angeles than ever opening another one in Arizona.
None of this contradicts your argument. If anything, I think it bolsters the point that “if we just fire half the bureaucrats we’ll get half the red tape” is laughable bullshit.