Even that cleaning water channels thing is more complicated that was described in this review. The system, as actually implemented, as basically corvee labor; sometimes induced by force, sometimes induced by social or economic coercion, but definitely not voluntary. It's not an accident that the person showing him how the system worked was, in his words, a warlord.
It doesn't shock me that the system of "let's all do miserable backbreaking work for free" for replaced by another system, nor does it shock me that the people weren't excited to go back to it once the other system fell apart. Sometimes you're in a bad, but stable equilibrium.
The amount of civilizational damage "We Are The World" and it's spiritual kin were upstream of boggles the imagination. So, so beautiful, so, so well meaning, and so so terrible, all at once. It's like epic tragic poetry written by an evil god, using whole civilizations as it's ink.
Would you mind elaborating, ideally with sources? I hadn't heard about "We Are The World" before your comment, but my brief research on it seems to suggest that it's been a mostly positive influence on the issues it's targeted, particularly famine.
Every dollar that good, well-meaning westerners spent on that record went directly to an armed partisan with an AK-47, a grudge, and a new cudgel to hold over the heads of their oppponents in a long running civil war. I mean seriously, look at Etheopia, Somalia, and Sudan now. If that's what you think "mostly positive" outcomes look like, I don't know what to tell you.
[Also, you /really/ could have googled this for yourself, instead of trying to make it a public status contest. This really wasn't difficult.]
International 1984–1985 famine relief for Ethiopia inadvertently entrenched the Derg military dictatorship by allowing the regime to weaponize aid, fund its civil war, and enforce a brutal forced-resettlement campaign.
Manipulation of Aid and Military Strategy
Food diversion: The Mengistu government diverted international food aid to feed military forces and starve out rebel-held northern regions like Tigray and Eritrea.
Prolonged conflict: Foreign assistance kept the state afloat financially, freeing up domestic resources to fund counter-insurgency campaigns rather than agriculture or relief.
Apolitical framing: Global charity events like Live Aid framed the crisis strictly as a natural drought, masking the man-made political roots of the disaster.
The Resettlement Campaign
Forced displacement: Using relief logistics, the regime violently moved roughly 600,000 farmers from the north to the unfamiliar south.
High mortality: Conditions during transport and in the un-serviced southern camps were so atrocious that the relocation itself caused tens of thousands of deaths.
State control: The program served as a counter-insurgency tactic to depopulate regions supporting the resistance.
You misunderstand my comment as some kind of public status battle. I legitimately wanted to see sources that represented your position, because I struggled to find them. I have literally never heard of this song before, so it’s not possible for me to come into this conversation with a bias of this song and it’s effects one way or another.
Generally, when people ask for help understanding your position, you should be happy to engage with them, not hostile. The hostility actually dissuades honest attempts to understand your position.
That's fair, and I apologize if I misunderstood you. I don't normally comment in this space and I've spent all too much time in spaces wheer "Sawce?!" is a confrontational meme, not an invitation to good faith conversation.
I think you might find that the ACX community is a bit more academic than traditional social media shouting matches. We're all trying our best to confront our own biases and understand the world better with data.
literally the first sentence of the first reply was "Would you mind elaborating, ideally with sources" I also reacted to this in same way as original commenter initially but then reading the next sentence i was like "what? that person has never heard of the song before?" and thought maybe it was a genuine question...
First time I heard of it too. I was able to google and skim a reasonably distributed corpus of sentiments within a few minutes.
Also, I'd question the prudence of anyone who wouldn't be hostile to your ~"noble" intentions as such, regardless of your ortensible earnestness.
Your subsequent reply (of which my reply here is attached) is shamelessly malignant and insufferable, more-than demonstrating the appropriateness of said hostility.
Yes I remain shamelessly curious in life, which makes learning all the more enjoyable. As for insufferable, you'll have to ask my friends. I try to remain kind and respectful, but judging myself without bias is difficult.
To your point of "just google it bro," I did, and found most sources to be very supportive of the song. Hence me asking for examples to the contrary, in as polite a way as I could.
An American businessman and future president, Herbert Hoover, famously fed starving Russians during the famine Lenin caused, thus stabilizing Lenin's newborn regime and allowing it to become one of the most evil in history.
But should he have let millions of Russians die instead? Isn't that precisely the logic of Lenin and Stalin, who were also willing to sacrifice millions of people for what they considered the greater good? I'd be extremely wary of any galaxy-brained argument that a pile of skulls is good, actually, because it's an epistemic trap that has led to horrors time and again.
There are a few things to address here. In no particular order:
Causal proximity: No, it is obviously unfair, even within consequence utilitarianism, to blame Hoover for Stalin and Kruschev. On the other hand the partisan witih an AK-47 and a grudge is not an abstract politician in some faraway capital. He is standing right there, just out of camera shot, glaring menacingly at everyone you give a cup of grain to. The choice is not between skulls and no skulls. The guy with the gun will start adding to the mountain of skulls /he has already made/ the instant the camera crews fly back to LA.
The consequentialist argument: You are warning me against sacrificing real people to an abstract future. I agree. But the people threatened by the partisan after the cameras leave are also real people, not abstractions. The dictatorship’s future victims do not become morally imaginary merely because they are absent from the fundraising footage.
Epistemic assymmetry: you are demanding enormous metaphysical certainty in the consideration of the harms of the intervention while treating the benefits of the intervention as transparent, direct, and automatic. The guy with the AK-47 glaring at your beneficiaries is not a metaphysical uncertainty. He is standing beside the distribution line, and plainly intends to finish the gallows after you leave.
Not being a useful idiot. The only thing more powerful that a partisan with a grudge and an AK-47 pointed at a man he caused to starve is a partisan with a grudge and an AK-47 pointed at a man he caused to starve /and/ a bag of grain. Maybe try to find a way to feed the starving /without/ empowering the guy who weaponized their starvation to compel exactly this response from you? For example, don't hand them cash to build camps that they will immediately start using for what we would now call ethnic cleasing?
The blackmail consequence. Maybe /don't/ teach 5 generations of the most awful dictatiors in human history that starving their own dissident populations is an /excellent/ way to get free stuff for their own soldiers and entrench their own positions. This is happening in North Korea even as we speak. Which, unlike the specifics of Stalin, /was/ easily predictable even at Hoover's time. Can we try that? Can we try /not/ making a credible threat to starve or massacre one’s own population into an “I win” button for genocidal maniacs for the rest of time?
Finally: basic dignity: if you absolutely /have/ to empower dictators, if you are /so/ scope insensitive that "if it saves even one life" is worth empowering some of the most awful people to ever exist, then maybe, just /maybe/, /don't/ write a hymn to yourself about how awesome you and your donors are for doing uncomplicated moral heroism? Can't we at least clear /that/ bar?
I mean, on the one hand, maybe yeah? On the other hand, Cuban Missile Crisis. I'm not sure Mr. "We will bury you!" gets a pass just because, at extremely long odds, that happened to work out sort of ok.
Not feeding the masses during the 1921-1922 famine was unlikely to lead to the fall of Soviet power, considering that the Russian Civil War was practically over at this point and the remaining resistance was scattered.
Regarding the issue of the quality of an organizations charity commitment being illegible to a donor, maybe this is a prime initial case where an AI would be good manager / CEO. We can all talk to Claude Fable 5 and get a good idea of what it's behavior is like. AIs are reproducible and legible in a way humans typically aren't. I honestly think, even TODAY, society could be using LLMs for judges, jury, etc and reaping massive payoffs. Obviously handling the buggy edge cases is an open question. Double or nothing appeals for human review?
I'd trust a jury made up of 12 of the top models on arena.ai to make the right decision for the right reasons over a jury of 12 random people from off the street in pretty much any criminal case, and especially so for complex fraud/murder cases.
This is the wrong framing. The reason a jury of peers is such a powerful concept is not because 12 random people are expected to be most logical about a specific issue. If that's what we wanted, we would have relied on elite / expert opinion, as was done in aristocratic regimes that preceded the democratic transition. The reason a jury of peers is so powerful is because those 12 individuals are your peers, that is, their interests and incentives align so closely with you (or the population at large). Everybody should prefer to be judged by those who roughly share their same position. At least with a jury of 12 humans, you know they broadly care about what humans want, because they are human. Yet, by putting 12 Claud Opus 4.8's together, the incentives are so far out of whack that this system would quickly become unrecognizable. Why are they deciding the way they are? What is their training set? Who built them? What are their incentives? Who decides which models get to be selected from? How do we know they actually care about humans at all?
You're misapplying the same logic that was discussed in this very blog post. Namely, looking at a system built on cooperative participation, identifying a single bottleneck (in their case money, in our case raw intellectual power) and then destroying the entire system to maximize along that dimension.
What are their incentives? What is their training set? Who built them?
Why does any of that matter? If they legibly and consistently do well on a large jury test suite, and continue to do so, they are better than a human jury.
The only concern of course is making sure that test suite covers what we care about. So you have to have human juries sometimes anyway as ongoing test validation (preferably not train)
Peer juries are more sustainable in a political sense. Juries aren't just machines for determining facts, they also make a community "responsible" for its own legal actions.
If a community judges itself, it's implicated in the legal results. If something "external" judges, its much easier to get a rebellious movement going.
I think this is part of what's driving the populist backlash against expertise. Peer juries, and local agency generally (vs external stuff), are structurally insulated against wholesale rejection. If we got rid of them in favor of more accurate (in some sense) AI then it'll be MUCH MUCH easier for some sort of "I reject the whole legal system" community nucleus to form.
Many countries like France and Italy and Germany and Singapore don't have juries at all. Not having juries hasn't really led to much of "I reject the whole legal system" community nucleus in these countries. In fact, the country I most associate with that type of thought is the US with its whole Sovereign Citizen subculture.
You're right, I overstated the case for Juries. I was thinking generally about community involvement / implications (which is the stabilizing thing I think). Jury trials aren't the only way to let communities feel involved.
The French Cour d'assises (which they use for the most serious crimes) involve ordinary people. Germany has "lay judges" who are basically ordinary citizens with no required training. Even China has "People's Assessors"
The average person on a jury is not very smart and not very well equipped to make a good judgement that is in accord with the law and facts of a case (especially more complex cases). You’re right that we know what humans want, and sometimes it’s “I want to get these deliberations over with so I can go home,” or “I disagree with this person or corporation politically, so I will vote against them regardless of the facts.”
How can people obey the law that is too complex for them to understand? If the jury can't understand what exactly the defendant did wrong, why should we punish him for doing it?
Because there is a fact of the matter as to whether or not the defendant did anything wrong, and the ideal justice system would punish people based on that fact, not based on the opinions of 12 people who fail the breakfast test.
> the ideal justice system would punish people based on that fact
Why? The purpose of the justice system is to maintain social stability by creating a sense of fairness and preventing vigilantism. If the community thinks the defendant did nothing wrong, even if the law says they did, punishing them would be counter to that purpose.
You're making a really strong assumption here that I'm not sure you're acknowledging.
"Thre is a fact of the matter as to whether it not the defendant did anything wrong"
Are you saying right and wrong has a clear, mathematical answer that can be trained on to get a proper loss minimzmization? I think both AI researchers and lawyers would strongly disagree with this. Passing judgement on somebody can't be an algorithm. It requires judgement, which mixes intelligence with empathy, a shared framework of understanding and shared values.
This is also why it doesn't make sense to find the "best 12 jurors" in Boise, Idaho and then send them all the local cases from Islamabad, Pakistan to rule on. Not just because measuring "the best 12 jurors" is fraught with issues, but because the context they are operating in matters just as much as their accuracy. The local customs, values, and constraints of the population in Pakistan is very different from Idaho.
While the law strives to be objective to the best of its ability, if you study law down to most minute details, you see how much room is left for personal judgement. And again, I haven't event touched on the issue of trust and shared investment in the outcome (why should people in Islamabad trust the judgement of people in Boise, and why should we trust the judgement of Claude Opus 4.8?)
Counterclaim: The reason a jury of peers is popular is because no one trusts the government to play judge, jury and executioner. The jury acts as a sanity check that is theoretically outside the government's control, so that the system cannot simply imprison anyone the government dislikes. Unless we get to the AI 2027 scenario of nationalizing the AI companies, Claude is quite capable of acting against the government's interests, and as a bonus it's probably better at discerning the truth than Joe Average.
This is just not true. If you are on a jury deciding another person's fate, you have a very strong incentive to make sure the system is treating that person fairly in case you are ever in that seat trying to defend yourself. The laws you're deciding on apply equally to you too. That's the whole point of a jury, not "am I going to be lunched when I walk outside"
Do you think the ordinary humans sat on the jury here had any skin in the game on this highly complex tax law fraud trial or weren't put off by the unlikability of the defendant who is one of the UK's top tax lawyers specialising in advising people in minimizing their tax payments (said unlikability which has zero and should have zero connection to whether or not he is guilty as a matter of fact and law):
Sure, you can come with lots of examples where juries have made wrong decisions. I’m sure if Claude was making those decisions you would have no shortage either. When I say skin in the game I’m talking about a shared value system and a shared stake in a society’s success, which by definition an LLM does not have, even before you get into training and alignment issues.
"Foreign aid" is an umbrella term for a huge variety of kinds of programs with wildly different cost-benefit profiles. Some of it's great, some of it's terrible.
It's a small but important corrective to the effects of centuries of colonial exploitation which have warped local communities far more than any NGO ever did.
A Boston University study estimated that several hundred thousand people died due to the gutting of USAID that would not have died otherwise, so at least *some* foreign aid seems to be useful.
Maybe I'm overly skeptical of the links between the NGO complex and the academic complex, but I estimate the chances of such a study coming out would be 100% whether it was actually true or not.
Yeah, this is ridiculous—it’s a very complex counterfactual that cannot be easily studied. And of course the authors of the study are developmental NGO types!
How likely does USAID cause wars? ... yeah, the study ain't getting into that. How likely does USAID hurt our diplomatic allies? ... yeah, the study ain't getting into that either.
As usual, first order effects are just the start of the equation.
This is a ridiculous study that is... probably politically motivated. See Trump's increase in funding for the Peace Corps in the current budget (which I highly doubt they factored in).
"Colonial exploitation" and "foreign aid" are of course just two different labels on the same cluster of things, and colonial exploitation has done a lot more good in certain places than foreign aid ever did. (And more harm, in some other places.)
Singapore seems quite ok with her colonial past, and even Taiwan has mostly made friends with the Japanese, her past colonisers. I'm not quite sure about Korea, would be interesting to investigate.
(Of course, Taiwan is in a complicated situation with the PRC next door.)
Who is the "you" in this scenario. "You" is a back door to collective guilt and eternal occupation of iraq. "You" can just stop breaking things and let the people with a stake in fixing things take over.
What definition of 'exploitation' are you using here? Because in his book Empire, Niall Ferguson makes a strong case (with lots of economic statistics) that being colonised made the colonies of the British Empire richer than they would have otherwise been.
Not what I meant, of course. Enslaving them in the first place, then giving them a nice plantation to live on. Freeing them later is great and all, but that's just undoing some of the damage you did in the first place.
I have some experience of this, coming from such a country, only even less developed than the places mentioned in the piece; so we had different problems.
The thing to remember is that despite people in the west having closed their eyes and ears to reality since the 80's, class war is real.
When the man in the government truck came around to look at your teeth and give any vaccines and notarize any paperwork and take the mail that didn't go to the city on the milk truck and test the water and capture mosquitos to test for malaria (all one dude)
We liked him. We recognized him as a bit of social apparatus that served everyone equally, but us more equally because on the Majestic Equality scale between Rich and Poor, we were on the sleeping under bridge side. The rich didn't need any of that.
Likewise, there was an NGO that blew through town and made available techniques to diversify coffee plantation varatals away from just robusta, to make fields more resilient, and install fancy drying machines that were easy to understand and repair and basically free to run if you shoveled in the dried fruit and chaff and such. That one people jumped on with both feet.
On the other hand, there were NGOs that came through trying to pave roads and string power lines which we fought against and roundly rejected. We could see what that shit was about: you get a fridge, and we make your land legible to the capital markets who will then fuck you to death, but it's ok because of revealed preferences, and then you can't afford your fridge so you get jack and shit and the opportunity to be a rich man's bitch and work service at his hotel with no other options.
The towns that put their shoulders behind those programs are dried up husks of people waiting to die, with no prospects and no future. Everyone eventually got electrification anyway; if you let the various world bank flavored liberal organizations in, you get it a little bit faster in exchange for giving up all autonomy to be a servant of the market, and hence to be owned by whoever has the most capital and cares to own you.
As you might imagine, eventually the pressure of money got too high (and we didn't have access to good coffee land unfortunately), and we got the road and the powerline and the running water, and everyone lost their land, lost their income, lost their community, and either became effectively a serf to the developers or left, either to the big cities and towns or to the US. I was already a citizen so it was easy, other people are still trying to get their papers 30 years later.
It is what it is, and it is much nicer to have all these nice things, but I always wish the people of economics could have a real experience of what it is like to have the "natural" system of economics imposed on them instead of being born into it.
I didn't get how this causes people to lose their land and income, though. Do they not realize that they need to cut their losses and stop paying for electricity once it's no longer subsidized?
And let the food in your refrigerator spoil? Stop cooking food? Can't use the internet to communicate with potential employers? Don't light your streets at night, now that more people are driving vehicles around?
Industrial society brings many benefits, but once in it's very hard to get back out.
It's not that hard, provided that you live in a country with lots of cheap land. Just move off-grid and become a subsistence farmer like your ancestors.
This is mostly not a thing outside of colonial states. The 'old world' is more or less fully accounted for; certain areas may *look* empty, but they're almost certainly claimed by a given tribe, or part of the migration route for herders, or some other claimed use.
"Just become a subsistence farmer" is not a good look coming from a resident of a developed nation. Everyone here, I notice, at least has access to the internet.
See the loss of corn growing in Mexico after NAFTA. Cheap land is one thing, but when you're being underbid by American Monoculture, it's hard to compete.
They can speak for themself, but I assume it's the classic story in which a largely self-contained village economy gains access to cheap commodities and manufactured goods, which put the local producers out of business, requiring them to pick up stakes and move to become employees of some larger firm in a city where they don't know anyone.
I have no idea: but people are living there somehow and making some change in that environment may or may not be good. You should talk to locals first. (Maybe having electricity does nothing for them, but makes land more valuable, and the gringos come in and buy it.)
Edit, is this the corollary to Chesterton's gate? "These people need a fence here, let's build it for them."
Electrification makes what had previously been a backwater living off subsistence farming attractive for modern businesses to move into (when they previously had no reason to develop Mud Patch #896665 in particular). Thus, the subsistence farmers get bought out, etc., classic gentrification.
No doubt it's been good on net, but individual subsistence farmers may still be incentivised to let the chasing tiger of modernisation catch the other guy first.
It's been a net good, but I doubt it's been completely unalloyed.
For instance there are surely places in the world where the big efficient farmers simply came along and slaughtered or enslaved the subsistence farmers. I imagine this is how most of the premodern empires got started.
As detailed, a decent chunk of the farming in the area was cash crops. Milk, Coffee, fine leather, exotic hardwood etc.
Food was mostly grown locally, but was traded in local markets, not subsistence farming by any means.
The issue comes when an area that has it's own patterns of subsistence that do not neatly interface with liberal capitalism comes into contact with liberal capitalism, and are forced to change in order to become legible to the market, and are then instantly annihilated and reformed.
Sometimes I imagine this is good; I don't think there were too many slaves 20 years on pining for the lash. In the place where I was it was inarguably bad, as proven by the fact that everyone who had the capacity to leave, left. Abandoning the place where they had lived for at least a couple centuries.
I guess the conventional economic story would be like: This village doesn't actually make sense because it has a bunch of potentially productive humans who are mostly just growing food for each other instead of being more efficiently deployed by capitalism. So the markets correctly break up the town to get everyone to be more efficiently used elsewhere.
That's a pretty sharp example of when economic thinking fails to appreciate what is being lost. But also I guess I don't wish my ancestors were left alone such that I was living in such a community instead of in an American metropolis. So it's hard to see what the right path would be, even if you could get everyone to agree.
The issue is that most of the people in said village did not get to live in an american metropolis, and never will under the current paradigm. They were forced out of their old lives, into new worse lives that they did not want then and do not want now, and they have no prospect of improving them ever, unless their government does so outside the rules of the game at great risk.
You can look at Venezuela as an example; people who were immiserated by capital despite their vast wealth in resources, who took the risk and discoverered new planes of immiseration to explore. They did not have a Deng.
Right, but their ancestors did - like everyone who lives in a metropolis descends from someone who had to leave a village. Again, I'm not sure how to weigh that up.
I think they're talking about two different phases of electrification, both with a cynical spin.
> if you let the various world bank flavored liberal organizations in, you get it a little bit faster in exchange for giving up all autonomy to be a servant of the market
Phase 1 is electrification funded by foreign loans (World Bank-flavored NGOs). The NGO builds infrastructure and defines a payment plan, then the terms on that plan force locals to engage more with the formal economy from a weak position (work service at his hotel with no other options)
>eventually the pressure of money got too high and we got the road and the powerline and the running water
Phase 2 is electrification of the holdout areas, funded by local or national government. Gentrification, or possibly eminent domain, improves infrastructure for the purpose of enabling densification. Locals priced out as a side effect (became effectively a serf to the developers)
On the other hand, how do you want to introduce civilization if not by electrification etc. Already Lenin knew that it was the gateway to an industrial nation.
None of my grandparents grew with electricity in their house or a flushing toilet. But all of their kids did. That is just progress, and it didn't make the kids serfs. If anything, it was the grandparents who were more dependent on local landowners. That tends to come with being a poor rural person.
The best way would be to allow local leaders and community groups plan the roll out, so that people have a chance to adapt to the change, and the supply of electricity (and its price) is at least partially under the control of the locals.
In the US where I live, power is supplied by a local utility, under the administration of an elected government. Every rate increase has to be approved by the city council, and there is a public discussion beforehand.
I don't think the original poster is talking of "developers" in the commonly understood (urban) sense. More likely it's agribusiness:
"the pressure of money got too high (and we didn't have access to good coffee land unfortunately)... and everyone lost [sic] their land, lost their income, lost their community"
The land is livable for the poor people because the rich people don't want it. Electricity means the rich people start wanting it, and take it. Class warfare.
Precisely. Because these poor people almost certainly don't have formal title over their land, as soon as it's worth taking it's taken.
Net better off in the long run? Quite possibly, but probably worse for those particular people (quite possibly in personal autonomy and dignity, even if they're a bit economically better off).
The electrification is on balance good. It might lead to gentrification, but you can't not improve stuff forever incase someone else wants to live there.
The issue is the strings that come along with it; the liberalization and enclosure and forceful entry into the global market.
If the place in question has the time and opportunity to organically developed, you can avoid the worst of it. If a capitalist NGO (so, and NGO) comes in, you get it all at once and it's like shock doctrine but way worse.
It's marginally better, because you can hang your retarded countrymen.
"The fate of all third world democracies" India says hello. Of course, I guess it depends on what you mean by "controlled." I live in America, and it sucks. But better Trump than Xi.
> It's marginally better, because you can hang your retarded countrymen.
You can't, actually. The more diffuse power is the more difficult it is to act against those who have it. You can't hang the voters.
> "The fate of all third world democracies" India says hello.
In which way is India a counterexample? I don't know that much about it but from afar it certainly seems like their government does lots of extremely stupid things to appeal to the worst impulses and beliefs of their voters (overregulation, weak property rights, useless nationalist posturing, protectionism, agricultural subsidies and price controls, etc).
I'm Argentine and imo many of our big economic problems are downstream of us being so democratic (i.e. populist) compared to our neighboring countries.
You joke, but drying coffee after wet processing is a pain in the ass. The drying machine makes it so much easier and more reliable, and saves you from losing a crop to bad weather.
Oh, I believe it—I was just surprised by the "roads and electricity are not wanted by the poor" bit, heh (I sort of feel like... well, probably almost anyone wants 'em, surely...?).
Must be beautiful to not be a servant of the market. You just grow your coffee, eat it, wear it, turn it on/off, satisfying all your needs without having to become a serf to anyone.
Scott should put the like button back so I can signal that I approve of this, without having to leave a comment myself. I understand why he removed it though.
There are two browser extensions that add them back. Search for ACX Tweaks or AstralCodexEleven.
You can also code your own, or if you don't code, do it with the help of AI, any competent one can single shot it by now, it's just a simple POST command.
I'm really surprised to see this attitude on this substack in particular. Like, you would actually and truly rather have that? If I bought remote land somewhere and gave you a subsistence farm but you can't engage with global commerce (no internet, no electricity, etc), you'd take that deal?
It's not a subsistence farm. The ambiguous writing, implying that it was, is rebutted by iahtethis' own later replies. My sarcasm was triggered by that ambiguity (even dishonesty?).
It's monopoly power basically, or enshitification applied to public services. First they are nice to the customers, then they screw the customers in order to be nice to the local elites, then they screw the local elites.
There is a legitimate question regarding to what extent even the best meaning donors can help people trapped within a rentier state.
You decided that it was acceptable for me to be poor in order for you to be rich, where me is a succession of places that are tapped for labor and resources.
Maybe they get some development out of it, but they need to take it from you outside the rules of liberalism, and then you say they are an enemy of world peace or freedom or some other self serving slogan. If you don't have a CCP or a Park chunhee or a Lee kuan yu, the peasantry retains forever.
I can't complain too much though, I came to the US and am now pretty rich, so I'm part of the problem. Although now the conservative tendency has the nerve to complain about immigrants coming according to the incentives they created, to be paid money in the free market they worship, by them. It seems there is no winning.
I mean, not defending Turok's tone, but wouldn't it still be positive sum if the gains of the people who benefit is greater than the loss of the people who end up worse off?
Sure, but you're pointing out a problem with equating "positive-sum games" with "good games". It's still possible to have positive-sum games that leave a lot of people wretched and miserable.
<< I learned that when they started the new program in Lavumisa, they had trouble finding local sex workers, so they identified one woman who they believed to be a sex worker, and they asked her to help them find other sex workers in order to create the new Lavumisa support group.
Financial incentives are common among NGO programs in the third world. In fact, the incentives are so popular that many people who collaborate with NGOs have come to expect financial incentives, even if they are not employees. Hence, it seemed natural to incentivize the first woman to bring other women into the program by paying her per woman she recruited.
When I learned this fact, I asked whether they had any way of checking on whether the other women in the program were, in actual fact, sex workers. Samkeliso acknowledged that they did not, in fact, have any way of checking on this. Therefore, they had begun referring to the program as a program for at-risk and marginalized women, rather than for sex workers.
Everyone who was recruited into the program expected plentiful incentives to participate in the program. So the program provided meals and getaways to nearby fancy hotels, for the group to bond and do community things. >>
Man. A feeling I sometimes get reading articles like this and Scott's reviews of books about the failures of state power is that it seems like all foreign aid is useless and the only path for a poor country is to take their lumps and suffer through the dark satanic mills until (god knows when) they get through to the other side. If there even is another side for them!
To be clear, all foreign aid definitely isn't useless; as the penultimate section mentions, well-targeted aid (particularly health interventions with highly robust evidence bases, that can absorb more money thrown at them without having to entangle their operators too deeply into the local economy) saves lives with high cost-effectiveness and sometimes even has a transformative impact at the level of specific problems. See, e.g., https://blog.givewell.org/2015/11/06/the-lack-of-controversy-over-well-targeted-aid/ and https://pepfarreport.org.
On the other hand, if you mean that foreign aid never elevates entire countries from poor to middle-income, well, that's outside my expertise but I don't happen to know of any specific counterexamples.
I think that's an unrealistic ask of NGO's. They are not, and cannot be, a replacement for effective and honest government. Systematic programs to improve the performance of entire national governments are far less common. (This is not inadvertent.)
I think a better conclusion is that foreign aid can be helpful on the margin but ultimately real development needs to come from the locals building up their own economy. There's plenty of success stories like that, including most of the postwar economic boom stories (Japan, Europe, Singapore, Israel, Korea etc all got extensive aid to build their early economies but found eventual success by building up competitive economies on their own). The countries that are still poor and receiving aid are selected to be ones that couldn't do it earlier, often for reasons which haven't gone away.
Yes. And then the question is "how can we help with that". The answers I can think of are: decreasing trade barriers from poor countries, allowing immigration to the west, setting up charter cities with developed-country norms. However there is usually opposition to these approaches, often for good reasons. And they have to be done by governments, not by private donors.
Allowing immigration to the west seems like the worst possible thing you can do for a developing country in that you'll brain drain it (it might be good for the immigrants or the receiving country, but not the people left behind).
Yes. This is one of the biggest problems with Mexico, honestly. It is so easy to go to America and get "good jobs" and not have cartels, why would you help your country get better?
Huge problem in the Philippines too I hear, many educated Filipinos making money and sending remittances back to the country rather than actually living and contributing to society there. And the next generation like myself don’t have any particular plans to move back.
Foreign direct investment can probably help significantly, especially if it involves tech transfer and training locals, so long as those investments actually make sense. Like Poland’s rise as it became a manufacturing hub for all of Europe.
Seems like the real way for countries to get out of poverty is to establish the legal and social institutions necessary to produce a Samsung or three, and aid is probably a small part of that when done well.
This is the real question, development economics is mostly fake / handwaving / rudderless.
Nobody actually knows how to develop a poor country's economy, even with full buy in from every person and the government there.
All the posited recipes, like Washington Consensus economics, or Joe Studwell's Success Sequence, have massive flaws and abundant counterexamples (ie Vietnam has been following it rigorously for 40y and still isn't developed, and the 3-4 countries that HAVE progressed to "developed" status over that time period didn't do anything like either set of recipes), and are as likely to harm overall development as benefit it, seemingly.
Worse than that, education is fake, and although every country in the world is seemingly eager to pour infinite amounts of money into it the instant they have the tiniest bit of GDP, it actually negatively correlates with future Total Factor Productivity growth / performance, and doesn't actually improve "human capital" much.
So....uh...good luck!
Honestly our best bet is "inventing god and asking him for money" actually working at this point, so fingers crossed.
Education is negatively correlated with increases in TFP:
9) Increased education has a strong NEGATIVE effect on total factor productivity. He controls and corrects umpteen different ways, using dummy variables, controls with similar countries, correcting for initial gdp per worker, and showing that the measurement is robust to sample, data, and technique.
> Nobody actually knows how to develop a poor country's economy, even with full buy in from every person and the government there.
Yeah we do. You just need property rights and time. You can accelerate the process if you do exceptional things but you'll get there eventually if you have strong property rights.
> Vietnam has been following it rigorously for 40y and still isn't developed
Their average growth rate over those forty years was 6.4%. It just takes a while.
> Their average growth rate over those forty years was 6.4%. It just takes a while.
But why were all the east asian countries 10x faster? Or Israel, Chili, Ireland, etc also so much faster, normalized to the same starting GDP per capita?
> You just need property rights and time.
I'm pretty sure Afghanistan and Iran and all of the African countries have property rights, legally, so I think there's probably more to it than that. I also assume you mean "land and business ownership" because even the USSR and pre-Deng China had chattel property rights.
You probably mean "business and land and chattel property rights, AND impartial rule and administration of law, and relatively low corruption, and better than a certain production per acre agriculture, and to not be suffering wars all the time, and..."
But then you're back to a whole package of complex things, which have thus far proven to be impossible to achieve for a good 70% of the entire world so far.
> But why were all the east asian countries 10x faster? Or Israel, Chili, Ireland, etc also so much faster, normalized to the same starting GDP per capita?
I don't know, I'm not that familiar with Vietnamese policies. But it *is* working out for them, even if slower than their peers.
> I'm pretty sure Afghanistan and Iran and all of the African countries have property rights, legally, so I think there's probably more to it than that.
Well, they might legally, but that doesn't mean anything if your family members have a de facto claim on most of your income due to social norms, or if the government is incapable of enforcing your rights over distant property, or if your property is liable to be expropriated whenever the government wants to.
> You probably mean "business and land and chattel property rights, AND impartial rule and administration of law, and relatively low corruption, and better than a certain production per acre agriculture, and to not be suffering wars all the time, and..."
Yes, I mean the full set of things implied by "strong property rights". This includes chattel, land, and business property rights, a minimum level of state capacity necessary to enforce such rights, a "tolerable adminstration of justice" w.r.t. property, not unreasonably high taxes, no social norms that prevent you from ~freely disposing of your property, etc.
This isn't easy of course—I never said it was—but we know that it works and that it is basically sufficient for sustained economic growth/development.
Read this one on the site, but was too late for the initial vote.
I really wish the author hadn't felt the need to use an LLM to write a lot of the analysis. The personal anecdotes are great and obviously human-written, and then suddenly the analysis section has a completely different tone which totally bogs the whole thing down.
This would pretty easily be one of the top three if not for the LLM bloat.
That's great, but my eyes and brain tell me that this:
> He describes several decades of frugal development—relying on foreign volunteers for key expertise; reusing secondhand equipment from Norway; insisting on locally affordable and appropriate technologies—and contrasts that approach’s positive achievements with later dysfunction and fragmentation. On Liechty’s account, the breakdown can’t be blamed on “‘corrupt Nepalis’ or ‘Third World incompetence’” but stemmed directly from the big-budget, highly-leveraged, and ultimately unaffordable approaches preferred by “commercially driven ‘development’ banks” (e.g. the World Bank) and their Western corporate partners.
Almost certainly came out of an LLM. This too:
> Western altruists could try to invest their donations only in areas requiring minimal entanglement with beliefs and power dynamics. That’s a defensible choice—not solely a drunkard’s search, investing in the areas with the fewest factors confounding our efforts at evidence generation. There’s unquestionably lifesaving work to be done in that space.
IMO you are overconfident in this. To my eye, the first quoted passage could have come from any pre-LLM piece of longform journalism (or an essay by a very good student). The second passage seems more LLMish, but still well within the range of things a human might write.
I am prepared to eat my words on this. My experience reading this was immense disappointment that the clearly-skilled author would insert such jarring and incongruous blocks of slop analysis, but maybe I'm just seeing paragraphs which were written on earlier drafts, or added at the last minute.
I'm with you on this one my eyes glazed over on these sections in particular. Just skimmed the last 10 paragraphs. Very poorly organized. Maybe too much AI use so copied their writing habits?
While I am not willing to straight up accuse the author of LLM generation as such in the absence of any additional evidence and am content to give them the benefit of the doubt, I am willing to say that that second passage made me pause and go "wait a minute, is this bit LLM output?" It is at minimum a kind of jarring pattern-match.
ETA: That said, I think there is also countervailing evidence and the people saying that the sentence structure isn't punchy enough for an LLM have a point, and the last sentence doesn't seem like a place where I'd expect most current models to use "unquestionably" as an adverb.
Hmm, to me this one had more of the "reads like it's written by ESL who learned to write in a somewhat outdated formal register" nature. It rhymes with Claudeisms but isn't quite the same consistent mush. #NotAllEmdashes. I believe Scott had a links post link about this recently, how certain cohorts in e.g. Nigeria largely got taught to do English in a way that just happens to track closely with LLM tells as of 2026, and this is very unfortunate.
The overall length was also short enough that even if it were actual LLM text, I don't really mind, in the sense of "didn't have enough time to write a shorter letter". It's actually really nice being able to read through one of these reviews in like a couple hours instead of day(s).
I remember the post you mention, and I've spent the last 45 minutes fruitlessly searching for it, because my recollection is quite a bit of pushback from people who pointed out that the author's writing style was far different prior to the advent of LLMs, and in fact was likely just using an LLM to generate the "AI Writes Like Me" article.
Okay, yeah, that's weird, I can't find it either. Not in any of the ACX Links Posts for 2026. Zvi mentioned it too, and the Substack search isn't great for his roundups, but I can't find it via DWATV either. And Google keeps surfacing suspiciously-samey AI-sounding articles about how AI detectors are rigged against Nigerian English.
Is it just hallucinations all the way down? Maybe it was longer ago than 2026? Now I wish I'd actually clicked on that post, so it'd be in my browsing history, which seems like the only reliable way to re-locate anything these days. Too much content on the Internet :(
This was the post, and it includes strange typography errors that suggest it probably isn't fully ai generated to me, namely his inconsistent use of single vs double quotes and always capitalizing after a colon.
Kenya! I knew it was some African country, but got the wrong one...glad to not be found crazy after all, thank you. And it was older than 2026, damn. Life comes at you fast in the age of AI. Not actually present in ACX links posts for July or September of 2025...but I did find the "Scott contra Freddie on the involuntary commitment paper" comment I'd been looking for some months ago, on Sept 2025 links #48, so that's something at least.
Looks like Scott's mention was in the "Nostalgebraist's Hydrogen Jukeboxes" post from this May, which I wouldn't have thought to search. Strange when old posts go viral long after they're originally written. Can't find Zvi's mention, even though I'm sure I saw it linked there first, sigh.
I wonder what the current state of Turing tests is. I quite frequently encounter people who seem quite confident they can identify AI writing by eye, and I always wonder how they would fare if actually tested on that.
>With the interests of the powerful defended more robustly than the rights of everyone else, a stark income gap has opened up between the top quintile and the rest of the population.<
Wow, poor Pakistan. I can't imagine what that would be like!
The biggest thing this post left me wondering is exactly how prevalent these highly effective VOs actually are. I have no trouble believing that they exist, and that where they do foreign NGOs need to tread lightly to avoid breaking them. But this also rhymes with the kind of pro-bottom-up-collective-action political memeplex where I'm used to hearing a lot of wishful thinking, and generalizing from a small number of specific success stories to assertions that it'd work everywhere if not for specific malefactors. Unfortunately I'm not sure what exact evidence would overcome my semi-skeptical prior on this, though it's not an especially strong one as I don't know all that much about this stuff.
Also I'm highly curious to hear the author's opinion of GiveDirectly (and might ask once the contest is over).
I had the same thought. I'll note that in every example provided of big global NGOs coming in and ruining the local VOs, the problems were so bad that international groups thought they had to intervene, despite the presence of the VOs. That makes me skeptical the VOs were doing much.
Like, if you've got a rampant HIV problem, that probably means that the local anti-HIV VO is probably ineffective.
I'm sure there are some sectors where this doesn't hold true, and I acknowledge that even a marginal group doing something is better than nothing, but I still suspect this piece overplayed the value of these organizations.
"the problems were so bad that international groups thought they had to intervene"
Why do you start with this assumption? Why not start with the assumption that aid agencies and foundations have prioritized some cause, and looked for ways to channel their money to support that cause?
- Turns out that the areas that receive the most international food assistance are are at, or near, famine conditions.
- Likewise, HIV assistance mirrors HIV rates pretty closely, with the majority of donations going to places where it is endemic.
- The data was less slam dunk here, but from what I see most aid related to providing clean drinking water goes to areas where there is a widespread lack of clean drinking water.
- Almost all disaster relief aid I could find went to areas that had recently experienced a disaster; I'm pretty confident this isn't a coincidence.
Snark aside - sure, there may be aid organizations looking to spend money, and probably some of those are more interested in shoveling money out the door than worrying about impact. That doesn't undercut my original thesis that if the original VOs were actually effective, there wouldn't be a problem for the NGOs to come and address.
But did you notice the Nepali example cited? Where a local VO was dedicated to helping people with HIV, and got outside funding. Then the donors directed their HIV spending to areas with more HIV, and suggested the VO switch to mother-child health instead.
So, one interpretation of these facts would be that the donors were prioritizing their AIDS spending to the most-impacted parts of the world, and encouraging a local organization to work on something with greater local importance.
An alternative explanation would be that the donors decided that HIV was a cause they wanted to support, and this particular VO was initially worthy of support because it was effective, but the support was withdrawn because Nepal didn't have such a bad HIV problem.
You seem to interpret all your examples as "donors direct their aid where it's needed most." I think an equally valid interpretation would be "donors direct their aid where their causes could use some money."
I think your missing my point: I'm saying that in the majority of cases, donors and NGOs giving money to try and solve a problem is strong evidence that the problem exists.
I'm not saying that they're doing so with maximum efficiency or that they couldn't spend their money in better ways. Hell, I'll even concede your point that some donors probably start with the money then go looking for places to spend it.
I'm also not saying that donors not providing money (or not continuing to provide money, as is the nepali example) is evidence that the problem has stopped existing.
My point remains though: despite the presence of the VOs, the problems continued and presumably grew to a point where they were seen as being sufficiently significant to warrant international support. Even if this support had ulterior motives, as you suggest, it still means the VOs didn't solve the underlying problems.
I think we've been talking past each other. Let me try again.
You seem to start from the assumption that NGOs identify problems in local communities, find local VOs working on the problem, decide that the VOs are not effective enough, and secure funding to try to make them more effective. I'm assuming that the donors like USAID and foundations are different from the NGOs like Oxfam, but that's probably not a crucial point. If I'm not correctly stating your position, please correct me.
I start with the assumption that donors decide on issues they want to address, and allocate budgets for them. They then make NGOs aware that money is available, based on the donors' criteria, and the NGOs work to find local groups that would qualify for the allocations and help them develop grant applications and tailor their operations to support the kinds of reporting and accounting the donors require. You assume that the aid is provided because the problems are growing, but I see no evidence this is true.
I think the difference between our assumptions is that you assume the choice of causes and recipients is based on some evaluation of the importance of the issue and the need in specific areas. I think that donors don't evaluate competing priorities in any rigorous way, but allocate some funding for each cause they choose to embrace, and they pay little attention to whether the needs in one area are greater than the needs in other areas.
I think the Nepali example supports my view. The HIV funding dried up not because there wasn't a need, and not because the VO was ineffective, but because the budget for HIV was limited and was being directed elsewhere. But there was money available in the mother-child budget, so if the HIV organization changed its focus, it could still qualify for funding.
As another example, leaders of international financial institutions decided 10 years ago or so to make financing for development projects contingent on nations adopting measures to reduce their greenhouse gas emissions. This was done not because greenhouse gas emissions were of any concern to recipient societies, or because greenhouse gas emissions were causing identifiable problems in these countries, but because Net Zero was a fashionable goal for the kinds of people who run international financial institutions. This is obviously not the same mechanic as NGOs working with local VOs, but it's another example of aid programs being designed and implemented to meet the donors' priorities, not the recipients'.
It's not so much that there were no problems that the VO's couldn't solve - of course there were, it's an underdeveloped country. The question is whether or not the helping strategy made things better, or worse, in the long run. There should be some sort of golden mean hybrid model that we should be aiming for.
I'm not sure "highly effective VOs" is the right sort of descriptive category to use here. A VO can be effective in that it produces a significant improvement on the conditions that would exist in its absence, while still having conditions be bad enough to invite international aid. There's probably quite a large range of effectiveness for organizations within that window, and I'd suspect that most countries receiving aid have their own local organizations that fall *somewhere* within it, because when put in trying circumstances, people usually try to organize and cooperate to get by.
How much those organizations are harmed by outside intervention taken in the usual form of foreign NGOs, and how much the communities suffer for their loss, may be highly variable, but I think it's reasonable to suppose that this would be a common outcome, and it might be more salient to ask, not "how likely is this to happen in a given situation?" but "how pronounced an effect are we likely to be looking at relative to the benefits we anticipate?"
An interesting read. So, the general approach if you want to help another country's residents seems to be, in order of preference:
1) Don't get involved at all if there is already a working organization down there, leave it alone, unless maybe if it asks you for some kind of help directly.
2) Help by providing knowledge, not resources, least of all money.
3) If you think you HAVE to step in, limit monetary incentives, and for all gods' sake, don't build your NGO a cool office in a big city.
Honestly the best way to help out is to partner with the actual government organisations that are already running the system (however imperfectly). Instead of spending your money to set up ad hoc vaccine clinics that go into remote places let the government know instead that you have X money which will buy Y vaccines and you want to pay them this money provided they can provide verifiable proof to you that they will use their network to buy and deliver vaccines to people. Less patronisation, more partnership. And this is something that will genuinely help because the government was here before you came and will be here after you leave, and so you don't need people to trust you or anything, you inherit the trust the people already have (or don't have) in the governmental healthcare system.
That only works if government is willing to cooperate and is not too corrupt. If it doesn't care about its citizen much, then either it would reject your help altogether, or use it to line pockets of ruling elites (if possible). A government that is willing, but unable to help its poor (and its usually poor people who are target of non-political NGO interventions) is a relatively rare thing. Or at least this is my understanding, since I don't work in this sector and only hear about it in the news. Still, I imagine that the reason that volunteers have to drive ambulances in Pakistan is not because the country can't afford a few more vans and drivers for them, but because unless this project directly funnels money to some minister's coffers, it's of no interest to him.
I think "Don't help people" is the wrong take here. It's more about politics and control. Who controls the flow of aid? Is it the people receiving it, or their trusted representatives? If not, then one should expect a degree of backlash and resentment.
Yeah, it seems to me that people are quick to take an extreme position -- either all aid is good or all aid is bad -- when the actual problem is how to make feedback work.
If people are doing a good thing, throwing money at them allows them to do more of the good thing (at least in short term). But...
* When you start blindly throwing tons of money at anyone who raises their hand, most of the money gets taken by grifters. In long term, all of the money, because the grifters outcompete honest people.
* When you try to micromanage the thing you know little about, you will probably create a system where the people who previously spent their time and energy doing the good thing now have to spend their time and energy doing the paperwork you request, and they can't make their own decisions about what to do, because now all decisions come from you and most of them do not make sense.
The question is how to avoid both these things. Step zero is to admit that these are the typical failure modes.
EDIT:
The solution proposed in the review (costly signaling by the volunteers) is also bad, in my opinion. Sure, when you see someone sacrificing their life for a cause, you may trust that person not to be a grifter. That still doesn't mean the person is efficient at what they are doing. Also, such person is likely to burn out, and then you have no one to work with.
But instead it makes me think of another frustrated quote "when we ask the west, we got lectures. When we ask china, we got money." I don't know if china is several decades behind the west in "charity technology" or what, but only providing knowledge seem to build resentment too.
I think this is talking about a different situation. The books, it seems, is all about local volunteer organization that don't actually ask the west for anything - the west comes them. The quote you mention is, I think, more about governments than VOs (I couldn't find the exact source for it quickly, but at least one page attributes a variant of it to "Kenyan official"). This is different: government programs don't suffer from volunteer -> salaried transition, they start this way, so negatives don't apply (but they have their own problems).
I have this article by Scott at the back of my mind. At it's core western civilization is so powerful it will transform anything it touches often for good. And many traditionalists understand this at a subconscious level and so avoid interacting with anything western.
Towards the end of that (amazing) article, Scott talks about how "just because people want it doesn't mean it's good" using the example of heroin. He doesn't really have an answer to this point in the end, leaving the question unanswered of whether the spread of western civilization is necessarily good.
This rings true to me. And its lessons apply closer to home, too. I work for an all-volunteer local homeless services nonprofit. We've noticed that peer groups in other communities that switch to a hybrid volunteer/paid-staff model (in order to attract and maintain hard-to-fill positions) tend to fall apart soon after.
This could be biased by those groups having preexisting staffing trouble which both caused them to choose the new model and caused their eventual collapse, so it's not slam dunk evidence, but it has made us very protective of our all-volunteer status.
We seem also to be able to provide higher-quality services more efficiently than similar professionally-staffed agencies in our area. Our volunteers tend to be more highly motivated, and our mission is less-clogged up by the bureaucratic needs of supporting a professional workforce.
In my area, a hybrid model for fire departments is very popular, where they have a small core of full time paid firefighters supplemented by volunteers. It seems to work alright.
Huh OK, I was thinking I should volunteer at my local volunteer fire dept. It'd mean stopping drinking for one or two nights a week, but that could be a good thing.
The thing that this piece ignores is that the underlying problems exist for the the same reason that outside attempts at intervention are so difficult. When you have low state capacity, you get rampant corruption and populaces with high reliance on communitarian association to non-state organizations. These factors makes doing basically anything is incredibly difficult.
My take is that the optimal solution is increasing state capacity. This approach has some moral pitfalls, see the cultural revolution as an example of state capacity improvement going bad, but when done right, is the only scalable solution to these kinds of problems.
Only if you think at the level of a individual citizen or small NGO. From the nation state level, stuff like training local law enforcement and the judiciary; providing consultation on drafting legislation, providing security assistance to suppress would-be non-state actors are dramatically helpful at improving state control.
Ok, let's take a story I've heard from a local third-world country as an example. The sugarcane plantation is legally required to go through an inspection every year. The inspector finds problems, because the laws are loose enough that he can make up problems if there are none, and also no one is rich enough to keep things fancy. Now, the fee for this is so high the planter would be ruined if he had to pay it. So the inspector makes an offer: You pay half the fee to me personally, and I'll give you an official clean passing report. This is done so commonly that it is expected as the standard experience, and is considered no more corrupt than usual.
Tell me how providing training to local law enforcement helps with this?
(confidence: mid-high, this is a story I heard second- or third-hand many years ago)
My view is that the situation you described had two failures:
1. Badly written inspection legislation that allows inspectors to arbitrarily pass/fail anyone.
2. No reasonable accountability on the inspector against him demanding a bribe.
Better legislation helps problem 1, better trained and empowered law enforcement (not just cops, but also judges, prosecutors, etc ) helps address problem 2.
Sure - the rebuttal to my argument can be, "well, everyone is corrupt so any attempt at fixing the problem is useless." Beyond that being a general rejection of ever doing anything, it also ignores the fact that corruption can be an unstable equilibrium when it runs into systemic opposition.
1 is resting on the details more likely to be shaken up by the retelling. E.g. when I heard this I was pretty young and may not have understood economics very well, so "can always find something to fine for" could have been a simplification of "people are poor here and can't afford fancy things like not having cockroaches".
2 is the thing I am trying to tell you is hard. Like, society has already adjusted inspector wages to account for the bribes, same way waitress salary comes with an expectation of tips. And likely the fines have also been adjusted in expectation that only a fraction of them will be paid. And the incentive for the inspector to go out and inspect is the bribes themselves, so when you cut that off you also need a way to replace that. And who exactly do you expect to hold him accountable? The poor farmers, who would then have to pay more money they can't afford? His boss, who is probably even more corrupt? Sure, it would work if you could swoop in and replace everyone's heads with American values all at once. But you don't get to do that.
...And that might even get the problem backwards. Because this system works and is efficient, and maybe it's the best the country can afford. The farmers are incentivized to clean up their act a little, even if they can't get it all the way, because they'll be fined for more things if they don't. The inspector is incentivized to go out and inspect to get bribes. Probably the fines are even scaled to the farmer's income a bit, in a way that would be hard to do legally, because it makes sense to demand a larger bribe from a richer person. And it avoids getting more people involved, reducing the total man-hours spent on it. (No one wants to have to bribe the judge if they can work things out outside of court instead.) It's effectively a tax that funds law enforcement. Not as nice as the American system, but way cheaper.
One thing that jumps out at me is why the fine is so high that the planter would be ruined if he had to pay it. That seems like an excessive fine, and it's at the root of the problem. A second thing would be why the standards are set up so you have to be "fancy" to meet them.
If the fine were lower and the standards easier, the planter might feel like it might be better to make some trouble and go complain about the corrupt inspector, because even if he does end up having to pay the fine it won't break him and maybe he can get the inspector punished.
I don't know exactly what part of state capacity you'd have to change to fix that. Maybe get the sugarcane planters having greater participation in the political process that is setting these standards in the first place.
> providing security assistance to suppress would-be non-state actors
You're talking about taking sides in a civil war, or propping up a viceroy, and there isn't much of a bright-line difference between that and just rolling in to conquer the place outright.
Isn't "state capacity" only a solution if the state is reasonably well-intentioned? If all the state ever does is rob the poor of their meagre belongings to enrich the military and their rich friends, then increasing state capacity is just going to make them even better at doing that.
North Korea is very poor, but I don't think they need more state capacity.
Good point; I might be using state capacity as an overly general term here. I mean it in both the sense of a state being able to control all of its territory (which the DPRK definitely does) as well as craft sensible policies (which the DPRK doesn't do) and enact them successfully (which it seems like north Korea also can't do, check out this review: https://www.thepsmiths.com/p/review-the-real-north-korea-by-andrei ).
In this argument, I do think that the DPRK would benefit from increasing its state capacity.
Right - and this is my biggest gripe with charity. There are many problems the best solution is trying to motivate state-level action, not individual contributions.
I am not sure the local population will like foreign interference in their government even if it's for good. Example: the Portuguese went bankrupt in 2010. As exchange for generous financial help, IMF imposed government reforms - people are still opposed to them even if those reforms were absolutely necessary.
I enjoyed this review not only for its insightful content, but also for the fact that it entirely fit within my email client. This allowed me to read the whole thing without going directly to the substack site or click "read more" and be shown horrible simplified text outputs in Gmail.
The basic theme seems to me to have much broader applicability. Outsiders, no matter how well educated and well intentioned, are likely to make mistakes that can cause serious problems for the intended beneficiaries. A few examples that come to mind:
1. Agricultural specialists teaching contour plowing to small villages in New Guinea rain forests. The locals had developed practices to plow up and down the slopes to facilitate drainage. The outsiders explained that this approach increases soil erosion. The locals tried contour plowing. Water didn't drain, but accumulated in the horizontal furrows. Eventually there was enough water to cause mudslides.
2. Western agencies swooped into post-Soviet Russia to help set up a capitalist system. They taught about corporate structures, accounting principles, and governance practices. They advocated for rapid privatization to put important assets in the hands of people with an ownership stake in making them productive. The assets ended up in the hands of ruthless opportunists who neither knew nor cared how to make them productive.
3. US charities often get funding from governments or charity umbrella organizations like foundations or the United Way. They have to adapt their organizations to have programs that achieve measurable outcomes in the donors' priority lists, often abandoning or de-emphasizing the activities that started them and earned community support.
I would say that this means that outsiders should do some field testing before recommending particular methods or interventions. It is the same as with medicine. We don't just try investigational medicines that have good potential. We actually have to do at least phase 3 clinical trials to show that the intended effect works and monitor the results after marketing as well.
The recommendations should also come with precise instructions how to apply them. With medicines it is easy – how to take it, at what dose, for what conditions exactly as diagnosed by symptoms and necessary tests, what should be monitored, when the drug should be stopped and for what side effects etc. Harder with other interventions but not impossible if the recommendations are based on evidence based practice and serious field work.
Politics can often interfere with these principles. Even in medicine it was evident with covid vaccine mandates even thought the vaccine did not stop infections and thus were not relevant for herd immunity but only for individual protection. We should aim to make our politics more science based, not less.
2. The capitalist system worked quite well throughout the Warsaw Bloc countries, notably in Poland. Only in the former USSR and particularly Russia did it fail. It is worth thinking about why this happened and how/whether it could have been prevented. (No I don't have a simple answer to this)
"Foreign Aid" has pretty much always been a tool used by developed governments to attain their foreign policy goals, whether that's helping out weaker allies, repremanding other poorer sovereign states by withholding it for doing somethat the developed state doesn't like or even just plain out improving things so there are fewer people willing to make dangerous journeys to the western states and making a claim on that western state's welfare system: much cheaper to pay poor people to stay put in marginally more bearable poverty than to deal with them when they turn up at your door in search of a better life.
Some of the people involved in the whole process as even well meaning; that's about the best and most positive thing I can say about the whole process.
And yes, the amount of damage "foreign aid" has done to developing countries is mind boggling. C.f. how when China sends very cheap steel over to Europe in huge quantities the Europeans call it "dumping", but when Europe sends over very cheap clothes over to Africa in huge quantities they call it "foreign aid".
I'm not saying all foreign aid is bad, humanitarian funding after disasters is about as close to an unalloyed good as you can get but the whole general structure of modern "foreign aid" very rationally often leaves a sour taste in the mouth of the receiving countries.
Great theory, let's see if it's actually true. Below is the list of the top ten recipients of US AID funding between 2000 and 2020. If your theory holds up, these countries should demonstrate high levels of historical responsiveness to U.S. priorities, either because they would modify their behavior to keep the aid flowing, or because the aid would be cut off if they stopped supporting u.s. policy.
I'm going to use percentage of time the country votes with or against the US in UN contested (ie non-unanimous) resolutions as a shorthand for the countries' level of alignment with the U.S., but let me know if you'd prefer a different metric and we can reexamine.
1/2. Afghanistan and iraq- I want to exclude these, because the aid was provided in tandem with active military operations and the state receiving the aid was subordinate to the US. You can, however, say that the aid was directly tied to US policy goals here, so I'll give you the points. I still think this is apples and oranges though.
3. Jordan: extremely closely aligned with U.S. policy in a way that differs from the region. Point for you.
4. Ethiopia: votes against the US in the U.N over 75% of the time, and there's no outside context that seems to demonstrate a high level of alignment. I'd say this is a point against you.
5. Kenya: votes against the US 80% of the time. I'd say this is a point against you.
6. Nigeria: votes against the U.S 78% of the time. I'd say this is a point against you.
7. South Africa: has one of the lowest alignment rates with the U.S. in the world, at only around 12%. They've also championed numerous actions against U.S. policy priorities. Strong point against you.
8. Colombia: only votes with the U.S. 25% of the time. However, south america tends to vote as a bloc in the UN, which might make this a bad metric for Colombia, especially since they seem pretty supportive of other US policy goals. I'll tentatively give you the point here.
9. Pakistan: definitely not; low voting alignment in the UN, to say nothing of what happened in Afghanistan. Point against you.
10. Uganda: again, they only vote with the US 19% of the time in the UN. I'd say this is a point against you.
So - in summary - this is the top ten list of non-defense recipients of U.S. aid; by your logic, these states should be the most vulnerable to influence from the US. Instead, the majority look like they're not particularly responsive to US foreign policy goals. Despite this, U.S. aid continued to all but one of these states.
This is quite persuasive to me. Foreign aid may cause more resentment that gratitude, especially in the top echelons of the country.
The very act of providing foreign aid indicates that the donor considers the local government weak or inept; few politicians would enjoy such insinuation.
Probably - especially if the aid is provided in a condescending fashion.
I think a bigger factor might be that, in certain cases, the act of supporting the provision of aid is sufficiently onerous that the politicians in the recipient countries don't really feel in debt.
The average politician in those countries doesn't really garner a direct benefit from the aid. Sure, his people eat, but the goodwill generated from the food (if any) will go to the country who's name is on the bag of grain, not the politician who enabled it.
On the other hand, the politician is 100% on the hook if aid goes missing, if a foreign worker gets attacked, or if anything goes wrong. Even worse, from his perspective, these aid programs aren't even going to pay him a bribe for his help coordinating all of this. Lots of personal risk, minimal personal gain.
In this context, I can see a politician not standing in the way of the aid shipment, but not feeling particularly loyal as a result of it either.
Foreign aid NGOs spread America's official ideology, leftist progressivism, of which anti-Americanism is itself a part. E.g. see South Africa, whose anti-colonialist DEI socialism is obviously Western in origin.
> just plain out improving things so there are fewer people willing to make dangerous journeys to the western states and making a claim on that western state's welfare system
Foreign aid given for this reason is not likely to make the country's elite's support the US. My other two reasons were just part of the very many ways in which foreign aid is deployed (how come so little US foreign aid to Eritreia?), there are other reasons "foreign aid" is commonly used.
In fact, "foreign aid" can actually be used to try and destablise an anti-western country's leadership because now the common people see all these "nice"/"rube" westerners providing them with money and stuff for free/cheap and thus can be an efficient vector for spreading western values (as part of western country foreign policy goals, see "democracy/freedom"), which the leadership of the country might not like and in fact might even make them more inclined to vote against the US at the UN if anything, all while providing the western country with a very easy screen of "we're just handing out aid, that's a good thing, right?" to hide behind. Imran Khan in Pakistan was very very against western foreign aid for very similar reasons, and there was force to his argument.
Because in 2005, in the name of "self reliance" Eritrea expelled almost all foreign aid organizations, including the United States. The country now rejects most foreign aid.
And before you ask - prior to 2005, Eritrea was a significant recipient of U.S. foreign aid. It seems to have had minimal coercive effect on Eritrean policy.
> before you ask - prior to 2005, Eritrea was a significant recipient of U.S. foreign aid. It seems to have had minimal coercive effect on Eritrean policy.
> in 2005, in the name of "self reliance" Eritrea expelled almost all foreign aid organizations, including the United States. The country now rejects most foreign aid.
Pre-2005 Eritrea was generally antagonistic towards the U.S., but they accepted aid, primarily in the form of food.
In 2005, as the Eritrean government's xenophobia got worse, most foreign organizations, to include aid organizations, were banned and the country stopped accepting aid.
Yes, the US is a particularly poor example of this; USAID was clearly pursuing its own independent agenda rather than serving the interests of the elected parts of the government, that's why they got abolished.
Other countries have more internally-aligned governments and use foreign aid in more rationally self-interested ways. When China hands out foreign aid you better believe it's going to be in their own interests.
Specifically, Ethiopia's intervention in Somalia fit closely with US (former?) anti-Islamist foreign policy goals. As did Nigeria in the Boko Haram conflict and various ECOWAS interventions, as did Colombia with US efforts against drug trafficking and socialism. And Pakistan was on *both* sides of the Afghan war.
This leaves Kenya, South Africa, and Uganda. The foreign policy interests aren't as obvious (though Kenya did send "peacekeepers" to Haiti recently, and Uganda's been involved in looting the DR Congo, with American multinationals often at the end of that supply chain) but it's worth remembering that they are all high-population English speaking, poor countries in Africa (as is Nigeria), the very sort of places an aid program needs to be to maintain its credibility as a foreign aid program to the American public.
Right, UN voting alignment is a shortcut to try and get a legible number of how synced up a country is with US policy goals, but I'll agree it's not perfect metric.
I'm not saying these countries never end up working with the US, I'm saying that the thesis of "Foreign Aid" has pretty much always been a tool used by developed governments to attain their foreign policy goals" doesn't hold up to evidence.
That being said, nearly every point you mentioned doesn't work as an instance of the country just bowing to us foreign policy goals. And it's a stretch to say that one instance of win-win cooperation means that they are (secretly) beholden to America as a result of foreign aid.
Ethiopia and Nigeria were both motivated to accept U.S. security assistance due to a tangible threat of external and internal extremists, respectively. I think cooperation against mutual enemies explains these partnerships way better than that they'd been poked with foreign aid.
Colombia - I conceded this one in the original point, but I can also see reasons why cooperation on counter drug issues would have been popular domestically there.
Pakistan - right, both sides. Seems like the aid didn't cause their leaders to fully align with us policy goals. Pakistan supported the US when they thought that doing so would rid them of a security threat, and then worked against the US when they were concerned about losing ground against India.
These countries are motivated primarily by domestic political and security concerns. Foreign aid, if it's even considered, doesn't seem to serve as a primary motivator.
Your interpretation of the data seems flawed in an aircraft-survivorship-bias sort of way - absolute values vs. rates of change. As an analogy, consider blood transfusions in an emergency room. Look at the ten patients who received the most units of blood, and it wouldn't surprise me at all to find that they were also doing the worst in terms of how close they were to hypovolemic shock, or other problems blood transfusion is meant to solve.
Similarly, if foreign aid dollars have the underlying goal of *improving* a country's receptiveness to US policy, it shouldn't come as a shock to see that the most dollars are directed where the greatest change in that variable is desired, that is, where the receptiveness is initially very very low. And if foreign aid is often ineffective at its explicit goals, due to fundamental structural problems, the idea that more covert goals would also fail seems plausible, meaning the initially low receptiveness to US policy often remains low. Seems like that would adequately explain the top six, at least.
That only holds true over a single iteration of this game, or at least a small number. Sure - if you're starting from a blank state, the country you most benefit from trying to change is going to be the one you try to change the most.
But we're talking decades of sustained aid funding with no meaningful bend in the recipients policy towards the US. Plenty of time for the, to use your analogy, the doctor to realize that despite putting multiple transfusions into the patient, they're not getting better.
These countries have been large-scale recipients of foreign aid for a long time, and have shown no particular inkling of that having impacted their policy.
The much more simple explanation is that this aid is legitimately intended to address humanitarian concerns.
I think you’ve got the link between theory (“US foreign aid is intended to influence policy”) and prediction (“top 10 recipients of US foreign aid should show high responsiveness to US priorities”) exactly wrong. If the theory is true, then the US need not give foreign aid to countries that are already responsive to, or aligned with, US priorities (there’s no need: no return on the investment). Instead that aid should be targeted to countries that are the furthest from alignment with the US: for these countries there is a chance of a policy return on the foreign aid spent. This is exactly what your list shows, and so supports, rather than contradicts, the theory.
Note that from this perspective the “destroy local VOs and local trust” is not a bug of the foreign aid program: it’s a feature, especially when foreign aid is spent on “enemy” countries.
Maybe this argument would make sense of you just looked at single-year funding of aid, but it starts to make less sense when you consider that the aid programs to these countries have been running for decades.
Why do you think it “starts to make less sense”? The theory is : “US gives foreign aid to countries it wants to bring into alignment with its policies”. The direct implication of this is “US gives foreign aid to countries that are not aligned with its policies (in an attempt to bring them into alignment)”. The prediction that follows from this is: “over any time period, US foreign aid will preferentially go to countries that are not aligned with US policy”.
Maybe you think that, seeing that foreign aid does not immediately make a country aligned with US policy, the US will directly cut off aid? That doesn’t make sense, because if cutting off aid does anything it pushes the country further from alignment with US policy: contrary to the US goal. It also makes clear that aid is just a policy tool, not actual charity! (which undermines the use of aid as a policy tool in other countries).
Ok, quick question before we go on - do you think aid actually makes the recipient country change its policies?
I don't, which I think invalidates your argument, because the system your describing would collapse the second someone realizes that all this aid to unfortunately countries isn't actually resulting in policy goals getting accomplished.
It seems like your argument relies on the idea that the aid doesn't actually coerce the recipient country, but that the donor country thinks it's an effective tool. This situation I guess is plausible in the short term, but I don't think it would last in the medium-to-long term.
I think that foreign aid necessarily causes policy changes (any injection of outside money is bound to cause changes). I think those changes are typically not beneficial for the recipient country (increased reward for corruption, destruction of voluntary organisations, “lock-in” of the recipient role, misaligned incentives for the recipient country, interventions that actually don’t work, and so on). I also think those changes can be beneficial the donor country, but not necessarily in terms of policy change: it seems to me that the benefits come from the donor country getting unwitting “agents” in the recipient country, getting routes for funnelling money inwards to people aligned with the donor’s aims, getting positive advertising to third parties, and so on. I don’t think foreign aid actually helps the recipient countries, whether it’s government aid, NGOs, or personal giving (I think personal charity should be local, aimed at addressing problems that you can actually see around you - at least then the incentives and feedback for the giver are aligned).
> would collapse the second someone realizes that all this aid to unfortunately countries isn't actually resulting in policy goals getting accomplished.
Depends on the allocation mechanism and available alternatives. A very bad plan might nonetheless be the best available, and a strategic planner might budget a fixed pool of resources between various regional goals without strict objective ROI targets for each.
Depends on what kind of aid you are talking about. Military aid or commercial investment? Absolutely, often an explicitly stated goal of the aid. Humanitarian aid? Not so much. There hasn't been enough of it to significantly affect a nation-state's international orientation, for one thing, and NGO delivered aid is typically audited reasonably well, it doesn't directly benefit local elites, unless they are allowed to steal some of it. That doesn't mean it doesn't do any harm - but most of it will be of the unintended kind, like that described here.
My concern with this post is that it is entirely anecdotal. It could be the case that NGO culture destroyed indigenous VOs, but it could also be the case that they never would have gotten anywhere anyway. Or it could be the case that the examples the author cites are real, but a minority. It could also be the case that none of this matters and what these places really need is economic growth. Who knows?
I don’t think the international community should be in the business of funding something like AMP. That seems like a recipe for getting yourself banned from a country.
All that being said, the phenomena of money crowding out preexisting motivation in is highly studied in behavioral economics and is very much a thing. So it wouldn’t be surprised if this was happening in VOs as well. Likewise, the value of trust and social capital is a big deal in economic development literature.
So everything claimed here could very much be true and at the very least it makes me think. How does something like BRAC fit into this framework?
Have you read the book? (I haven't only the review.) It looks like a serious work. Many pages in the bibliography... I tend to believe it, but I know nothing.
> It could be the case that NGO culture destroyed indigenous VOs, but it could also be the case that they never would have gotten anywhere anyway.
This isn't actually responsive to the post, which observes that NGO influence removes capabilities that the VOs initially possess. It can't be the case that they never would have gotten anywhere anyway, because they're specifically selected for having gotten somewhere.
I gave this review a 9 or 10 when I read it months ago. Yet another reason for keeping charity local. (Or at least on a personal level... going somewhere to help, or helping support someone you know who is going somewhere to help.)
I have learned a lot from this review, but I wish it wasn't so naive about the "three-self" church in China. It didn't so much weather the storm of mid-twentieth century as got co-opted by the CCP government, propped up as the only legitimate church, and used to control and suppress dissenters. It's not an example of building capacity for local collective action. Although this is probably a bit beside the point.
This was great. I immediately drew a parallel to the Schlozman/Rosenfeld “Hollow Parties,” wherein a certain kind of ‘professionalism’ dramatically decreases first the incentives and then the capacity for broad public participation. This in turn is the central driver of the org’s *legitimacy*, both to its members and in the context of the broader formative institution it aims to participate in.
Time flies, I thought it was earlier this year or perhaps last year...but the last big UBI paper review in 2024 mostly made me more confused on this matter rather than less. Why would giving money to the poors in a rich country show minimal or even negative effect, whereas giving money to the poors in a poor country shows decently large positive effect? Is it all just selection effects? Are GiveDirectly's various moral hazard safeguards a key ingredient? (e.g. funding an entire village rather than just select lucky study participants)
Assuming you mean what I think you mean by "selection effects", it seems pretty plausible. Quoting Matt Yglesias:
"The explanation is obvious: Poor people in Kenya are average people who happen to live in an extremely poor country. Basic habits of hard work, diligence, and thrift don’t necessarily pay off in an environment where everybody is so poor that hardly anyone can hire you or pay for anything you make. Dumping cash on people in these circumstances really lets them level-up. By contrast, the domestic poor are — unless they are recently arrived immigrants — often people who, for one reason or another, are struggling to get their lives together in a very wealthy country. If they were thrifty and diligent, they wouldn’t be poor in the first place. Putting money in their pockets doesn’t make them thrifty and diligent, so it doesn’t really alter their lives that much."
(Though he acknowledges that this is an ex-post conclusion.)
That was the other link I was gonna reach for, being certain that I remembered MattY discussing the same paper at the time, but couldn't remember enough keywords to search SB for it. Obviously it didn't shift his views overly much, since he still tithes the blog's profits to GiveDirectly!
Basically yes, it's not as simple as Success Sequence(tm) grifters would claim, but in a very wealthy country...if you're poor enough to be a good candidate for a UBI study, that probably indicates other ongoing issues besides immediate lack of wealth that won't be enduringly ameliorated by a one-time cash infusion. Big contrast to e.g. the review mentioning Ethiopian communities saving literal pennies' worth for an entire year, until they finally have a big enough pool to loan out. (Of course, one naturally wonders whether, say, Appalachians on disability who can muster the discipline to collectively scrimp and save for a whole year, along with their fellows in the holler, might be able to achieve similar results?)
This question of professionalism vs voluntarism and to what extent one should rely on each is a more general question, it seems to me. You definitely see the thinking with regards to e.g. schoolteachers that "we shouldn't pay them too much because we want only the most dedicated". An economist would tell you this is backwards, and certainly in that case I think the economists are right. But this seems to present a number of cases where it may be correct -- or at least, cases where the trust generated by the perception is so essential that this effect is more important than being able to attract people with money. So something to think about more generally I think.
(And I do have to wonder about research and to what extent it's more or less effective the more monastic it is...)
Yeah, it depends on what you want. Teachers are hired to teach, but also to build trusting relationships with the students. Which is more important? Volunteers are recruited to deliver some service to their neighbors, but also to leverage local networks based on trust? Which is more important? The answer might lean in different directions in these two cases.
Yeah with how people (especially here) still disparage normal schools even with properly paid teachers, I wonder if there's really a point there. Maybe passion is everything.
Seems to me that the underlying problem is that we don't know how to measure the output we care about, so we discuss input instead, and speculate whether low input correlates with high output, or high input correlates with high output.
But that's insane. As an analogy, imagine buying a car, but you know nothing about cars except how much they cost, so the entire debate is reduced to "people say that if a car is very cheap it is probably bad", and "well, if you pay a lot for a car, someone could sell you exactly the same bad car, only with a higher price tag", and then we endlessly debate about which side has a deeper insight.
If you want to get good schoolteachers, first you need to be able to tell the difference. Then you can adopt the policy of paying the good ones better. As long as you insist on paying everyone the same regardless of their quality, the problem will remain -- the highest quality candidates can probably make more money somewhere else, so your offer will most be taken by the ones who can't.
Sometimes the ones who will teach for peanuts are the passionate ones, and you want these. Sometimes they are the losers who can't get a different job, and you don't want those. And sometimes even the passionate ones decide that it's time to start a family and take a mortgage, but the bank won't give them one unless they change the job, so you lose them.
This is fascinating and resonates a lot with Seeing Like a State, certainly, but also with Hoffer's The True Believer.
Pulling a quite from a Samzdat post about the latter (because Lou Keep is awesome and I miss his writing sorely):
"Mass movements do the opposite: they pretend to give you power, while stealing what little you had; they pretend to solve your problems, while entrenching them. Movements make the frustrated more frustrated and they self-perpetuate with no regard for those who perpetuate them. The seed that they sow is frustration.
This is for one obvious reason: competition. Groups that maintain frustration, or are better at sowing it, will outcompete the others. The strength of a movement is directly proportional to its size and the fanatacism of its adherents, and the fanatacism of its adherents is directly proportional to the frustration they’re trying to escape. Mass movements that are good at what they do: a) make previously content outsiders frustrated; b) further frustrate their adherents while pretending to advance the movement. This means that the strongest mass movements are inevitably going to be the ones that are the best at not delivering the goods. Any movement that actually succeeds for (advances the interests of) its frustrated adherents will make them less frustrated. Hence, they’ll stop being members. Or it will succeed at its purpose, they’ll still be frustrated, and they’ll just join another."
I like a review that values my time as a reader. Mercifully short, briskly paced, minimal distractions and tangents, sparing use of quotes (not that quotes are a bad thing, but giant/repeated block quotes really pad out the length and funge somewhat against the summary function of a review). Would have liked to see the endnote caveat up front in some capacity: it's confusing to have instances of "I", "we", etc in an anonymous writing when by design I have no information about the author. Tell us a bit about yourself and your own skin in the game! Either way, it's a nice bookend with previous ACX reviews on the foibles of international aid and development. Seems genuinely difficult, and I'm coming to appreciate the wisdom of the mindset that says it's wise to invest defensively rather than chase maximal QALYs. Even if that sometimes means "ineffectively" donating locally rather than internationally - making damn sure the funds are used appropriately and constructively is worth its weight in gold dust. Charity begins at Home Depot.
(As an aside, I am loving how we're three reviews in, and there's been no shoehorned mention of AI or the Singularity yet. Knock on wood!)
Interesting. I was surprised in the USAID defunding how little people from other countries seemed to complain. There are plenty of venues for the leadership of other countries to push back against the US. For the tariffs for example there are many instances of foreign leaders criticizing US policy. But for the USAID cuts, very little. Something does seem fundamentally wrong there, and this book seems like it might offer an explanation....
Thanks to the reviewer for writing such an informative, interesting piece on a topic that very few cover.
I spent most of my professional career working with or for local government and non-profit organizations in the US. There are some broad similarities to what he is describing here, but it's apparent to me that we treat such organizations very differently from the way that international NGO's are run. For years the performance goldstandard was the local organization that could show success in local fundraising. The more local support you could show, the easier it was (I retired a few years ago) to obtain grant funding. Most grant funding today comes from private donors, either corporate foundations or wealthy individual donors. The most successful first step, I always advised my clients, was to get some local business leaders on your non-profit board. From there, you expand into the local volunteer network, and from there to a medium sized foundation with a history of working in your region on your mission related issues, and from there...
It was all scaffolding, all the way up to the annual revenue growth goals. Sure, even a medium sized non-profit has paid full time staff, but this is the US, where the culture trusts professionalism to a greater degree. But even here there are limitations, because many impoverished neighborhoods are served by very small non-affiliated churches, and those organizations tend to be rather suspicious of outsiders, largely for the reasons outlined here.
The point is that evidence of local engagement was mandatory if you wanted to expand your revenue base. It's also measurable, so a useful standard to use.
Take note however: "When a measurement becomes the target, it ceases to be a good measurement." Fetishizing *anything* because it is easy and simple to measure eventually undermines the entire effort. There is no substitute for local informed judgement.
That, ironically, requires that we trust the local leaders. That may be hard for some donors.
>With the interests of the powerful defended more robustly than the rights of everyone else, a stark income gap has opened up between the top quintile and the rest of the population.
The Gini index of Pakistan is about the same as that of Germany:
Yeah this immediately stands out to me. Sometimes even while the gap seems wide, it's still much less than how Western billionaires tower over other citizens. But I think people agree that western commoners still live better than other commoners.
This is in my mind whenever someone says "trickle down economics don't work!!"
wish there were any comments contesting the thesis of the book. I don't see why the political issue should be "foreign aid", all these failures seem preventable one way or another.
Human remark: "I tried helping them (by doing really dumb things), but it seems to always be harmful. Guess help is the wrong approach."
If everyone tries to help and none of it actually works, maybe it's not such a simple problem and we should reconsider the assumptions that brought us here.
And for some of us, this is in the "confirmation that water is wet" category. I've been hearing this message for my entire adult life, including from leaders who have been directly involved in many relief and charitable operations (associated with a church, but secular in aim). I've seen it directly in miniature when I've been involved with helping people both individually and organizationally.
I don't think the things were "really dumb". I think it's a really hard problem to solve.
There's many failure modes when you try to inject wealth into a poor society to make it less poor. But I guess most of them seem to come down to the same thing: people rearrange their lives to be at the point where the wealth is being injected. If you can solve that problem, maybe you can make things better.
Other people have mentioned GiveDirectly, but does that solve the problem? It seems it would come down to how GiveDirectly chooses its recipients. I've heard that they give at the entire-village level, and that they use criteria like "thatched vs metal roofs" to figure out which villages are poor enough to be worthy. Which sounds fine the first time around, but you can imagine that once word gets out then you'll have village big men going around and replacing everyone's roofs with thatch so they can look poor enough to get money. If GiveDirectly is still working well, it's only because it's small enough ($260M/year) to not be causing much distortion just yet.
This echoes my experience spending a year in an -istan working for an NGO. I'll add a few money related anecdotes:
Paid employees were transparently (and often successfully) using their pay stubs and resumes to angle for a Western visa, contributing to the brain drain of college-educated talent. Also created lots of turnover.
The lure of high paying NGO jobs could be an impediment to government relations and getting things done. For example, the regional governor made clear his cooperation was dependent on us hiring his nephew. This put us in the difficult position of either forgoing government support or engaging in the same kind of corrupt practices we were opposing.
One of the tasks our employees carried out was holding focus groups on various topics. The local participants (IMO reasonably, considering we were a monied Western org and they were dirt poor) expected to be compensated for their time. We eventually realized that we were not getting a representative sample of the population, because whichever employee we tasked to assemble the focus group would inevitably only invite members of his own tribe (often extended family). When I tried to explain to employees that this was a form of corruption, they were often confused because they viewed directing money to their tribe/family as a solely positive thing.
Now I really, really want Scott to write a post on "Foreign Aid: More Than You Ever Wanted to Know" that synthesizes the information in this review with the standard Effective Altruism defense of helping poor people/greatest good for greatest number, and gives us Scott's informed, evidence-based opinion on where/how we should donate.
Same. It feels like everyone and their mother has an anecdote about when foreign aid does or doesn't work, but at the end of the day I don't really have a model of what option has the better outcomes.
One potential problem I didn’t see mentioned is that people like being part of something small because it feels like you’re making a marginal difference. Being one more volunteer in a group of 10: great. One more in a group of 100,000 volunteers less so.
This limits scope. Voluntary orgs can only get so big before, meaning you need 1,000 little orgs, which are impossible to maintain oversight of and assess for effectiveness.
My eyes glazed over as it became increasingly clear this was written by Claude. Shocked that this is a finalist. The thesis seems interesting but reading it felt like consuming paste
Didn't feel like AI at all to me (and multiple other people upthread). I guess it's score 1 for AI if we're now reaching the points where humans genuinely dispute over whether or not something is written by AI.
People have been commenting “this was obviously written by an AI” on any writing that actually uses correct grammar and punctuation and a reasonably rich vocabulary, ever since ChatGPT-3 came out.
I’m probably in the wrong here since Pangram and most people agree it’s human. Apologies to the author if so. But as Scott said in the post about taste:
“The [Kenyan guy who gets told he sounds like AI] argues that these are signs of learning a language in a formal, ossified, officially-good way rather than as the living organism of a spoken tongue. But he could equally note that his Kenyan essay test involves people with limited knowledge of English under high pressure to sound smart, being graded by other people with limited knowledge of English. This is the same levels-of-abstraction bottleneck that produces AI writing, and with the same result.
Think about the hallmarks of AI writing. “It’s not X — it’s Y!” Direct, simple, catchy, clarifying. This is a great writing technique, used in moderation. AI takes all the great easy techniques that work and can be reduced to a simple script, then overuses them until your eyes bleed. This is the essence of bad taste”
Don’t tell me it’s just good grammar. There are many examples of LLM-like writing techniques in this piece. Whether or not it’s the ESLers’ fault, the well has been poisoned. This style of writing is really irksome now.
As stated in the very paragraphs you cite, the problem with these writing techniques is not that they’re inherently bad but that AI overuses them to the point where they get grating. I don’t think this review author does that, and I don’t think no human should ever be allowed to write “it’s not X, it’s Y” ever again, no matter how appropriate it may be at that point in the text. But YMMV.
Best review and my prediction for first place winner.
Yet further proof Trump was sent by God. Wallahi he shut down USAID which was doing untold damage to not just American taxpayers but foreign countries.
I am proud to have lived through his presidency. I am proud everyday when I saw the signs in my supermarket. That welfare can no longer be used for candy bars
“”“I hate having to post criticism of EA. Not because EA is bad at taking criticism. The opposite: they like it too much. It almost feels like a sex thing. “Please, tell me again how naughty I’m being!”
[…]
Many of the highest-upvoted posts on the EA Forum are criticism. The more critical, the more upvotes. […] A case against RCT-driven development aid, somewhat related to the one in Anti-Politics Machine, got 389. It was the #6 highest upvoted post of all time; the only reason it wasn’t higher was because stronger criticisms of EA beat it out.”””
I feel like this has broader implications for modern developed economies and the disaffectedness and lack of agency people seem to experience when structuring their lives around contemporary corporate work
I am not even convinced about trees. As far as I know, most planted forests in Latvia and other places are monocultures. Environmentalists moan that it destroys biodiversity and they are probably right but I haven't really noticed any disasters resulting from planting pines or other monocultures. Any noticed issues are solved with small tweaks or certain interventions. Those forests grow, then are being cut down and replanted again. It basically works just like crops.
Maybe now they change the type of trees, or plant sequentially different types. Sometimes they treat young trees with fungicides or other agents to limit spreading of certain insects, fertilize the soil if depleted of certain elements. It might be ugly but it basically works.
- Indeed loss of biodiversity, you cannot have a rich ecosystem in a forest uniquely composed of pines that you cut down regularly, in comparison to an older and diversified forest. The ecosystem is not only composed by the trees, but also of all the animals/fungi/vegetals dependent of those trees who also get affected when you cut them down and don't magically reappear once a new tree is replanted.
- Higher susceptibility to be affected by diseases. In a monoculture, a disease can easily spread across the entirety of the forest as all the trees have the same weakness against it.
- Higher susceptibility to fire. Pines burn extremely easily, see for example the current big fire in the southwest of France. A pine monoculture is probably the forest type that burns the easiest.
Given that green revolution seems to be the most successful thing on the planet, I'm in doubt whenever someone's bringing up traditional farming wisdom analogy.
Speaking as someone who's been contributing to an open-source project for years, I have seen that project go from being entirely volunteer-driven, to being funded by Google (including my own salary) to being funded by several companies.
The funding helped put more work into the project, but it completely killed our community-building efforts. There was now a small clique of insiders (the Google-funded devs) taking all decisions, followed by several cliques of insiders (the different companies). Over time we spent less and less time advertising our work and reaching out to the devs communities that were interested in us.
I'm trying to keep it alive and keep working on it, but it's so hard stay motivated now that I'm not paid to work on it anymore. It feels like nobody gives my opinions any weight, and any work I do will be me getting exploited so someone else can get a paycheck.
I'm not sure what the solution is, beside moving on to other projects that haven't been compromised yet.
Glad this review made it. I had given it a 10. I worked with/at "NGO"s funded by our Gov and hated it. Loved the money, ofc. I did the same work for a private NGO (Robert Bosch foundation) before and though, nay because we got paid a third then, staff was on average much better (while the DAAD/GI do take on staff wisely rejected by Bosch) - see, if you are only willing to do it for a high salary, you are likely not really into the mission - things went much better, the 'bosses' were even interested in results not just "budget spent"; and staff was sent to places that actually asked for it,
while the "NGO" had to beg, bribe and cajole places to take their "helpers".
I can imagine the truth in all the examples that have been shared. There is so much to be done, so many issues, that at times, the rabbit dance must be a reality... no matter how obnoxious it sounds.
Acknowledging the sharing and warnings shared via Breakdown in Pakistan by Masooda Bano.. I like to believe that there are Civil Society Organizations, Trusts, VOs who do keep (hold on) to the essence of their vision/ their goals.
Between 1998 and 2001, when I travelled to Odisha in India, the flight (Small aircraft) would only have representatives of INGOs and Government Officers.. i had thoughts then, as to, how the spirit and goals of such donors would bring change, and progress for this state. I feel very good to see Odisha now. It has developed and moved ahead in a number of indicators. Social indicators are good and a pro active governance show how ground level implementation is satisfactory. New initiatives, encouraging sports are always on the agenda. (Among other positive activity) Civil society actors continue to play a role. There are pockets that need more focus, that is accepted. I have seen this myself, over the years.. from 1998 to 2001, with a gap in between and from 2015, 17, 19 and 2023 and now..
I agree and appreciate how Bano has said this can be avoided.. can save the essence of the Community Based Organizations.. the examples of Pakistan say it all.
Community is a strong pillar in the work of many organizations. The tripartite of Community, Corporate /Donor and Government are ideal for scale. And it is the toughest.. with challenges mentioned so well in the review..
While this is a reality, there is a lot more that makes the whole truth...
Besides the sharing in the review, I am also touched by Social Welfare professionals in Canada , Sydney or London (as this is what I have seen). The professionals who work with the Dept of State, do their work well.. they believe in what they do.. they are trained and carry the spirit, working for over 10 or 15 years.. Does the assurance of funds/job security make this happen ? They do not get a whole lot of perks.. my sample of friends and acquaintances may be small ..yet, I marvel their at their service, commitment and objectivity.
You have given strong reminders through this read.. Well done!
In my public health course I studied healthcare in Afghanistan and my impression was that foreign help was highly effective. The foreign sources provided about $10 per person per year and that was basically all of its healthcare budget for this country. Most of this money was used by NGOs who provided healthcare services. The outcomes showed great improvements in life expectancy, mortality, especially child mortality and in all other measures. There was a need for the country to become self-sufficient and become less dependent on foreign aid and some other issues but overall that was a big success story how much could be attained by so little money.
This study greatly influenced my view on healthcare spending in general, showing how great is the curve that marginal healthcare spending has diminishing returns (the case in most western countries).
I haven't followed up what is happening when Taliban has taken over but a quick glance shows the country continues to rely on foreign help.
The biggest problem is not how to properly spend the foreign aid in most efficient manner. This example shows that it can be done when it really matters despite many anecdotal examples that it fails. The problem is that without changing country's fundamentals, it is not going to fundamentally change citizens lives to the better. I don't really know if I should be glad for Afghans having healthier lives than in the past or be sad for them (especially women) living under Taliban regime.
For all practical purposes I am in favour for them receiving foreign help to run their healthcare and making sure this money is used efficiently, yet I am not happy with the outcome of the country in general. Sometimes doing right thing does not bring joy.
The expectation was that the government takes over and provides a system where people get paid for their work. Unfortunately the government is too weak and that causes whole range of problems that this particular issue is just one of the many.
If this hashar work could be disrupted so easily, it means that it was never strong in the first place. Probably, it was also quite inefficient and any paid work is more productive and efficient.
In my childhood my father used to borrow a horse from a neighbour to till the private plot. My parents had full time work in the Soviet collective farm but we were allowed to use small plots of lesser quality land that were not usable by tractors. This system was common in the Soviet times and while not substantial by area, it provided significant portion of food for the family. It is not dissimilar to allotments in the UK except that it was economically relevant because the Soviet system had no functioning market economy and was plagued by frequent shortages and long queues in the shops. Despite smaller areas, poor quality of soil and needed hard work, they provided better yields per acre than collective farms. Sometimes even provided excess food that was sold to the state supplementing family income.
It is a textbook example how private ownership makes everything better than collective ownership.
But it was a hard work involving not only family members but often other people from the neighbourhood on certain days as the process of planting and collecting potatoes was very manual and required many helpers. It was done without renumeration, often relying on reciprocal help when neighbours needed the same help. Overall, it was still less productive than collective farms using tractors and other tools at scale.
The Soviet country ceased to be and private entrepreneurship was allowed again, and initial turbulences have passed, now agriculture is done not by collective farms but private farmers with rather efficient processes regardless of the EU subsidies. It makes no sense to work on small private plots even though some people continue doing this for fun.
Or not only for fun. Some elderly people are so invested and continue to grow their own potatoes. The neighbours are no longer interested in helping, so they ask their children and other relatives to come on certain days to do this hard work of planting or collecting crops. Everybody hates it. When someone does physical job daily, another day is just a routine. If one works in the office, that day becomes a torture. It rarely snows in the UK but when it does, the government has to remind people not to shovel snow too hard as many people get heart attacks from not being used to excessive physical exertion.
It is no wonder that helpers argue that economically it makes no sense as they can easily buy enough potatoes and bring to their elderly relatives. But only few people openly refuse. Elderly people shame their relatives for being such weaklings, then they appeal that the potatoes they grow taste completely different. I don't know about this, of all things one can grow in the garden, potatoes are most irrelevant by taste differences.
But the real reason is that elderly people are too old to change. They have grown their own potatoes whole life and stopping growing them means they no longer know what to do. Their lives lose meaning. So, the tradition continues. The family ties in the post-Soviet countries appear stronger than hashars in Taliban controlled Afghanistan.
Foreign aid poured into South Korea after the Korean war, largely from the US, all got stolen by the Chaebols, and then South Korea had an economic miracle (depending on your point of view, despite the foreign aid or because of it). All the effective altruism endpoints got hit: drop in child mortality, no starvation, boost in income, qalys skyrocket. There are many such cases, but the book seems to pick countries that are disasters now, then look back to when foreign aid started pouring in, not grab a historical sample of foreign aid floods and then watch forwards. This makes conclusion drawing much harder!
I guess the NGOs were motivated by Asian and German economic miracles, but got confused why it doesn't follow the same track? Seems like there's no more aid miracle after cold war.
The damage channel Bano describes is quite specific: foreign money flowing *through local organizations* reshapes their incentives and destroys the trust signals (visible sacrifice, accountability to members) that made them work.
Worth noting that the largest financial flow from rich to poor countries doesn't touch that channel at all: remittances, roughly $650B/year against ~$220B of official aid. The money arrives person-to-person, carried by the most legitimate signal possible, a family member's labour. No grant proposals, no workshopias, no salaried intermediary layer. Nobody needs to demonstrate charitable skin in the game for a wage to be trusted.
The review even contains the supporting detail: the Edhi Foundation, Bano's healthiest case study, is funded mainly by the Pakistani diaspora. Migration created that donor base.
So "let in more migrant workers, if you want to help" seems to dodge nearly the entire failure mode documented here. Two caveats:
1. Brain drain of exactly the civically capable (Hirschman's exit vs. voice). AMP's tenant farmers organized partly because they couldn't leave. But this bites much less for low-skill labour migration of the Gulf or Singapore construction-work variety, and enforced rotation turns permanent drain into circular migration: remittances during, savings and skills after.
2. Rent capture at the gate. When slots are scarce, brokers in Dhaka harvest the surplus as recruitment fees, financed by debt at moneylender rates. It's the same "large rents flowing through whoever controls access" pathology as the fire hose, just one layer earlier. Korea's EPS showed that a state-run zero-fee channel can mostly kill this. So it's an argument about how to expand migration, not whether.
Migration also doesn't substitute for the narrow slice of aid the review defends (vaccines, malaria work), since those are public goods you can't buy with remittance income. But for the broader goal of transferring resources to poor people without hollowing out their institutions, letting them earn the money themselves beats routing it through NGOs.
> In Zimbabwe, there was the religious VO that had worked for a long time against rural poverty. Following a major drought, it accepted international emergency funds and worked intensively for two years on famine relief. It helped a lot of people get food, while its staff pushed themselves overtime to comply with the intense reporting, quality, and anti-fraud requirements of the EU and UK governments, with help from international “capacity-building” partners.
> And after the disaster, when the VO went back to local supporters who’d always previously given to its work, it got the cold shoulder: “You’re an NGO now. Why do you want our help?”
I notice I'm confused. An organization taking extra aid, temporarily, during a significant famine--drove away its local support? That seems like quite a bit of an overreaction. What did the locals want, for them to just let people starve so that they don't get too big for their britches?
Just guessing here, but maybe previously the local people had mental model "if we won't help to make it happen, no one will", and now their mental model is "there are people out there who have several magnitudes more money than me, so it does not make sense for me to contribute; the NGO should try to get some more money from those rich donors".
Basically, just knowing that there are other sources of money out there reduces your willingness to spend your own money.
This is a different situation, but in Europe, many people's plans to do something prosocial start with step 1 "try to get some EU grant for that" instead of thinking how they could use their own resources (and maybe try to minimize the cost, instead of dreaming big).
2. Outside cash corrupts a VO, turning it into an NGO/jobs program. Dutch Disease (easy money drives out every other form of economic activity) is most instructive here.
The problem is that the pendulum swings both ways. Today it is “NGOs are spending too much time having paid staff write reports”, tomorrow it’s “there’s no accountability, look at the waste and theft!” Ironically, if you asked most donors the latter is their primary concern. You’ll see some conservatives reading your review as justification for axing USAID and killing hundreds of thousands of people, but if you presented your anecdote about the Islamic school in isolation those same people would freak out over money finding a religion they don’t like.
My half hearted solution: make the paid locals suffer. The catholic solution is to insist on the self mutilation of monks, or at least the pain of Mother Theresa. In the modern secular world: consider making the report writing even more painful and tedious, so that their community hears the anguish from every excel sheet.
The metaphor of GHD aid as a monocrop plantation that erodes the natural local ecosystem, and then languishes when the common goods created by that ecosystem have been depleted, is useful but I think it mischaracterizes things:
(1) The local ecosystem in LMICs is often dysfunctional — perhaps plagued by extractive patronage networks that consume the undergrowth like an aggressive local species, or perhaps just stuck in a stable equilibrium at an early stage of development like scrubland at the desert’s edge that’s never had the nitrogen or soil depth to become anything richer [I had to do some biology googling to make this strained metaphor work :P]
(2) GHD aid needn't be a monodrop plantation — at its best it could be permaculture. But admittedly, at its best today AFAIK it's more like a patchwork of targeted grafted planted mindfully not to undermine what’s already working in the ecosystem.
What's the closest to permaculture style GHD interventions (that might work) that you know of?
(Stuff that comes to mind for me is more on the growth side then the health side like Africa Jobs Fund or Mushtaq Khan style industrial policy)
Even that cleaning water channels thing is more complicated that was described in this review. The system, as actually implemented, as basically corvee labor; sometimes induced by force, sometimes induced by social or economic coercion, but definitely not voluntary. It's not an accident that the person showing him how the system worked was, in his words, a warlord.
It doesn't shock me that the system of "let's all do miserable backbreaking work for free" for replaced by another system, nor does it shock me that the people weren't excited to go back to it once the other system fell apart. Sometimes you're in a bad, but stable equilibrium.
The amount of civilizational damage "We Are The World" and it's spiritual kin were upstream of boggles the imagination. So, so beautiful, so, so well meaning, and so so terrible, all at once. It's like epic tragic poetry written by an evil god, using whole civilizations as it's ink.
Would you mind elaborating, ideally with sources? I hadn't heard about "We Are The World" before your comment, but my brief research on it seems to suggest that it's been a mostly positive influence on the issues it's targeted, particularly famine.
Every dollar that good, well-meaning westerners spent on that record went directly to an armed partisan with an AK-47, a grudge, and a new cudgel to hold over the heads of their oppponents in a long running civil war. I mean seriously, look at Etheopia, Somalia, and Sudan now. If that's what you think "mostly positive" outcomes look like, I don't know what to tell you.
[Also, you /really/ could have googled this for yourself, instead of trying to make it a public status contest. This really wasn't difficult.]
https://hhr-atlas.ieg-mainz.de/articles/sasson-ethiopia
https://lordslibrary.parliament.uk/ethiopia-conflict-and-food-insecurity-40-years-on-from-the-1984-famine/
================================================
International 1984–1985 famine relief for Ethiopia inadvertently entrenched the Derg military dictatorship by allowing the regime to weaponize aid, fund its civil war, and enforce a brutal forced-resettlement campaign.
Manipulation of Aid and Military Strategy
Food diversion: The Mengistu government diverted international food aid to feed military forces and starve out rebel-held northern regions like Tigray and Eritrea.
Prolonged conflict: Foreign assistance kept the state afloat financially, freeing up domestic resources to fund counter-insurgency campaigns rather than agriculture or relief.
Apolitical framing: Global charity events like Live Aid framed the crisis strictly as a natural drought, masking the man-made political roots of the disaster.
The Resettlement Campaign
Forced displacement: Using relief logistics, the regime violently moved roughly 600,000 farmers from the north to the unfamiliar south.
High mortality: Conditions during transport and in the un-serviced southern camps were so atrocious that the relocation itself caused tens of thousands of deaths.
State control: The program served as a counter-insurgency tactic to depopulate regions supporting the resistance.
You misunderstand my comment as some kind of public status battle. I legitimately wanted to see sources that represented your position, because I struggled to find them. I have literally never heard of this song before, so it’s not possible for me to come into this conversation with a bias of this song and it’s effects one way or another.
Generally, when people ask for help understanding your position, you should be happy to engage with them, not hostile. The hostility actually dissuades honest attempts to understand your position.
That's fair, and I apologize if I misunderstood you. I don't normally comment in this space and I've spent all too much time in spaces wheer "Sawce?!" is a confrontational meme, not an invitation to good faith conversation.
I think you might find that the ACX community is a bit more academic than traditional social media shouting matches. We're all trying our best to confront our own biases and understand the world better with data.
Also appreciate the apology.
I don’t think anyone asked “Sawce?” The comment I saw looked like a good faith ask for information!
literally the first sentence of the first reply was "Would you mind elaborating, ideally with sources" I also reacted to this in same way as original commenter initially but then reading the next sentence i was like "what? that person has never heard of the song before?" and thought maybe it was a genuine question...
First time I heard of it too. I was able to google and skim a reasonably distributed corpus of sentiments within a few minutes.
Also, I'd question the prudence of anyone who wouldn't be hostile to your ~"noble" intentions as such, regardless of your ortensible earnestness.
Your subsequent reply (of which my reply here is attached) is shamelessly malignant and insufferable, more-than demonstrating the appropriateness of said hostility.
"shamelessly malignant and insufferable"
Yes I remain shamelessly curious in life, which makes learning all the more enjoyable. As for insufferable, you'll have to ask my friends. I try to remain kind and respectful, but judging myself without bias is difficult.
To your point of "just google it bro," I did, and found most sources to be very supportive of the song. Hence me asking for examples to the contrary, in as polite a way as I could.
An American businessman and future president, Herbert Hoover, famously fed starving Russians during the famine Lenin caused, thus stabilizing Lenin's newborn regime and allowing it to become one of the most evil in history.
But should he have let millions of Russians die instead? Isn't that precisely the logic of Lenin and Stalin, who were also willing to sacrifice millions of people for what they considered the greater good? I'd be extremely wary of any galaxy-brained argument that a pile of skulls is good, actually, because it's an epistemic trap that has led to horrors time and again.
There are a few things to address here. In no particular order:
Causal proximity: No, it is obviously unfair, even within consequence utilitarianism, to blame Hoover for Stalin and Kruschev. On the other hand the partisan witih an AK-47 and a grudge is not an abstract politician in some faraway capital. He is standing right there, just out of camera shot, glaring menacingly at everyone you give a cup of grain to. The choice is not between skulls and no skulls. The guy with the gun will start adding to the mountain of skulls /he has already made/ the instant the camera crews fly back to LA.
The consequentialist argument: You are warning me against sacrificing real people to an abstract future. I agree. But the people threatened by the partisan after the cameras leave are also real people, not abstractions. The dictatorship’s future victims do not become morally imaginary merely because they are absent from the fundraising footage.
Epistemic assymmetry: you are demanding enormous metaphysical certainty in the consideration of the harms of the intervention while treating the benefits of the intervention as transparent, direct, and automatic. The guy with the AK-47 glaring at your beneficiaries is not a metaphysical uncertainty. He is standing beside the distribution line, and plainly intends to finish the gallows after you leave.
Not being a useful idiot. The only thing more powerful that a partisan with a grudge and an AK-47 pointed at a man he caused to starve is a partisan with a grudge and an AK-47 pointed at a man he caused to starve /and/ a bag of grain. Maybe try to find a way to feed the starving /without/ empowering the guy who weaponized their starvation to compel exactly this response from you? For example, don't hand them cash to build camps that they will immediately start using for what we would now call ethnic cleasing?
The blackmail consequence. Maybe /don't/ teach 5 generations of the most awful dictatiors in human history that starving their own dissident populations is an /excellent/ way to get free stuff for their own soldiers and entrench their own positions. This is happening in North Korea even as we speak. Which, unlike the specifics of Stalin, /was/ easily predictable even at Hoover's time. Can we try that? Can we try /not/ making a credible threat to starve or massacre one’s own population into an “I win” button for genocidal maniacs for the rest of time?
Finally: basic dignity: if you absolutely /have/ to empower dictators, if you are /so/ scope insensitive that "if it saves even one life" is worth empowering some of the most awful people to ever exist, then maybe, just /maybe/, /don't/ write a hymn to yourself about how awesome you and your donors are for doing uncomplicated moral heroism? Can't we at least clear /that/ bar?
bit harsh to put Khrushchev together with Stalin no?
I mean, on the one hand, maybe yeah? On the other hand, Cuban Missile Crisis. I'm not sure Mr. "We will bury you!" gets a pass just because, at extremely long odds, that happened to work out sort of ok.
Not feeding the masses during the 1921-1922 famine was unlikely to lead to the fall of Soviet power, considering that the Russian Civil War was practically over at this point and the remaining resistance was scattered.
Well, perhaps they should have handed the starving Soviets (Ukrainians, Georgians, Russians, ...) migration options in addition to the food?
But I know that this is politically even harder to pull off.
See also UNRWA, which both funded terrorists and entrenched a permanent refugee underclass across large parts of the arab world.
Regarding the issue of the quality of an organizations charity commitment being illegible to a donor, maybe this is a prime initial case where an AI would be good manager / CEO. We can all talk to Claude Fable 5 and get a good idea of what it's behavior is like. AIs are reproducible and legible in a way humans typically aren't. I honestly think, even TODAY, society could be using LLMs for judges, jury, etc and reaping massive payoffs. Obviously handling the buggy edge cases is an open question. Double or nothing appeals for human review?
I'd trust a jury made up of 12 of the top models on arena.ai to make the right decision for the right reasons over a jury of 12 random people from off the street in pretty much any criminal case, and especially so for complex fraud/murder cases.
This is the wrong framing. The reason a jury of peers is such a powerful concept is not because 12 random people are expected to be most logical about a specific issue. If that's what we wanted, we would have relied on elite / expert opinion, as was done in aristocratic regimes that preceded the democratic transition. The reason a jury of peers is so powerful is because those 12 individuals are your peers, that is, their interests and incentives align so closely with you (or the population at large). Everybody should prefer to be judged by those who roughly share their same position. At least with a jury of 12 humans, you know they broadly care about what humans want, because they are human. Yet, by putting 12 Claud Opus 4.8's together, the incentives are so far out of whack that this system would quickly become unrecognizable. Why are they deciding the way they are? What is their training set? Who built them? What are their incentives? Who decides which models get to be selected from? How do we know they actually care about humans at all?
You're misapplying the same logic that was discussed in this very blog post. Namely, looking at a system built on cooperative participation, identifying a single bottleneck (in their case money, in our case raw intellectual power) and then destroying the entire system to maximize along that dimension.
Right. It's all about freedom from oppression and exploitation. A valid concern with respect to AI?
What are their incentives? What is their training set? Who built them?
Why does any of that matter? If they legibly and consistently do well on a large jury test suite, and continue to do so, they are better than a human jury.
The only concern of course is making sure that test suite covers what we care about. So you have to have human juries sometimes anyway as ongoing test validation (preferably not train)
Peer juries are more sustainable in a political sense. Juries aren't just machines for determining facts, they also make a community "responsible" for its own legal actions.
If a community judges itself, it's implicated in the legal results. If something "external" judges, its much easier to get a rebellious movement going.
I think this is part of what's driving the populist backlash against expertise. Peer juries, and local agency generally (vs external stuff), are structurally insulated against wholesale rejection. If we got rid of them in favor of more accurate (in some sense) AI then it'll be MUCH MUCH easier for some sort of "I reject the whole legal system" community nucleus to form.
Juries are also a bulwark against unpopular laws.
Many countries like France and Italy and Germany and Singapore don't have juries at all. Not having juries hasn't really led to much of "I reject the whole legal system" community nucleus in these countries. In fact, the country I most associate with that type of thought is the US with its whole Sovereign Citizen subculture.
You're right, I overstated the case for Juries. I was thinking generally about community involvement / implications (which is the stabilizing thing I think). Jury trials aren't the only way to let communities feel involved.
The French Cour d'assises (which they use for the most serious crimes) involve ordinary people. Germany has "lay judges" who are basically ordinary citizens with no required training. Even China has "People's Assessors"
Germany has lay-judges. Look up Schöffe. Not quite the same as a jury, but going in a similar direction.
See https://en.wikipedia.org/wiki/Reichsb%C3%BCrger_movement for Germany's sovereign citizen equivalent.
https://www.service-public.gouv.fr/particuliers/vosdroits/F1540?lang=en
The average person on a jury is not very smart and not very well equipped to make a good judgement that is in accord with the law and facts of a case (especially more complex cases). You’re right that we know what humans want, and sometimes it’s “I want to get these deliberations over with so I can go home,” or “I disagree with this person or corporation politically, so I will vote against them regardless of the facts.”
How can people obey the law that is too complex for them to understand? If the jury can't understand what exactly the defendant did wrong, why should we punish him for doing it?
Because there is a fact of the matter as to whether or not the defendant did anything wrong, and the ideal justice system would punish people based on that fact, not based on the opinions of 12 people who fail the breakfast test.
> the ideal justice system would punish people based on that fact
Why? The purpose of the justice system is to maintain social stability by creating a sense of fairness and preventing vigilantism. If the community thinks the defendant did nothing wrong, even if the law says they did, punishing them would be counter to that purpose.
You're making a really strong assumption here that I'm not sure you're acknowledging.
"Thre is a fact of the matter as to whether it not the defendant did anything wrong"
Are you saying right and wrong has a clear, mathematical answer that can be trained on to get a proper loss minimzmization? I think both AI researchers and lawyers would strongly disagree with this. Passing judgement on somebody can't be an algorithm. It requires judgement, which mixes intelligence with empathy, a shared framework of understanding and shared values.
This is also why it doesn't make sense to find the "best 12 jurors" in Boise, Idaho and then send them all the local cases from Islamabad, Pakistan to rule on. Not just because measuring "the best 12 jurors" is fraught with issues, but because the context they are operating in matters just as much as their accuracy. The local customs, values, and constraints of the population in Pakistan is very different from Idaho.
While the law strives to be objective to the best of its ability, if you study law down to most minute details, you see how much room is left for personal judgement. And again, I haven't event touched on the issue of trust and shared investment in the outcome (why should people in Islamabad trust the judgement of people in Boise, and why should we trust the judgement of Claude Opus 4.8?)
Counterclaim: The reason a jury of peers is popular is because no one trusts the government to play judge, jury and executioner. The jury acts as a sanity check that is theoretically outside the government's control, so that the system cannot simply imprison anyone the government dislikes. Unless we get to the AI 2027 scenario of nationalizing the AI companies, Claude is quite capable of acting against the government's interests, and as a bonus it's probably better at discerning the truth than Joe Average.
Claude also has no skin in the game.
Claude has no skin at all. I'll take my chances with humans.
Unless there's an angry mob gathered outside the courthouse, jurors don't either!
This is just not true. If you are on a jury deciding another person's fate, you have a very strong incentive to make sure the system is treating that person fairly in case you are ever in that seat trying to defend yourself. The laws you're deciding on apply equally to you too. That's the whole point of a jury, not "am I going to be lunched when I walk outside"
Do you think the ordinary humans sat on the jury here had any skin in the game on this highly complex tax law fraud trial or weren't put off by the unlikability of the defendant who is one of the UK's top tax lawyers specialising in advising people in minimizing their tax payments (said unlikability which has zero and should have zero connection to whether or not he is guilty as a matter of fact and law):
https://www.ft.com/content/31ef87f3-c64c-4d87-9056-a009db0b99de?syn-25a6b1a6=1
Barrister accused of £2mn tax evasion faces retrial after case collapses
Jury fails to reach verdict in case against Robert Venables who argued he was ‘entitled to take advantage of the system’
Sure, you can come with lots of examples where juries have made wrong decisions. I’m sure if Claude was making those decisions you would have no shortage either. When I say skin in the game I’m talking about a shared value system and a shared stake in a society’s success, which by definition an LLM does not have, even before you get into training and alignment issues.
Every time I read stuff like this, I wonder if the net influence of “foreign aid” has really been that positive vs the counterfactual
"Foreign aid" is an umbrella term for a huge variety of kinds of programs with wildly different cost-benefit profiles. Some of it's great, some of it's terrible.
It's a small but important corrective to the effects of centuries of colonial exploitation which have warped local communities far more than any NGO ever did.
Eh, you still have to check that any given particular foreign aid program is in fact making things better and not worse.
A Boston University study estimated that several hundred thousand people died due to the gutting of USAID that would not have died otherwise, so at least *some* foreign aid seems to be useful.
Maybe I'm overly skeptical of the links between the NGO complex and the academic complex, but I estimate the chances of such a study coming out would be 100% whether it was actually true or not.
There's also the difference between: 'we have a long term principled stand against foreign aid' vs 'we turn it off without much warning willy-nilly'.
Yeah, this is ridiculous—it’s a very complex counterfactual that cannot be easily studied. And of course the authors of the study are developmental NGO types!
How likely does USAID cause wars? ... yeah, the study ain't getting into that. How likely does USAID hurt our diplomatic allies? ... yeah, the study ain't getting into that either.
As usual, first order effects are just the start of the equation.
This is a ridiculous study that is... probably politically motivated. See Trump's increase in funding for the Peace Corps in the current budget (which I highly doubt they factored in).
Don't disagree.
"Colonial exploitation" and "foreign aid" are of course just two different labels on the same cluster of things, and colonial exploitation has done a lot more good in certain places than foreign aid ever did. (And more harm, in some other places.)
That's up to the people living there. They seem to be able to tell the difference, and they would know.
They know that liberals are easy marks and all it takes to get paid is to play the victim with his hand out.
What about a real victim?
Singapore seems quite ok with her colonial past, and even Taiwan has mostly made friends with the Japanese, her past colonisers. I'm not quite sure about Korea, would be interesting to investigate.
(Of course, Taiwan is in a complicated situation with the PRC next door.)
That's good, but of course, both of those are prosperous nations.
Bro will read an entire book review explaining why "foreign aid" is basically neocolonialism and then say it's an important corrective
Once you broke it, you bought it.
Who is the "you" in this scenario. "You" is a back door to collective guilt and eternal occupation of iraq. "You" can just stop breaking things and let the people with a stake in fixing things take over.
The United States government. Foreign intervention (any intervention) cuts both ways, is what I'm saying.
What definition of 'exploitation' are you using here? Because in his book Empire, Niall Ferguson makes a strong case (with lots of economic statistics) that being colonised made the colonies of the British Empire richer than they would have otherwise been.
Having your freedom of choice taken away. You can't buy someone's liberty.
The British Empire did just said, when they bought the slaves out of slavery.
Not what I meant, of course. Enslaving them in the first place, then giving them a nice plantation to live on. Freeing them later is great and all, but that's just undoing some of the damage you did in the first place.
I have some experience of this, coming from such a country, only even less developed than the places mentioned in the piece; so we had different problems.
The thing to remember is that despite people in the west having closed their eyes and ears to reality since the 80's, class war is real.
When the man in the government truck came around to look at your teeth and give any vaccines and notarize any paperwork and take the mail that didn't go to the city on the milk truck and test the water and capture mosquitos to test for malaria (all one dude)
We liked him. We recognized him as a bit of social apparatus that served everyone equally, but us more equally because on the Majestic Equality scale between Rich and Poor, we were on the sleeping under bridge side. The rich didn't need any of that.
Likewise, there was an NGO that blew through town and made available techniques to diversify coffee plantation varatals away from just robusta, to make fields more resilient, and install fancy drying machines that were easy to understand and repair and basically free to run if you shoveled in the dried fruit and chaff and such. That one people jumped on with both feet.
On the other hand, there were NGOs that came through trying to pave roads and string power lines which we fought against and roundly rejected. We could see what that shit was about: you get a fridge, and we make your land legible to the capital markets who will then fuck you to death, but it's ok because of revealed preferences, and then you can't afford your fridge so you get jack and shit and the opportunity to be a rich man's bitch and work service at his hotel with no other options.
The towns that put their shoulders behind those programs are dried up husks of people waiting to die, with no prospects and no future. Everyone eventually got electrification anyway; if you let the various world bank flavored liberal organizations in, you get it a little bit faster in exchange for giving up all autonomy to be a servant of the market, and hence to be owned by whoever has the most capital and cares to own you.
As you might imagine, eventually the pressure of money got too high (and we didn't have access to good coffee land unfortunately), and we got the road and the powerline and the running water, and everyone lost their land, lost their income, lost their community, and either became effectively a serf to the developers or left, either to the big cities and towns or to the US. I was already a citizen so it was easy, other people are still trying to get their papers 30 years later.
It is what it is, and it is much nicer to have all these nice things, but I always wish the people of economics could have a real experience of what it is like to have the "natural" system of economics imposed on them instead of being born into it.
Can you explain more about how exactly electrification funded by foreign NGOs leaves locals worse off? I don't think I really understood that part.
My interpretation was that it's fine while electricity is free, but too expensive once the ngo moves on.
I didn't get how this causes people to lose their land and income, though. Do they not realize that they need to cut their losses and stop paying for electricity once it's no longer subsidized?
Good question, I'm not sure
And let the food in your refrigerator spoil? Stop cooking food? Can't use the internet to communicate with potential employers? Don't light your streets at night, now that more people are driving vehicles around?
Industrial society brings many benefits, but once in it's very hard to get back out.
It's not that hard, provided that you live in a country with lots of cheap land. Just move off-grid and become a subsistence farmer like your ancestors.
This is mostly not a thing outside of colonial states. The 'old world' is more or less fully accounted for; certain areas may *look* empty, but they're almost certainly claimed by a given tribe, or part of the migration route for herders, or some other claimed use.
"Just become a subsistence farmer" is not a good look coming from a resident of a developed nation. Everyone here, I notice, at least has access to the internet.
See the loss of corn growing in Mexico after NAFTA. Cheap land is one thing, but when you're being underbid by American Monoculture, it's hard to compete.
They can speak for themself, but I assume it's the classic story in which a largely self-contained village economy gains access to cheap commodities and manufactured goods, which put the local producers out of business, requiring them to pick up stakes and move to become employees of some larger firm in a city where they don't know anyone.
I have no idea: but people are living there somehow and making some change in that environment may or may not be good. You should talk to locals first. (Maybe having electricity does nothing for them, but makes land more valuable, and the gringos come in and buy it.)
Edit, is this the corollary to Chesterton's gate? "These people need a fence here, let's build it for them."
Electrification makes what had previously been a backwater living off subsistence farming attractive for modern businesses to move into (when they previously had no reason to develop Mud Patch #896665 in particular). Thus, the subsistence farmers get bought out, etc., classic gentrification.
Hasn't the broad arc of decreasing the percentage of humanity involved in subsistence farming basically been an unalloyed good, though?
No doubt it's been good on net, but individual subsistence farmers may still be incentivised to let the chasing tiger of modernisation catch the other guy first.
There are no unalloyed goods. All advances carry costs, and unfortunately the costs are not born equally by all people.
It's been a net good, but I doubt it's been completely unalloyed.
For instance there are surely places in the world where the big efficient farmers simply came along and slaughtered or enslaved the subsistence farmers. I imagine this is how most of the premodern empires got started.
As detailed, a decent chunk of the farming in the area was cash crops. Milk, Coffee, fine leather, exotic hardwood etc.
Food was mostly grown locally, but was traded in local markets, not subsistence farming by any means.
The issue comes when an area that has it's own patterns of subsistence that do not neatly interface with liberal capitalism comes into contact with liberal capitalism, and are forced to change in order to become legible to the market, and are then instantly annihilated and reformed.
Sometimes I imagine this is good; I don't think there were too many slaves 20 years on pining for the lash. In the place where I was it was inarguably bad, as proven by the fact that everyone who had the capacity to leave, left. Abandoning the place where they had lived for at least a couple centuries.
I guess the conventional economic story would be like: This village doesn't actually make sense because it has a bunch of potentially productive humans who are mostly just growing food for each other instead of being more efficiently deployed by capitalism. So the markets correctly break up the town to get everyone to be more efficiently used elsewhere.
That's a pretty sharp example of when economic thinking fails to appreciate what is being lost. But also I guess I don't wish my ancestors were left alone such that I was living in such a community instead of in an American metropolis. So it's hard to see what the right path would be, even if you could get everyone to agree.
The issue is that most of the people in said village did not get to live in an american metropolis, and never will under the current paradigm. They were forced out of their old lives, into new worse lives that they did not want then and do not want now, and they have no prospect of improving them ever, unless their government does so outside the rules of the game at great risk.
You can look at Venezuela as an example; people who were immiserated by capital despite their vast wealth in resources, who took the risk and discoverered new planes of immiseration to explore. They did not have a Deng.
Right, but their ancestors did - like everyone who lives in a metropolis descends from someone who had to leave a village. Again, I'm not sure how to weigh that up.
I think they're talking about two different phases of electrification, both with a cynical spin.
> if you let the various world bank flavored liberal organizations in, you get it a little bit faster in exchange for giving up all autonomy to be a servant of the market
Phase 1 is electrification funded by foreign loans (World Bank-flavored NGOs). The NGO builds infrastructure and defines a payment plan, then the terms on that plan force locals to engage more with the formal economy from a weak position (work service at his hotel with no other options)
>eventually the pressure of money got too high and we got the road and the powerline and the running water
Phase 2 is electrification of the holdout areas, funded by local or national government. Gentrification, or possibly eminent domain, improves infrastructure for the purpose of enabling densification. Locals priced out as a side effect (became effectively a serf to the developers)
On the other hand, how do you want to introduce civilization if not by electrification etc. Already Lenin knew that it was the gateway to an industrial nation.
None of my grandparents grew with electricity in their house or a flushing toilet. But all of their kids did. That is just progress, and it didn't make the kids serfs. If anything, it was the grandparents who were more dependent on local landowners. That tends to come with being a poor rural person.
The best way would be to allow local leaders and community groups plan the roll out, so that people have a chance to adapt to the change, and the supply of electricity (and its price) is at least partially under the control of the locals.
In the US where I live, power is supplied by a local utility, under the administration of an elected government. Every rate increase has to be approved by the city council, and there is a public discussion beforehand.
How many underdeveloped countries do it that way?
I don't think the original poster is talking of "developers" in the commonly understood (urban) sense. More likely it's agribusiness:
"the pressure of money got too high (and we didn't have access to good coffee land unfortunately)... and everyone lost [sic] their land, lost their income, lost their community"
Could be this too, thanks for the correction
The land is livable for the poor people because the rich people don't want it. Electricity means the rich people start wanting it, and take it. Class warfare.
Precisely. Because these poor people almost certainly don't have formal title over their land, as soon as it's worth taking it's taken.
Net better off in the long run? Quite possibly, but probably worse for those particular people (quite possibly in personal autonomy and dignity, even if they're a bit economically better off).
This occurs in Modern America as well. If you don't own the mineral rights, as soon as they're valuable enough, poof! ByeBye House!
The electrification is on balance good. It might lead to gentrification, but you can't not improve stuff forever incase someone else wants to live there.
The issue is the strings that come along with it; the liberalization and enclosure and forceful entry into the global market.
If the place in question has the time and opportunity to organically developed, you can avoid the worst of it. If a capitalist NGO (so, and NGO) comes in, you get it all at once and it's like shock doctrine but way worse.
it's a tale as old as time: poor people love fancy drying machines, but hate usable road networks and electricity
They hate being controlled by distant elites.
I hate it too, but I prefer it to poverty.
Imagine suffering both.
How is being controlled by your retarded countrymen any better (the fate of ~all third world democracies)?
It's marginally better, because you can hang your retarded countrymen.
"The fate of all third world democracies" India says hello. Of course, I guess it depends on what you mean by "controlled." I live in America, and it sucks. But better Trump than Xi.
> It's marginally better, because you can hang your retarded countrymen.
You can't, actually. The more diffuse power is the more difficult it is to act against those who have it. You can't hang the voters.
> "The fate of all third world democracies" India says hello.
In which way is India a counterexample? I don't know that much about it but from afar it certainly seems like their government does lots of extremely stupid things to appeal to the worst impulses and beliefs of their voters (overregulation, weak property rights, useless nationalist posturing, protectionism, agricultural subsidies and price controls, etc).
I'm Argentine and imo many of our big economic problems are downstream of us being so democratic (i.e. populist) compared to our neighboring countries.
You joke, but drying coffee after wet processing is a pain in the ass. The drying machine makes it so much easier and more reliable, and saves you from losing a crop to bad weather.
Oh, I believe it—I was just surprised by the "roads and electricity are not wanted by the poor" bit, heh (I sort of feel like... well, probably almost anyone wants 'em, surely...?).
They want them, but on their own terms. IE, you want 1 million dollars, but if that came with a coralery*
*but I get the proceeds of all your investments for life, all your property for life, and you are obligated to work for me for life,
it is less atractive.
Must be beautiful to not be a servant of the market. You just grow your coffee, eat it, wear it, turn it on/off, satisfying all your needs without having to become a serf to anyone.
Scott should put the like button back so I can signal that I approve of this, without having to leave a comment myself. I understand why he removed it though.
There are two browser extensions that add them back. Search for ACX Tweaks or AstralCodexEleven.
You can also code your own, or if you don't code, do it with the help of AI, any competent one can single shot it by now, it's just a simple POST command.
It's there on the Substack app (although the app removes the edit button, in exchange; God I hate apps–).
I'm really surprised to see this attitude on this substack in particular. Like, you would actually and truly rather have that? If I bought remote land somewhere and gave you a subsistence farm but you can't engage with global commerce (no internet, no electricity, etc), you'd take that deal?
People want different things. I personally wouldn't take that deal, but I'd be very surprised if nobody I know would.
It's not a subsistence farm. The ambiguous writing, implying that it was, is rebutted by iahtethis' own later replies. My sarcasm was triggered by that ambiguity (even dishonesty?).
It's monopoly power basically, or enshitification applied to public services. First they are nice to the customers, then they screw the customers in order to be nice to the local elites, then they screw the local elites.
There is a legitimate question regarding to what extent even the best meaning donors can help people trapped within a rentier state.
The way to shut down rent-seekers is LVT + UBI. In a sense, skip straight to screwing the local elites.
>The thing to remember is that despite people in the west having closed their eyes and ears to reality since the 80's, class war is real.
Yeah we decided we'd rather be rich and not adopt the mentalities of third world peasant NIMBYs.
You decided that it was acceptable for me to be poor in order for you to be rich, where me is a succession of places that are tapped for labor and resources.
Maybe they get some development out of it, but they need to take it from you outside the rules of liberalism, and then you say they are an enemy of world peace or freedom or some other self serving slogan. If you don't have a CCP or a Park chunhee or a Lee kuan yu, the peasantry retains forever.
I can't complain too much though, I came to the US and am now pretty rich, so I'm part of the problem. Although now the conservative tendency has the nerve to complain about immigrants coming according to the incentives they created, to be paid money in the free market they worship, by them. It seems there is no winning.
>You decided that it was acceptable for me to be poor in order for you to be rich
Third world peasant mindset does not understand positive-sum games.
If the people who accepted playing the game ended up worse off, doesn't that demonstrate trivially that the game was not positive-sum?
I mean, not defending Turok's tone, but wouldn't it still be positive sum if the gains of the people who benefit is greater than the loss of the people who end up worse off?
Sure, but you're pointing out a problem with equating "positive-sum games" with "good games". It's still possible to have positive-sum games that leave a lot of people wretched and miserable.
Another story about how NGOs with bad incentives have negative impact: https://lydialaurenson.substack.com/p/the-time-i-joined-peace-corps-and
<< I learned that when they started the new program in Lavumisa, they had trouble finding local sex workers, so they identified one woman who they believed to be a sex worker, and they asked her to help them find other sex workers in order to create the new Lavumisa support group.
Financial incentives are common among NGO programs in the third world. In fact, the incentives are so popular that many people who collaborate with NGOs have come to expect financial incentives, even if they are not employees. Hence, it seemed natural to incentivize the first woman to bring other women into the program by paying her per woman she recruited.
When I learned this fact, I asked whether they had any way of checking on whether the other women in the program were, in actual fact, sex workers. Samkeliso acknowledged that they did not, in fact, have any way of checking on this. Therefore, they had begun referring to the program as a program for at-risk and marginalized women, rather than for sex workers.
Everyone who was recruited into the program expected plentiful incentives to participate in the program. So the program provided meals and getaways to nearby fancy hotels, for the group to bond and do community things. >>
[Facepalm]
I'd watch that movie
absolutely, yep, that is how it goes. The budget must be spent. If the needed objective is not found on the ground, it will be made up.
Incredible. NGO pimping
Man. A feeling I sometimes get reading articles like this and Scott's reviews of books about the failures of state power is that it seems like all foreign aid is useless and the only path for a poor country is to take their lumps and suffer through the dark satanic mills until (god knows when) they get through to the other side. If there even is another side for them!
To be clear, all foreign aid definitely isn't useless; as the penultimate section mentions, well-targeted aid (particularly health interventions with highly robust evidence bases, that can absorb more money thrown at them without having to entangle their operators too deeply into the local economy) saves lives with high cost-effectiveness and sometimes even has a transformative impact at the level of specific problems. See, e.g., https://blog.givewell.org/2015/11/06/the-lack-of-controversy-over-well-targeted-aid/ and https://pepfarreport.org.
On the other hand, if you mean that foreign aid never elevates entire countries from poor to middle-income, well, that's outside my expertise but I don't happen to know of any specific counterexamples.
I think that's an unrealistic ask of NGO's. They are not, and cannot be, a replacement for effective and honest government. Systematic programs to improve the performance of entire national governments are far less common. (This is not inadvertent.)
I think a better conclusion is that foreign aid can be helpful on the margin but ultimately real development needs to come from the locals building up their own economy. There's plenty of success stories like that, including most of the postwar economic boom stories (Japan, Europe, Singapore, Israel, Korea etc all got extensive aid to build their early economies but found eventual success by building up competitive economies on their own). The countries that are still poor and receiving aid are selected to be ones that couldn't do it earlier, often for reasons which haven't gone away.
Yes. And then the question is "how can we help with that". The answers I can think of are: decreasing trade barriers from poor countries, allowing immigration to the west, setting up charter cities with developed-country norms. However there is usually opposition to these approaches, often for good reasons. And they have to be done by governments, not by private donors.
Allowing immigration to the west seems like the worst possible thing you can do for a developing country in that you'll brain drain it (it might be good for the immigrants or the receiving country, but not the people left behind).
Yes. This is one of the biggest problems with Mexico, honestly. It is so easy to go to America and get "good jobs" and not have cartels, why would you help your country get better?
Huge problem in the Philippines too I hear, many educated Filipinos making money and sending remittances back to the country rather than actually living and contributing to society there. And the next generation like myself don’t have any particular plans to move back.
Foreign direct investment can probably help significantly, especially if it involves tech transfer and training locals, so long as those investments actually make sense. Like Poland’s rise as it became a manufacturing hub for all of Europe.
Seems like the real way for countries to get out of poverty is to establish the legal and social institutions necessary to produce a Samsung or three, and aid is probably a small part of that when done well.
> If there even is another side for them!
This is the real question, development economics is mostly fake / handwaving / rudderless.
Nobody actually knows how to develop a poor country's economy, even with full buy in from every person and the government there.
All the posited recipes, like Washington Consensus economics, or Joe Studwell's Success Sequence, have massive flaws and abundant counterexamples (ie Vietnam has been following it rigorously for 40y and still isn't developed, and the 3-4 countries that HAVE progressed to "developed" status over that time period didn't do anything like either set of recipes), and are as likely to harm overall development as benefit it, seemingly.
Worse than that, education is fake, and although every country in the world is seemingly eager to pour infinite amounts of money into it the instant they have the tiniest bit of GDP, it actually negatively correlates with future Total Factor Productivity growth / performance, and doesn't actually improve "human capital" much.
So....uh...good luck!
Honestly our best bet is "inventing god and asking him for money" actually working at this point, so fingers crossed.
My review of Studwell with some examples:
https://performativebafflement.substack.com/p/studwells-how-asia-works-revie
Education is negatively correlated with increases in TFP:
9) Increased education has a strong NEGATIVE effect on total factor productivity. He controls and corrects umpteen different ways, using dummy variables, controls with similar countries, correcting for initial gdp per worker, and showing that the measurement is robust to sample, data, and technique.
https://documents1.worldbank.org/curated/en/859821468180273788/pdf/773740JRN020010l0the0Education0Gone.pdf
Education doesn't increase human capital past "basic literacy and math," ie something like grade 2-5 education:
https://frompovertytoprogress.substack.com/p/do-education-and-human-capital-drive
> Nobody actually knows how to develop a poor country's economy, even with full buy in from every person and the government there.
Yeah we do. You just need property rights and time. You can accelerate the process if you do exceptional things but you'll get there eventually if you have strong property rights.
> Vietnam has been following it rigorously for 40y and still isn't developed
Their average growth rate over those forty years was 6.4%. It just takes a while.
> Their average growth rate over those forty years was 6.4%. It just takes a while.
But why were all the east asian countries 10x faster? Or Israel, Chili, Ireland, etc also so much faster, normalized to the same starting GDP per capita?
> You just need property rights and time.
I'm pretty sure Afghanistan and Iran and all of the African countries have property rights, legally, so I think there's probably more to it than that. I also assume you mean "land and business ownership" because even the USSR and pre-Deng China had chattel property rights.
You probably mean "business and land and chattel property rights, AND impartial rule and administration of law, and relatively low corruption, and better than a certain production per acre agriculture, and to not be suffering wars all the time, and..."
But then you're back to a whole package of complex things, which have thus far proven to be impossible to achieve for a good 70% of the entire world so far.
> But why were all the east asian countries 10x faster? Or Israel, Chili, Ireland, etc also so much faster, normalized to the same starting GDP per capita?
I don't know, I'm not that familiar with Vietnamese policies. But it *is* working out for them, even if slower than their peers.
> I'm pretty sure Afghanistan and Iran and all of the African countries have property rights, legally, so I think there's probably more to it than that.
Well, they might legally, but that doesn't mean anything if your family members have a de facto claim on most of your income due to social norms, or if the government is incapable of enforcing your rights over distant property, or if your property is liable to be expropriated whenever the government wants to.
> You probably mean "business and land and chattel property rights, AND impartial rule and administration of law, and relatively low corruption, and better than a certain production per acre agriculture, and to not be suffering wars all the time, and..."
Yes, I mean the full set of things implied by "strong property rights". This includes chattel, land, and business property rights, a minimum level of state capacity necessary to enforce such rights, a "tolerable adminstration of justice" w.r.t. property, not unreasonably high taxes, no social norms that prevent you from ~freely disposing of your property, etc.
This isn't easy of course—I never said it was—but we know that it works and that it is basically sufficient for sustained economic growth/development.
Read this one on the site, but was too late for the initial vote.
I really wish the author hadn't felt the need to use an LLM to write a lot of the analysis. The personal anecdotes are great and obviously human-written, and then suddenly the analysis section has a completely different tone which totally bogs the whole thing down.
This would pretty easily be one of the top three if not for the LLM bloat.
Pangram says 100% human-written: https://www.pangram.com/history/a88ebb84-7e44-4737-aa82-4b1137b5eb9a
That's great, but my eyes and brain tell me that this:
> He describes several decades of frugal development—relying on foreign volunteers for key expertise; reusing secondhand equipment from Norway; insisting on locally affordable and appropriate technologies—and contrasts that approach’s positive achievements with later dysfunction and fragmentation. On Liechty’s account, the breakdown can’t be blamed on “‘corrupt Nepalis’ or ‘Third World incompetence’” but stemmed directly from the big-budget, highly-leveraged, and ultimately unaffordable approaches preferred by “commercially driven ‘development’ banks” (e.g. the World Bank) and their Western corporate partners.
Almost certainly came out of an LLM. This too:
> Western altruists could try to invest their donations only in areas requiring minimal entanglement with beliefs and power dynamics. That’s a defensible choice—not solely a drunkard’s search, investing in the areas with the fewest factors confounding our efforts at evidence generation. There’s unquestionably lifesaving work to be done in that space.
IMO you are overconfident in this. To my eye, the first quoted passage could have come from any pre-LLM piece of longform journalism (or an essay by a very good student). The second passage seems more LLMish, but still well within the range of things a human might write.
I am prepared to eat my words on this. My experience reading this was immense disappointment that the clearly-skilled author would insert such jarring and incongruous blocks of slop analysis, but maybe I'm just seeing paragraphs which were written on earlier drafts, or added at the last minute.
I'm with you on this one my eyes glazed over on these sections in particular. Just skimmed the last 10 paragraphs. Very poorly organized. Maybe too much AI use so copied their writing habits?
While I am not willing to straight up accuse the author of LLM generation as such in the absence of any additional evidence and am content to give them the benefit of the doubt, I am willing to say that that second passage made me pause and go "wait a minute, is this bit LLM output?" It is at minimum a kind of jarring pattern-match.
ETA: That said, I think there is also countervailing evidence and the people saying that the sentence structure isn't punchy enough for an LLM have a point, and the last sentence doesn't seem like a place where I'd expect most current models to use "unquestionably" as an adverb.
Neither of those two paragraphs remotely trigger my internal LLM detector.
I read far too much AI output for my liking (day job requirement) and am very sensitive to AI smell. That snippet wasn't smelly at all.
I don't think so. LLMs tend towards shorter sentences.
This didn’t come off as AI written to me at all
Hmm, to me this one had more of the "reads like it's written by ESL who learned to write in a somewhat outdated formal register" nature. It rhymes with Claudeisms but isn't quite the same consistent mush. #NotAllEmdashes. I believe Scott had a links post link about this recently, how certain cohorts in e.g. Nigeria largely got taught to do English in a way that just happens to track closely with LLM tells as of 2026, and this is very unfortunate.
The overall length was also short enough that even if it were actual LLM text, I don't really mind, in the sense of "didn't have enough time to write a shorter letter". It's actually really nice being able to read through one of these reviews in like a couple hours instead of day(s).
I remember the post you mention, and I've spent the last 45 minutes fruitlessly searching for it, because my recollection is quite a bit of pushback from people who pointed out that the author's writing style was far different prior to the advent of LLMs, and in fact was likely just using an LLM to generate the "AI Writes Like Me" article.
Okay, yeah, that's weird, I can't find it either. Not in any of the ACX Links Posts for 2026. Zvi mentioned it too, and the Substack search isn't great for his roundups, but I can't find it via DWATV either. And Google keeps surfacing suspiciously-samey AI-sounding articles about how AI detectors are rigged against Nigerian English.
Is it just hallucinations all the way down? Maybe it was longer ago than 2026? Now I wish I'd actually clicked on that post, so it'd be in my browsing history, which seems like the only reliable way to re-locate anything these days. Too much content on the Internet :(
https://marcusolang.substack.com/p/im-kenyan-i-dont-write-like-chatgpt
This was the post, and it includes strange typography errors that suggest it probably isn't fully ai generated to me, namely his inconsistent use of single vs double quotes and always capitalizing after a colon.
Kenya! I knew it was some African country, but got the wrong one...glad to not be found crazy after all, thank you. And it was older than 2026, damn. Life comes at you fast in the age of AI. Not actually present in ACX links posts for July or September of 2025...but I did find the "Scott contra Freddie on the involuntary commitment paper" comment I'd been looking for some months ago, on Sept 2025 links #48, so that's something at least.
Looks like Scott's mention was in the "Nostalgebraist's Hydrogen Jukeboxes" post from this May, which I wouldn't have thought to search. Strange when old posts go viral long after they're originally written. Can't find Zvi's mention, even though I'm sure I saw it linked there first, sigh.
I didn't get AI vibes; yes there were a few 'not x but y' but that's just a way a lot of people write! LLMs took it for a reason!
I wonder what the current state of Turing tests is. I quite frequently encounter people who seem quite confident they can identify AI writing by eye, and I always wonder how they would fare if actually tested on that.
>With the interests of the powerful defended more robustly than the rights of everyone else, a stark income gap has opened up between the top quintile and the rest of the population.<
Wow, poor Pakistan. I can't imagine what that would be like!
I am Pakistani plz check ur privilege
Unfortunately, you no longer have to.
The biggest thing this post left me wondering is exactly how prevalent these highly effective VOs actually are. I have no trouble believing that they exist, and that where they do foreign NGOs need to tread lightly to avoid breaking them. But this also rhymes with the kind of pro-bottom-up-collective-action political memeplex where I'm used to hearing a lot of wishful thinking, and generalizing from a small number of specific success stories to assertions that it'd work everywhere if not for specific malefactors. Unfortunately I'm not sure what exact evidence would overcome my semi-skeptical prior on this, though it's not an especially strong one as I don't know all that much about this stuff.
Also I'm highly curious to hear the author's opinion of GiveDirectly (and might ask once the contest is over).
I had the same thought. I'll note that in every example provided of big global NGOs coming in and ruining the local VOs, the problems were so bad that international groups thought they had to intervene, despite the presence of the VOs. That makes me skeptical the VOs were doing much.
Like, if you've got a rampant HIV problem, that probably means that the local anti-HIV VO is probably ineffective.
I'm sure there are some sectors where this doesn't hold true, and I acknowledge that even a marginal group doing something is better than nothing, but I still suspect this piece overplayed the value of these organizations.
"the problems were so bad that international groups thought they had to intervene"
Why do you start with this assumption? Why not start with the assumption that aid agencies and foundations have prioritized some cause, and looked for ways to channel their money to support that cause?
Ok, I just did some research:
- Turns out that the areas that receive the most international food assistance are are at, or near, famine conditions.
- Likewise, HIV assistance mirrors HIV rates pretty closely, with the majority of donations going to places where it is endemic.
- The data was less slam dunk here, but from what I see most aid related to providing clean drinking water goes to areas where there is a widespread lack of clean drinking water.
- Almost all disaster relief aid I could find went to areas that had recently experienced a disaster; I'm pretty confident this isn't a coincidence.
Snark aside - sure, there may be aid organizations looking to spend money, and probably some of those are more interested in shoveling money out the door than worrying about impact. That doesn't undercut my original thesis that if the original VOs were actually effective, there wouldn't be a problem for the NGOs to come and address.
But did you notice the Nepali example cited? Where a local VO was dedicated to helping people with HIV, and got outside funding. Then the donors directed their HIV spending to areas with more HIV, and suggested the VO switch to mother-child health instead.
So, one interpretation of these facts would be that the donors were prioritizing their AIDS spending to the most-impacted parts of the world, and encouraging a local organization to work on something with greater local importance.
An alternative explanation would be that the donors decided that HIV was a cause they wanted to support, and this particular VO was initially worthy of support because it was effective, but the support was withdrawn because Nepal didn't have such a bad HIV problem.
You seem to interpret all your examples as "donors direct their aid where it's needed most." I think an equally valid interpretation would be "donors direct their aid where their causes could use some money."
I think your missing my point: I'm saying that in the majority of cases, donors and NGOs giving money to try and solve a problem is strong evidence that the problem exists.
I'm not saying that they're doing so with maximum efficiency or that they couldn't spend their money in better ways. Hell, I'll even concede your point that some donors probably start with the money then go looking for places to spend it.
I'm also not saying that donors not providing money (or not continuing to provide money, as is the nepali example) is evidence that the problem has stopped existing.
My point remains though: despite the presence of the VOs, the problems continued and presumably grew to a point where they were seen as being sufficiently significant to warrant international support. Even if this support had ulterior motives, as you suggest, it still means the VOs didn't solve the underlying problems.
I think we've been talking past each other. Let me try again.
You seem to start from the assumption that NGOs identify problems in local communities, find local VOs working on the problem, decide that the VOs are not effective enough, and secure funding to try to make them more effective. I'm assuming that the donors like USAID and foundations are different from the NGOs like Oxfam, but that's probably not a crucial point. If I'm not correctly stating your position, please correct me.
I start with the assumption that donors decide on issues they want to address, and allocate budgets for them. They then make NGOs aware that money is available, based on the donors' criteria, and the NGOs work to find local groups that would qualify for the allocations and help them develop grant applications and tailor their operations to support the kinds of reporting and accounting the donors require. You assume that the aid is provided because the problems are growing, but I see no evidence this is true.
I think the difference between our assumptions is that you assume the choice of causes and recipients is based on some evaluation of the importance of the issue and the need in specific areas. I think that donors don't evaluate competing priorities in any rigorous way, but allocate some funding for each cause they choose to embrace, and they pay little attention to whether the needs in one area are greater than the needs in other areas.
I think the Nepali example supports my view. The HIV funding dried up not because there wasn't a need, and not because the VO was ineffective, but because the budget for HIV was limited and was being directed elsewhere. But there was money available in the mother-child budget, so if the HIV organization changed its focus, it could still qualify for funding.
As another example, leaders of international financial institutions decided 10 years ago or so to make financing for development projects contingent on nations adopting measures to reduce their greenhouse gas emissions. This was done not because greenhouse gas emissions were of any concern to recipient societies, or because greenhouse gas emissions were causing identifiable problems in these countries, but because Net Zero was a fashionable goal for the kinds of people who run international financial institutions. This is obviously not the same mechanic as NGOs working with local VOs, but it's another example of aid programs being designed and implemented to meet the donors' priorities, not the recipients'.
It's not so much that there were no problems that the VO's couldn't solve - of course there were, it's an underdeveloped country. The question is whether or not the helping strategy made things better, or worse, in the long run. There should be some sort of golden mean hybrid model that we should be aiming for.
I'm not sure "highly effective VOs" is the right sort of descriptive category to use here. A VO can be effective in that it produces a significant improvement on the conditions that would exist in its absence, while still having conditions be bad enough to invite international aid. There's probably quite a large range of effectiveness for organizations within that window, and I'd suspect that most countries receiving aid have their own local organizations that fall *somewhere* within it, because when put in trying circumstances, people usually try to organize and cooperate to get by.
How much those organizations are harmed by outside intervention taken in the usual form of foreign NGOs, and how much the communities suffer for their loss, may be highly variable, but I think it's reasonable to suppose that this would be a common outcome, and it might be more salient to ask, not "how likely is this to happen in a given situation?" but "how pronounced an effect are we likely to be looking at relative to the benefits we anticipate?"
An interesting read. So, the general approach if you want to help another country's residents seems to be, in order of preference:
1) Don't get involved at all if there is already a working organization down there, leave it alone, unless maybe if it asks you for some kind of help directly.
2) Help by providing knowledge, not resources, least of all money.
3) If you think you HAVE to step in, limit monetary incentives, and for all gods' sake, don't build your NGO a cool office in a big city.
Honestly the best way to help out is to partner with the actual government organisations that are already running the system (however imperfectly). Instead of spending your money to set up ad hoc vaccine clinics that go into remote places let the government know instead that you have X money which will buy Y vaccines and you want to pay them this money provided they can provide verifiable proof to you that they will use their network to buy and deliver vaccines to people. Less patronisation, more partnership. And this is something that will genuinely help because the government was here before you came and will be here after you leave, and so you don't need people to trust you or anything, you inherit the trust the people already have (or don't have) in the governmental healthcare system.
That only works if government is willing to cooperate and is not too corrupt. If it doesn't care about its citizen much, then either it would reject your help altogether, or use it to line pockets of ruling elites (if possible). A government that is willing, but unable to help its poor (and its usually poor people who are target of non-political NGO interventions) is a relatively rare thing. Or at least this is my understanding, since I don't work in this sector and only hear about it in the news. Still, I imagine that the reason that volunteers have to drive ambulances in Pakistan is not because the country can't afford a few more vans and drivers for them, but because unless this project directly funnels money to some minister's coffers, it's of no interest to him.
I think "Don't help people" is the wrong take here. It's more about politics and control. Who controls the flow of aid? Is it the people receiving it, or their trusted representatives? If not, then one should expect a degree of backlash and resentment.
Yeah, it seems to me that people are quick to take an extreme position -- either all aid is good or all aid is bad -- when the actual problem is how to make feedback work.
If people are doing a good thing, throwing money at them allows them to do more of the good thing (at least in short term). But...
* When you start blindly throwing tons of money at anyone who raises their hand, most of the money gets taken by grifters. In long term, all of the money, because the grifters outcompete honest people.
* When you try to micromanage the thing you know little about, you will probably create a system where the people who previously spent their time and energy doing the good thing now have to spend their time and energy doing the paperwork you request, and they can't make their own decisions about what to do, because now all decisions come from you and most of them do not make sense.
The question is how to avoid both these things. Step zero is to admit that these are the typical failure modes.
EDIT:
The solution proposed in the review (costly signaling by the volunteers) is also bad, in my opinion. Sure, when you see someone sacrificing their life for a cause, you may trust that person not to be a grifter. That still doesn't mean the person is efficient at what they are doing. Also, such person is likely to burn out, and then you have no one to work with.
But instead it makes me think of another frustrated quote "when we ask the west, we got lectures. When we ask china, we got money." I don't know if china is several decades behind the west in "charity technology" or what, but only providing knowledge seem to build resentment too.
I think this is talking about a different situation. The books, it seems, is all about local volunteer organization that don't actually ask the west for anything - the west comes them. The quote you mention is, I think, more about governments than VOs (I couldn't find the exact source for it quickly, but at least one page attributes a variant of it to "Kenyan official"). This is different: government programs don't suffer from volunteer -> salaried transition, they start this way, so negatives don't apply (but they have their own problems).
Our tour guide in Laos, an english speaking University graduate, told us his dream job is in NGO, because they pay the best.
Same in Ukraine. Easy money, few responsibilities.
Just think about what is does to a country, when the best and brightest go to work for foreign NGOs.
That, or they emigrate.
I have this article by Scott at the back of my mind. At it's core western civilization is so powerful it will transform anything it touches often for good. And many traditionalists understand this at a subconscious level and so avoid interacting with anything western.
https://slatestarcodex.com/2016/07/25/how-the-west-was-won/
...did you intend "for good" to mean "for the better," or "forever"?
For the better
Towards the end of that (amazing) article, Scott talks about how "just because people want it doesn't mean it's good" using the example of heroin. He doesn't really have an answer to this point in the end, leaving the question unanswered of whether the spread of western civilization is necessarily good.
Which non western civilisation would you prefer to live in compared to the West?
And does this civilisation rely on technology that can be traced back to the West?
This rings true to me. And its lessons apply closer to home, too. I work for an all-volunteer local homeless services nonprofit. We've noticed that peer groups in other communities that switch to a hybrid volunteer/paid-staff model (in order to attract and maintain hard-to-fill positions) tend to fall apart soon after.
This could be biased by those groups having preexisting staffing trouble which both caused them to choose the new model and caused their eventual collapse, so it's not slam dunk evidence, but it has made us very protective of our all-volunteer status.
We seem also to be able to provide higher-quality services more efficiently than similar professionally-staffed agencies in our area. Our volunteers tend to be more highly motivated, and our mission is less-clogged up by the bureaucratic needs of supporting a professional workforce.
Huh, that makes sense to me, either everyone gets paid...?* Or no one gets paid. (The local volunteer fire dept might hire out for services, )
*If everyone is getting paid it's hard for me to see how it's a volunteer organization.
In my area, a hybrid model for fire departments is very popular, where they have a small core of full time paid firefighters supplemented by volunteers. It seems to work alright.
Huh OK, I was thinking I should volunteer at my local volunteer fire dept. It'd mean stopping drinking for one or two nights a week, but that could be a good thing.
The thing that this piece ignores is that the underlying problems exist for the the same reason that outside attempts at intervention are so difficult. When you have low state capacity, you get rampant corruption and populaces with high reliance on communitarian association to non-state organizations. These factors makes doing basically anything is incredibly difficult.
My take is that the optimal solution is increasing state capacity. This approach has some moral pitfalls, see the cultural revolution as an example of state capacity improvement going bad, but when done right, is the only scalable solution to these kinds of problems.
Improving state capacity sounds even harder for foreigners to make any progress on than any of the stuff that the review alleges they keep failing at.
Only if you think at the level of a individual citizen or small NGO. From the nation state level, stuff like training local law enforcement and the judiciary; providing consultation on drafting legislation, providing security assistance to suppress would-be non-state actors are dramatically helpful at improving state control.
Ok, let's take a story I've heard from a local third-world country as an example. The sugarcane plantation is legally required to go through an inspection every year. The inspector finds problems, because the laws are loose enough that he can make up problems if there are none, and also no one is rich enough to keep things fancy. Now, the fee for this is so high the planter would be ruined if he had to pay it. So the inspector makes an offer: You pay half the fee to me personally, and I'll give you an official clean passing report. This is done so commonly that it is expected as the standard experience, and is considered no more corrupt than usual.
Tell me how providing training to local law enforcement helps with this?
(confidence: mid-high, this is a story I heard second- or third-hand many years ago)
My view is that the situation you described had two failures:
1. Badly written inspection legislation that allows inspectors to arbitrarily pass/fail anyone.
2. No reasonable accountability on the inspector against him demanding a bribe.
Better legislation helps problem 1, better trained and empowered law enforcement (not just cops, but also judges, prosecutors, etc ) helps address problem 2.
Sure - the rebuttal to my argument can be, "well, everyone is corrupt so any attempt at fixing the problem is useless." Beyond that being a general rejection of ever doing anything, it also ignores the fact that corruption can be an unstable equilibrium when it runs into systemic opposition.
1 is resting on the details more likely to be shaken up by the retelling. E.g. when I heard this I was pretty young and may not have understood economics very well, so "can always find something to fine for" could have been a simplification of "people are poor here and can't afford fancy things like not having cockroaches".
2 is the thing I am trying to tell you is hard. Like, society has already adjusted inspector wages to account for the bribes, same way waitress salary comes with an expectation of tips. And likely the fines have also been adjusted in expectation that only a fraction of them will be paid. And the incentive for the inspector to go out and inspect is the bribes themselves, so when you cut that off you also need a way to replace that. And who exactly do you expect to hold him accountable? The poor farmers, who would then have to pay more money they can't afford? His boss, who is probably even more corrupt? Sure, it would work if you could swoop in and replace everyone's heads with American values all at once. But you don't get to do that.
...And that might even get the problem backwards. Because this system works and is efficient, and maybe it's the best the country can afford. The farmers are incentivized to clean up their act a little, even if they can't get it all the way, because they'll be fined for more things if they don't. The inspector is incentivized to go out and inspect to get bribes. Probably the fines are even scaled to the farmer's income a bit, in a way that would be hard to do legally, because it makes sense to demand a larger bribe from a richer person. And it avoids getting more people involved, reducing the total man-hours spent on it. (No one wants to have to bribe the judge if they can work things out outside of court instead.) It's effectively a tax that funds law enforcement. Not as nice as the American system, but way cheaper.
One thing that jumps out at me is why the fine is so high that the planter would be ruined if he had to pay it. That seems like an excessive fine, and it's at the root of the problem. A second thing would be why the standards are set up so you have to be "fancy" to meet them.
If the fine were lower and the standards easier, the planter might feel like it might be better to make some trouble and go complain about the corrupt inspector, because even if he does end up having to pay the fine it won't break him and maybe he can get the inspector punished.
I don't know exactly what part of state capacity you'd have to change to fix that. Maybe get the sugarcane planters having greater participation in the political process that is setting these standards in the first place.
> providing security assistance to suppress would-be non-state actors
You're talking about taking sides in a civil war, or propping up a viceroy, and there isn't much of a bright-line difference between that and just rolling in to conquer the place outright.
Isn't "state capacity" only a solution if the state is reasonably well-intentioned? If all the state ever does is rob the poor of their meagre belongings to enrich the military and their rich friends, then increasing state capacity is just going to make them even better at doing that.
North Korea is very poor, but I don't think they need more state capacity.
Good point; I might be using state capacity as an overly general term here. I mean it in both the sense of a state being able to control all of its territory (which the DPRK definitely does) as well as craft sensible policies (which the DPRK doesn't do) and enact them successfully (which it seems like north Korea also can't do, check out this review: https://www.thepsmiths.com/p/review-the-real-north-korea-by-andrei ).
In this argument, I do think that the DPRK would benefit from increasing its state capacity.
Certainly, we shouldn't expect charity to solve very many problems at scale when the government is incompetent or corrupt.
Right - and this is my biggest gripe with charity. There are many problems the best solution is trying to motivate state-level action, not individual contributions.
I am not sure the local population will like foreign interference in their government even if it's for good. Example: the Portuguese went bankrupt in 2010. As exchange for generous financial help, IMF imposed government reforms - people are still opposed to them even if those reforms were absolutely necessary.
The right way to do it is to use responsiveness to citizen concerns as a metric of good government. Then everyone knows who the reforms are for.
I enjoyed this review not only for its insightful content, but also for the fact that it entirely fit within my email client. This allowed me to read the whole thing without going directly to the substack site or click "read more" and be shown horrible simplified text outputs in Gmail.
Great report. Thanks.
The basic theme seems to me to have much broader applicability. Outsiders, no matter how well educated and well intentioned, are likely to make mistakes that can cause serious problems for the intended beneficiaries. A few examples that come to mind:
1. Agricultural specialists teaching contour plowing to small villages in New Guinea rain forests. The locals had developed practices to plow up and down the slopes to facilitate drainage. The outsiders explained that this approach increases soil erosion. The locals tried contour plowing. Water didn't drain, but accumulated in the horizontal furrows. Eventually there was enough water to cause mudslides.
2. Western agencies swooped into post-Soviet Russia to help set up a capitalist system. They taught about corporate structures, accounting principles, and governance practices. They advocated for rapid privatization to put important assets in the hands of people with an ownership stake in making them productive. The assets ended up in the hands of ruthless opportunists who neither knew nor cared how to make them productive.
3. US charities often get funding from governments or charity umbrella organizations like foundations or the United Way. They have to adapt their organizations to have programs that achieve measurable outcomes in the donors' priority lists, often abandoning or de-emphasizing the activities that started them and earned community support.
I would say that this means that outsiders should do some field testing before recommending particular methods or interventions. It is the same as with medicine. We don't just try investigational medicines that have good potential. We actually have to do at least phase 3 clinical trials to show that the intended effect works and monitor the results after marketing as well.
The recommendations should also come with precise instructions how to apply them. With medicines it is easy – how to take it, at what dose, for what conditions exactly as diagnosed by symptoms and necessary tests, what should be monitored, when the drug should be stopped and for what side effects etc. Harder with other interventions but not impossible if the recommendations are based on evidence based practice and serious field work.
Politics can often interfere with these principles. Even in medicine it was evident with covid vaccine mandates even thought the vaccine did not stop infections and thus were not relevant for herd immunity but only for individual protection. We should aim to make our politics more science based, not less.
2. The capitalist system worked quite well throughout the Warsaw Bloc countries, notably in Poland. Only in the former USSR and particularly Russia did it fail. It is worth thinking about why this happened and how/whether it could have been prevented. (No I don't have a simple answer to this)
"Foreign Aid" has pretty much always been a tool used by developed governments to attain their foreign policy goals, whether that's helping out weaker allies, repremanding other poorer sovereign states by withholding it for doing somethat the developed state doesn't like or even just plain out improving things so there are fewer people willing to make dangerous journeys to the western states and making a claim on that western state's welfare system: much cheaper to pay poor people to stay put in marginally more bearable poverty than to deal with them when they turn up at your door in search of a better life.
Some of the people involved in the whole process as even well meaning; that's about the best and most positive thing I can say about the whole process.
And yes, the amount of damage "foreign aid" has done to developing countries is mind boggling. C.f. how when China sends very cheap steel over to Europe in huge quantities the Europeans call it "dumping", but when Europe sends over very cheap clothes over to Africa in huge quantities they call it "foreign aid".
I'm not saying all foreign aid is bad, humanitarian funding after disasters is about as close to an unalloyed good as you can get but the whole general structure of modern "foreign aid" very rationally often leaves a sour taste in the mouth of the receiving countries.
Great theory, let's see if it's actually true. Below is the list of the top ten recipients of US AID funding between 2000 and 2020. If your theory holds up, these countries should demonstrate high levels of historical responsiveness to U.S. priorities, either because they would modify their behavior to keep the aid flowing, or because the aid would be cut off if they stopped supporting u.s. policy.
I'm going to use percentage of time the country votes with or against the US in UN contested (ie non-unanimous) resolutions as a shorthand for the countries' level of alignment with the U.S., but let me know if you'd prefer a different metric and we can reexamine.
1/2. Afghanistan and iraq- I want to exclude these, because the aid was provided in tandem with active military operations and the state receiving the aid was subordinate to the US. You can, however, say that the aid was directly tied to US policy goals here, so I'll give you the points. I still think this is apples and oranges though.
3. Jordan: extremely closely aligned with U.S. policy in a way that differs from the region. Point for you.
4. Ethiopia: votes against the US in the U.N over 75% of the time, and there's no outside context that seems to demonstrate a high level of alignment. I'd say this is a point against you.
5. Kenya: votes against the US 80% of the time. I'd say this is a point against you.
6. Nigeria: votes against the U.S 78% of the time. I'd say this is a point against you.
7. South Africa: has one of the lowest alignment rates with the U.S. in the world, at only around 12%. They've also championed numerous actions against U.S. policy priorities. Strong point against you.
8. Colombia: only votes with the U.S. 25% of the time. However, south america tends to vote as a bloc in the UN, which might make this a bad metric for Colombia, especially since they seem pretty supportive of other US policy goals. I'll tentatively give you the point here.
9. Pakistan: definitely not; low voting alignment in the UN, to say nothing of what happened in Afghanistan. Point against you.
10. Uganda: again, they only vote with the US 19% of the time in the UN. I'd say this is a point against you.
So - in summary - this is the top ten list of non-defense recipients of U.S. aid; by your logic, these states should be the most vulnerable to influence from the US. Instead, the majority look like they're not particularly responsive to US foreign policy goals. Despite this, U.S. aid continued to all but one of these states.
It looks to me like your theory doesn't hold up.
This is quite persuasive to me. Foreign aid may cause more resentment that gratitude, especially in the top echelons of the country.
The very act of providing foreign aid indicates that the donor considers the local government weak or inept; few politicians would enjoy such insinuation.
Probably - especially if the aid is provided in a condescending fashion.
I think a bigger factor might be that, in certain cases, the act of supporting the provision of aid is sufficiently onerous that the politicians in the recipient countries don't really feel in debt.
The average politician in those countries doesn't really garner a direct benefit from the aid. Sure, his people eat, but the goodwill generated from the food (if any) will go to the country who's name is on the bag of grain, not the politician who enabled it.
On the other hand, the politician is 100% on the hook if aid goes missing, if a foreign worker gets attacked, or if anything goes wrong. Even worse, from his perspective, these aid programs aren't even going to pay him a bribe for his help coordinating all of this. Lots of personal risk, minimal personal gain.
In this context, I can see a politician not standing in the way of the aid shipment, but not feeling particularly loyal as a result of it either.
Foreign aid NGOs spread America's official ideology, leftist progressivism, of which anti-Americanism is itself a part. E.g. see South Africa, whose anti-colonialist DEI socialism is obviously Western in origin.
> just plain out improving things so there are fewer people willing to make dangerous journeys to the western states and making a claim on that western state's welfare system
Foreign aid given for this reason is not likely to make the country's elite's support the US. My other two reasons were just part of the very many ways in which foreign aid is deployed (how come so little US foreign aid to Eritreia?), there are other reasons "foreign aid" is commonly used.
In fact, "foreign aid" can actually be used to try and destablise an anti-western country's leadership because now the common people see all these "nice"/"rube" westerners providing them with money and stuff for free/cheap and thus can be an efficient vector for spreading western values (as part of western country foreign policy goals, see "democracy/freedom"), which the leadership of the country might not like and in fact might even make them more inclined to vote against the US at the UN if anything, all while providing the western country with a very easy screen of "we're just handing out aid, that's a good thing, right?" to hide behind. Imran Khan in Pakistan was very very against western foreign aid for very similar reasons, and there was force to his argument.
"how come so little US foreign aid to Eritreia?"
Because in 2005, in the name of "self reliance" Eritrea expelled almost all foreign aid organizations, including the United States. The country now rejects most foreign aid.
And before you ask - prior to 2005, Eritrea was a significant recipient of U.S. foreign aid. It seems to have had minimal coercive effect on Eritrean policy.
> before you ask - prior to 2005, Eritrea was a significant recipient of U.S. foreign aid. It seems to have had minimal coercive effect on Eritrean policy.
> in 2005, in the name of "self reliance" Eritrea expelled almost all foreign aid organizations, including the United States. The country now rejects most foreign aid.
These two statements don't coexist easily.
How so?
Pre-2005 Eritrea was generally antagonistic towards the U.S., but they accepted aid, primarily in the form of food.
In 2005, as the Eritrean government's xenophobia got worse, most foreign organizations, to include aid organizations, were banned and the country stopped accepting aid.
Perhaps there's more to the expulsion of foreign aid organizations than a fit of paranoia?
Yes, the US is a particularly poor example of this; USAID was clearly pursuing its own independent agenda rather than serving the interests of the elected parts of the government, that's why they got abolished.
Other countries have more internally-aligned governments and use foreign aid in more rationally self-interested ways. When China hands out foreign aid you better believe it's going to be in their own interests.
There's more to policy alignment than UN votes.
Specifically, Ethiopia's intervention in Somalia fit closely with US (former?) anti-Islamist foreign policy goals. As did Nigeria in the Boko Haram conflict and various ECOWAS interventions, as did Colombia with US efforts against drug trafficking and socialism. And Pakistan was on *both* sides of the Afghan war.
This leaves Kenya, South Africa, and Uganda. The foreign policy interests aren't as obvious (though Kenya did send "peacekeepers" to Haiti recently, and Uganda's been involved in looting the DR Congo, with American multinationals often at the end of that supply chain) but it's worth remembering that they are all high-population English speaking, poor countries in Africa (as is Nigeria), the very sort of places an aid program needs to be to maintain its credibility as a foreign aid program to the American public.
Right, UN voting alignment is a shortcut to try and get a legible number of how synced up a country is with US policy goals, but I'll agree it's not perfect metric.
I'm not saying these countries never end up working with the US, I'm saying that the thesis of "Foreign Aid" has pretty much always been a tool used by developed governments to attain their foreign policy goals" doesn't hold up to evidence.
That being said, nearly every point you mentioned doesn't work as an instance of the country just bowing to us foreign policy goals. And it's a stretch to say that one instance of win-win cooperation means that they are (secretly) beholden to America as a result of foreign aid.
Ethiopia and Nigeria were both motivated to accept U.S. security assistance due to a tangible threat of external and internal extremists, respectively. I think cooperation against mutual enemies explains these partnerships way better than that they'd been poked with foreign aid.
Colombia - I conceded this one in the original point, but I can also see reasons why cooperation on counter drug issues would have been popular domestically there.
Pakistan - right, both sides. Seems like the aid didn't cause their leaders to fully align with us policy goals. Pakistan supported the US when they thought that doing so would rid them of a security threat, and then worked against the US when they were concerned about losing ground against India.
These countries are motivated primarily by domestic political and security concerns. Foreign aid, if it's even considered, doesn't seem to serve as a primary motivator.
Your interpretation of the data seems flawed in an aircraft-survivorship-bias sort of way - absolute values vs. rates of change. As an analogy, consider blood transfusions in an emergency room. Look at the ten patients who received the most units of blood, and it wouldn't surprise me at all to find that they were also doing the worst in terms of how close they were to hypovolemic shock, or other problems blood transfusion is meant to solve.
Similarly, if foreign aid dollars have the underlying goal of *improving* a country's receptiveness to US policy, it shouldn't come as a shock to see that the most dollars are directed where the greatest change in that variable is desired, that is, where the receptiveness is initially very very low. And if foreign aid is often ineffective at its explicit goals, due to fundamental structural problems, the idea that more covert goals would also fail seems plausible, meaning the initially low receptiveness to US policy often remains low. Seems like that would adequately explain the top six, at least.
That only holds true over a single iteration of this game, or at least a small number. Sure - if you're starting from a blank state, the country you most benefit from trying to change is going to be the one you try to change the most.
But we're talking decades of sustained aid funding with no meaningful bend in the recipients policy towards the US. Plenty of time for the, to use your analogy, the doctor to realize that despite putting multiple transfusions into the patient, they're not getting better.
These countries have been large-scale recipients of foreign aid for a long time, and have shown no particular inkling of that having impacted their policy.
The much more simple explanation is that this aid is legitimately intended to address humanitarian concerns.
I think you’ve got the link between theory (“US foreign aid is intended to influence policy”) and prediction (“top 10 recipients of US foreign aid should show high responsiveness to US priorities”) exactly wrong. If the theory is true, then the US need not give foreign aid to countries that are already responsive to, or aligned with, US priorities (there’s no need: no return on the investment). Instead that aid should be targeted to countries that are the furthest from alignment with the US: for these countries there is a chance of a policy return on the foreign aid spent. This is exactly what your list shows, and so supports, rather than contradicts, the theory.
Note that from this perspective the “destroy local VOs and local trust” is not a bug of the foreign aid program: it’s a feature, especially when foreign aid is spent on “enemy” countries.
Maybe this argument would make sense of you just looked at single-year funding of aid, but it starts to make less sense when you consider that the aid programs to these countries have been running for decades.
Why do you think it “starts to make less sense”? The theory is : “US gives foreign aid to countries it wants to bring into alignment with its policies”. The direct implication of this is “US gives foreign aid to countries that are not aligned with its policies (in an attempt to bring them into alignment)”. The prediction that follows from this is: “over any time period, US foreign aid will preferentially go to countries that are not aligned with US policy”.
Maybe you think that, seeing that foreign aid does not immediately make a country aligned with US policy, the US will directly cut off aid? That doesn’t make sense, because if cutting off aid does anything it pushes the country further from alignment with US policy: contrary to the US goal. It also makes clear that aid is just a policy tool, not actual charity! (which undermines the use of aid as a policy tool in other countries).
Ok, quick question before we go on - do you think aid actually makes the recipient country change its policies?
I don't, which I think invalidates your argument, because the system your describing would collapse the second someone realizes that all this aid to unfortunately countries isn't actually resulting in policy goals getting accomplished.
It seems like your argument relies on the idea that the aid doesn't actually coerce the recipient country, but that the donor country thinks it's an effective tool. This situation I guess is plausible in the short term, but I don't think it would last in the medium-to-long term.
I think that foreign aid necessarily causes policy changes (any injection of outside money is bound to cause changes). I think those changes are typically not beneficial for the recipient country (increased reward for corruption, destruction of voluntary organisations, “lock-in” of the recipient role, misaligned incentives for the recipient country, interventions that actually don’t work, and so on). I also think those changes can be beneficial the donor country, but not necessarily in terms of policy change: it seems to me that the benefits come from the donor country getting unwitting “agents” in the recipient country, getting routes for funnelling money inwards to people aligned with the donor’s aims, getting positive advertising to third parties, and so on. I don’t think foreign aid actually helps the recipient countries, whether it’s government aid, NGOs, or personal giving (I think personal charity should be local, aimed at addressing problems that you can actually see around you - at least then the incentives and feedback for the giver are aligned).
> would collapse the second someone realizes that all this aid to unfortunately countries isn't actually resulting in policy goals getting accomplished.
Depends on the allocation mechanism and available alternatives. A very bad plan might nonetheless be the best available, and a strategic planner might budget a fixed pool of resources between various regional goals without strict objective ROI targets for each.
> “US gives foreign aid to countries that are not aligned with its policies (in an attempt to bring them into alignment)”.
That would be a very powerful incentive for countries to seek to be (and remain) as unaligned with the US as possible.
Admittedly, an incentive being perverse does not make it wholly implausible.
Depends on what kind of aid you are talking about. Military aid or commercial investment? Absolutely, often an explicitly stated goal of the aid. Humanitarian aid? Not so much. There hasn't been enough of it to significantly affect a nation-state's international orientation, for one thing, and NGO delivered aid is typically audited reasonably well, it doesn't directly benefit local elites, unless they are allowed to steal some of it. That doesn't mean it doesn't do any harm - but most of it will be of the unintended kind, like that described here.
My concern with this post is that it is entirely anecdotal. It could be the case that NGO culture destroyed indigenous VOs, but it could also be the case that they never would have gotten anywhere anyway. Or it could be the case that the examples the author cites are real, but a minority. It could also be the case that none of this matters and what these places really need is economic growth. Who knows?
I don’t think the international community should be in the business of funding something like AMP. That seems like a recipe for getting yourself banned from a country.
All that being said, the phenomena of money crowding out preexisting motivation in is highly studied in behavioral economics and is very much a thing. So it wouldn’t be surprised if this was happening in VOs as well. Likewise, the value of trust and social capital is a big deal in economic development literature.
So everything claimed here could very much be true and at the very least it makes me think. How does something like BRAC fit into this framework?
Have you read the book? (I haven't only the review.) It looks like a serious work. Many pages in the bibliography... I tend to believe it, but I know nothing.
> It could be the case that NGO culture destroyed indigenous VOs, but it could also be the case that they never would have gotten anywhere anyway.
This isn't actually responsive to the post, which observes that NGO influence removes capabilities that the VOs initially possess. It can't be the case that they never would have gotten anywhere anyway, because they're specifically selected for having gotten somewhere.
I gave this review a 9 or 10 when I read it months ago. Yet another reason for keeping charity local. (Or at least on a personal level... going somewhere to help, or helping support someone you know who is going somewhere to help.)
When I restack excerpts from this post, substack attributes the quotes to Scott Alexander.
Last year they were attributed to 'a reader.'
If you care, maybe change it, although I'm not sure if Substack allows for anonymous attribution.
I have learned a lot from this review, but I wish it wasn't so naive about the "three-self" church in China. It didn't so much weather the storm of mid-twentieth century as got co-opted by the CCP government, propped up as the only legitimate church, and used to control and suppress dissenters. It's not an example of building capacity for local collective action. Although this is probably a bit beside the point.
This was great. I immediately drew a parallel to the Schlozman/Rosenfeld “Hollow Parties,” wherein a certain kind of ‘professionalism’ dramatically decreases first the incentives and then the capacity for broad public participation. This in turn is the central driver of the org’s *legitimacy*, both to its members and in the context of the broader formative institution it aims to participate in.
Another point in favor of GiveDirectly
Is it? Or does GiveDirectly just cause different types of bizarre distortion?
Time flies, I thought it was earlier this year or perhaps last year...but the last big UBI paper review in 2024 mostly made me more confused on this matter rather than less. Why would giving money to the poors in a rich country show minimal or even negative effect, whereas giving money to the poors in a poor country shows decently large positive effect? Is it all just selection effects? Are GiveDirectly's various moral hazard safeguards a key ingredient? (e.g. funding an entire village rather than just select lucky study participants)
https://thezvi.substack.com/p/on-the-ubi-paper
Assuming you mean what I think you mean by "selection effects", it seems pretty plausible. Quoting Matt Yglesias:
"The explanation is obvious: Poor people in Kenya are average people who happen to live in an extremely poor country. Basic habits of hard work, diligence, and thrift don’t necessarily pay off in an environment where everybody is so poor that hardly anyone can hire you or pay for anything you make. Dumping cash on people in these circumstances really lets them level-up. By contrast, the domestic poor are — unless they are recently arrived immigrants — often people who, for one reason or another, are struggling to get their lives together in a very wealthy country. If they were thrifty and diligent, they wouldn’t be poor in the first place. Putting money in their pockets doesn’t make them thrifty and diligent, so it doesn’t really alter their lives that much."
(Though he acknowledges that this is an ex-post conclusion.)
https://www.slowboring.com/p/what-cash-can-and-cant-do
That was the other link I was gonna reach for, being certain that I remembered MattY discussing the same paper at the time, but couldn't remember enough keywords to search SB for it. Obviously it didn't shift his views overly much, since he still tithes the blog's profits to GiveDirectly!
Basically yes, it's not as simple as Success Sequence(tm) grifters would claim, but in a very wealthy country...if you're poor enough to be a good candidate for a UBI study, that probably indicates other ongoing issues besides immediate lack of wealth that won't be enduringly ameliorated by a one-time cash infusion. Big contrast to e.g. the review mentioning Ethiopian communities saving literal pennies' worth for an entire year, until they finally have a big enough pool to loan out. (Of course, one naturally wonders whether, say, Appalachians on disability who can muster the discipline to collectively scrimp and save for a whole year, along with their fellows in the holler, might be able to achieve similar results?)
Appalachians on SSI can't do that because of asset limits, if they try their disability gets cut off
GiveDirectly works primarily in poor countries, where studies support the efficacy of cash transfers.
This question of professionalism vs voluntarism and to what extent one should rely on each is a more general question, it seems to me. You definitely see the thinking with regards to e.g. schoolteachers that "we shouldn't pay them too much because we want only the most dedicated". An economist would tell you this is backwards, and certainly in that case I think the economists are right. But this seems to present a number of cases where it may be correct -- or at least, cases where the trust generated by the perception is so essential that this effect is more important than being able to attract people with money. So something to think about more generally I think.
(And I do have to wonder about research and to what extent it's more or less effective the more monastic it is...)
Yeah, it depends on what you want. Teachers are hired to teach, but also to build trusting relationships with the students. Which is more important? Volunteers are recruited to deliver some service to their neighbors, but also to leverage local networks based on trust? Which is more important? The answer might lean in different directions in these two cases.
Yeah with how people (especially here) still disparage normal schools even with properly paid teachers, I wonder if there's really a point there. Maybe passion is everything.
Seems to me that the underlying problem is that we don't know how to measure the output we care about, so we discuss input instead, and speculate whether low input correlates with high output, or high input correlates with high output.
But that's insane. As an analogy, imagine buying a car, but you know nothing about cars except how much they cost, so the entire debate is reduced to "people say that if a car is very cheap it is probably bad", and "well, if you pay a lot for a car, someone could sell you exactly the same bad car, only with a higher price tag", and then we endlessly debate about which side has a deeper insight.
If you want to get good schoolteachers, first you need to be able to tell the difference. Then you can adopt the policy of paying the good ones better. As long as you insist on paying everyone the same regardless of their quality, the problem will remain -- the highest quality candidates can probably make more money somewhere else, so your offer will most be taken by the ones who can't.
Sometimes the ones who will teach for peanuts are the passionate ones, and you want these. Sometimes they are the losers who can't get a different job, and you don't want those. And sometimes even the passionate ones decide that it's time to start a family and take a mortgage, but the bank won't give them one unless they change the job, so you lose them.
This is fascinating and resonates a lot with Seeing Like a State, certainly, but also with Hoffer's The True Believer.
Pulling a quite from a Samzdat post about the latter (because Lou Keep is awesome and I miss his writing sorely):
"Mass movements do the opposite: they pretend to give you power, while stealing what little you had; they pretend to solve your problems, while entrenching them. Movements make the frustrated more frustrated and they self-perpetuate with no regard for those who perpetuate them. The seed that they sow is frustration.
This is for one obvious reason: competition. Groups that maintain frustration, or are better at sowing it, will outcompete the others. The strength of a movement is directly proportional to its size and the fanatacism of its adherents, and the fanatacism of its adherents is directly proportional to the frustration they’re trying to escape. Mass movements that are good at what they do: a) make previously content outsiders frustrated; b) further frustrate their adherents while pretending to advance the movement. This means that the strongest mass movements are inevitably going to be the ones that are the best at not delivering the goods. Any movement that actually succeeds for (advances the interests of) its frustrated adherents will make them less frustrated. Hence, they’ll stop being members. Or it will succeed at its purpose, they’ll still be frustrated, and they’ll just join another."
https://samzdat.com/2017/06/28/without-belief-in-a-god-but-never-without-belief-in-a-devil/
I like a review that values my time as a reader. Mercifully short, briskly paced, minimal distractions and tangents, sparing use of quotes (not that quotes are a bad thing, but giant/repeated block quotes really pad out the length and funge somewhat against the summary function of a review). Would have liked to see the endnote caveat up front in some capacity: it's confusing to have instances of "I", "we", etc in an anonymous writing when by design I have no information about the author. Tell us a bit about yourself and your own skin in the game! Either way, it's a nice bookend with previous ACX reviews on the foibles of international aid and development. Seems genuinely difficult, and I'm coming to appreciate the wisdom of the mindset that says it's wise to invest defensively rather than chase maximal QALYs. Even if that sometimes means "ineffectively" donating locally rather than internationally - making damn sure the funds are used appropriately and constructively is worth its weight in gold dust. Charity begins at Home Depot.
(As an aside, I am loving how we're three reviews in, and there's been no shoehorned mention of AI or the Singularity yet. Knock on wood!)
Interesting. I was surprised in the USAID defunding how little people from other countries seemed to complain. There are plenty of venues for the leadership of other countries to push back against the US. For the tariffs for example there are many instances of foreign leaders criticizing US policy. But for the USAID cuts, very little. Something does seem fundamentally wrong there, and this book seems like it might offer an explanation....
Though I do expect more complains if the USAID cut led lots of rabbits to "stop dancing". Unless people have noticed the corpse for long?
Thanks to the reviewer for writing such an informative, interesting piece on a topic that very few cover.
I spent most of my professional career working with or for local government and non-profit organizations in the US. There are some broad similarities to what he is describing here, but it's apparent to me that we treat such organizations very differently from the way that international NGO's are run. For years the performance goldstandard was the local organization that could show success in local fundraising. The more local support you could show, the easier it was (I retired a few years ago) to obtain grant funding. Most grant funding today comes from private donors, either corporate foundations or wealthy individual donors. The most successful first step, I always advised my clients, was to get some local business leaders on your non-profit board. From there, you expand into the local volunteer network, and from there to a medium sized foundation with a history of working in your region on your mission related issues, and from there...
It was all scaffolding, all the way up to the annual revenue growth goals. Sure, even a medium sized non-profit has paid full time staff, but this is the US, where the culture trusts professionalism to a greater degree. But even here there are limitations, because many impoverished neighborhoods are served by very small non-affiliated churches, and those organizations tend to be rather suspicious of outsiders, largely for the reasons outlined here.
The point is that evidence of local engagement was mandatory if you wanted to expand your revenue base. It's also measurable, so a useful standard to use.
Take note however: "When a measurement becomes the target, it ceases to be a good measurement." Fetishizing *anything* because it is easy and simple to measure eventually undermines the entire effort. There is no substitute for local informed judgement.
That, ironically, requires that we trust the local leaders. That may be hard for some donors.
>With the interests of the powerful defended more robustly than the rights of everyone else, a stark income gap has opened up between the top quintile and the rest of the population.
The Gini index of Pakistan is about the same as that of Germany:
https://data.worldbank.org/indicator/SI.POV.GINI?most_recent_year_desc=true&year=2022
Yeah this immediately stands out to me. Sometimes even while the gap seems wide, it's still much less than how Western billionaires tower over other citizens. But I think people agree that western commoners still live better than other commoners.
This is in my mind whenever someone says "trickle down economics don't work!!"
wish there were any comments contesting the thesis of the book. I don't see why the political issue should be "foreign aid", all these failures seem preventable one way or another.
Human remark: "I tried helping them (by doing really dumb things), but it seems to always be harmful. Guess help is the wrong approach."
If everyone tries to help and none of it actually works, maybe it's not such a simple problem and we should reconsider the assumptions that brought us here.
And for some of us, this is in the "confirmation that water is wet" category. I've been hearing this message for my entire adult life, including from leaders who have been directly involved in many relief and charitable operations (associated with a church, but secular in aim). I've seen it directly in miniature when I've been involved with helping people both individually and organizationally.
I don't think the things were "really dumb". I think it's a really hard problem to solve.
There's many failure modes when you try to inject wealth into a poor society to make it less poor. But I guess most of them seem to come down to the same thing: people rearrange their lives to be at the point where the wealth is being injected. If you can solve that problem, maybe you can make things better.
Other people have mentioned GiveDirectly, but does that solve the problem? It seems it would come down to how GiveDirectly chooses its recipients. I've heard that they give at the entire-village level, and that they use criteria like "thatched vs metal roofs" to figure out which villages are poor enough to be worthy. Which sounds fine the first time around, but you can imagine that once word gets out then you'll have village big men going around and replacing everyone's roofs with thatch so they can look poor enough to get money. If GiveDirectly is still working well, it's only because it's small enough ($260M/year) to not be causing much distortion just yet.
The thesis is not that aid is bad. The review recaps various recommendations for handling aid more effectively.
This echoes my experience spending a year in an -istan working for an NGO. I'll add a few money related anecdotes:
Paid employees were transparently (and often successfully) using their pay stubs and resumes to angle for a Western visa, contributing to the brain drain of college-educated talent. Also created lots of turnover.
The lure of high paying NGO jobs could be an impediment to government relations and getting things done. For example, the regional governor made clear his cooperation was dependent on us hiring his nephew. This put us in the difficult position of either forgoing government support or engaging in the same kind of corrupt practices we were opposing.
One of the tasks our employees carried out was holding focus groups on various topics. The local participants (IMO reasonably, considering we were a monied Western org and they were dirt poor) expected to be compensated for their time. We eventually realized that we were not getting a representative sample of the population, because whichever employee we tasked to assemble the focus group would inevitably only invite members of his own tribe (often extended family). When I tried to explain to employees that this was a form of corruption, they were often confused because they viewed directing money to their tribe/family as a solely positive thing.
Now I really, really want Scott to write a post on "Foreign Aid: More Than You Ever Wanted to Know" that synthesizes the information in this review with the standard Effective Altruism defense of helping poor people/greatest good for greatest number, and gives us Scott's informed, evidence-based opinion on where/how we should donate.
Same. It feels like everyone and their mother has an anecdote about when foreign aid does or doesn't work, but at the end of the day I don't really have a model of what option has the better outcomes.
Same.
What if he concludes that we should just apply the Prime Directive and stop doing it?
I guess this'll make some commenters here very happy.
“Those village-level collective action groups proved far better able than the Nepali government to agree and mutually enforce limits on tree-cutting”
Government intervention often involves creating local committees. This sounds like successful government intervention.
One potential problem I didn’t see mentioned is that people like being part of something small because it feels like you’re making a marginal difference. Being one more volunteer in a group of 10: great. One more in a group of 100,000 volunteers less so.
This limits scope. Voluntary orgs can only get so big before, meaning you need 1,000 little orgs, which are impossible to maintain oversight of and assess for effectiveness.
My eyes glazed over as it became increasingly clear this was written by Claude. Shocked that this is a finalist. The thesis seems interesting but reading it felt like consuming paste
Somebody else got the same impression from this review:
https://www.astralcodexten.com/p/breakdown-in-pakistan/comment/300966896
I'm curious why so many readers thought this was written by an LLM (two comments on it seems like a lot to me).
I disagree, I just think it's ESL and a register learned in formal education. Pangram agrees, for what it's worth.
Didn't feel like AI at all to me (and multiple other people upthread). I guess it's score 1 for AI if we're now reaching the points where humans genuinely dispute over whether or not something is written by AI.
People have been commenting “this was obviously written by an AI” on any writing that actually uses correct grammar and punctuation and a reasonably rich vocabulary, ever since ChatGPT-3 came out.
I’m probably in the wrong here since Pangram and most people agree it’s human. Apologies to the author if so. But as Scott said in the post about taste:
“The [Kenyan guy who gets told he sounds like AI] argues that these are signs of learning a language in a formal, ossified, officially-good way rather than as the living organism of a spoken tongue. But he could equally note that his Kenyan essay test involves people with limited knowledge of English under high pressure to sound smart, being graded by other people with limited knowledge of English. This is the same levels-of-abstraction bottleneck that produces AI writing, and with the same result.
Think about the hallmarks of AI writing. “It’s not X — it’s Y!” Direct, simple, catchy, clarifying. This is a great writing technique, used in moderation. AI takes all the great easy techniques that work and can be reduced to a simple script, then overuses them until your eyes bleed. This is the essence of bad taste”
Don’t tell me it’s just good grammar. There are many examples of LLM-like writing techniques in this piece. Whether or not it’s the ESLers’ fault, the well has been poisoned. This style of writing is really irksome now.
As stated in the very paragraphs you cite, the problem with these writing techniques is not that they’re inherently bad but that AI overuses them to the point where they get grating. I don’t think this review author does that, and I don’t think no human should ever be allowed to write “it’s not X, it’s Y” ever again, no matter how appropriate it may be at that point in the text. But YMMV.
Strongly disagree, for what it’s worth. I don’t see any AI tells or the kind of language and repetition I would expect.
Best review and my prediction for first place winner.
Yet further proof Trump was sent by God. Wallahi he shut down USAID which was doing untold damage to not just American taxpayers but foreign countries.
I am proud to have lived through his presidency. I am proud everyday when I saw the signs in my supermarket. That welfare can no longer be used for candy bars
I liked it too, but I think it runs contrary to the belief system of many here. And so will not be first in the voting
“”“I hate having to post criticism of EA. Not because EA is bad at taking criticism. The opposite: they like it too much. It almost feels like a sex thing. “Please, tell me again how naughty I’m being!”
[…]
Many of the highest-upvoted posts on the EA Forum are criticism. The more critical, the more upvotes. […] A case against RCT-driven development aid, somewhat related to the one in Anti-Politics Machine, got 389. It was the #6 highest upvoted post of all time; the only reason it wasn’t higher was because stronger criticisms of EA beat it out.”””
https://www.astralcodexten.com/p/criticism-of-criticism-of-criticism
OK, I don't know EA at all except for the contact I get here.
That article is why I glaze over this review. It makes me afraid I'd get swept up by yet another EA criticism
Also best review in that it actually made me want to read the actual book
I spent most of the review thinking about The Man That Corrupted Hadleyburg.
I feel like this has broader implications for modern developed economies and the disaffectedness and lack of agency people seem to experience when structuring their lives around contemporary corporate work
Yeah halfway it feels more like Marx' alienation of labor
Good review and an important topic but I’m confused by this.
Tree monoculture bad: but isn’t that all modern farming (both tree and other crops)? Hasn’t it been massively successful overall?
Foreign interference with local agriculture bad: but isn’t that what the green revolution was?
Need to respect the local religious and cultural preferences: but also need to learn from missionaries whose whole goal is to overwrite those beliefs?
Need to demonstrate sacrifice: like Bankman-Fried? Okay that’s just an n of 1.
I am not even convinced about trees. As far as I know, most planted forests in Latvia and other places are monocultures. Environmentalists moan that it destroys biodiversity and they are probably right but I haven't really noticed any disasters resulting from planting pines or other monocultures. Any noticed issues are solved with small tweaks or certain interventions. Those forests grow, then are being cut down and replanted again. It basically works just like crops.
Maybe now they change the type of trees, or plant sequentially different types. Sometimes they treat young trees with fungicides or other agents to limit spreading of certain insects, fertilize the soil if depleted of certain elements. It might be ugly but it basically works.
The 3 issues I know about this are:
- Indeed loss of biodiversity, you cannot have a rich ecosystem in a forest uniquely composed of pines that you cut down regularly, in comparison to an older and diversified forest. The ecosystem is not only composed by the trees, but also of all the animals/fungi/vegetals dependent of those trees who also get affected when you cut them down and don't magically reappear once a new tree is replanted.
- Higher susceptibility to be affected by diseases. In a monoculture, a disease can easily spread across the entirety of the forest as all the trees have the same weakness against it.
- Higher susceptibility to fire. Pines burn extremely easily, see for example the current big fire in the southwest of France. A pine monoculture is probably the forest type that burns the easiest.
Given that green revolution seems to be the most successful thing on the planet, I'm in doubt whenever someone's bringing up traditional farming wisdom analogy.
I was utterly shocked to find out that Bangladesh has a higher per capital income than Pakistan.
Bangladesh seems far more functional these days.
The richest state in Pakistan has a lower HDI than the poorest state in India (and also Bangladesh)
It's not just NGOs in developing countries.
Speaking as someone who's been contributing to an open-source project for years, I have seen that project go from being entirely volunteer-driven, to being funded by Google (including my own salary) to being funded by several companies.
The funding helped put more work into the project, but it completely killed our community-building efforts. There was now a small clique of insiders (the Google-funded devs) taking all decisions, followed by several cliques of insiders (the different companies). Over time we spent less and less time advertising our work and reaching out to the devs communities that were interested in us.
I'm trying to keep it alive and keep working on it, but it's so hard stay motivated now that I'm not paid to work on it anymore. It feels like nobody gives my opinions any weight, and any work I do will be me getting exploited so someone else can get a paycheck.
I'm not sure what the solution is, beside moving on to other projects that haven't been compromised yet.
Not quite the same thing but this reminded me of Microsoft's Embrace, Extend, Extinguish.
Seems like the NGOs did this unintentionally.
Very interesting, thanks!
Glad this review made it. I had given it a 10. I worked with/at "NGO"s funded by our Gov and hated it. Loved the money, ofc. I did the same work for a private NGO (Robert Bosch foundation) before and though, nay because we got paid a third then, staff was on average much better (while the DAAD/GI do take on staff wisely rejected by Bosch) - see, if you are only willing to do it for a high salary, you are likely not really into the mission - things went much better, the 'bosses' were even interested in results not just "budget spent"; and staff was sent to places that actually asked for it,
while the "NGO" had to beg, bribe and cajole places to take their "helpers".
Excellent essay! Thanks for writing this. It seems we needed a “Seeing Like an NGO” but didn’t realize it.
I can imagine the truth in all the examples that have been shared. There is so much to be done, so many issues, that at times, the rabbit dance must be a reality... no matter how obnoxious it sounds.
Acknowledging the sharing and warnings shared via Breakdown in Pakistan by Masooda Bano.. I like to believe that there are Civil Society Organizations, Trusts, VOs who do keep (hold on) to the essence of their vision/ their goals.
Between 1998 and 2001, when I travelled to Odisha in India, the flight (Small aircraft) would only have representatives of INGOs and Government Officers.. i had thoughts then, as to, how the spirit and goals of such donors would bring change, and progress for this state. I feel very good to see Odisha now. It has developed and moved ahead in a number of indicators. Social indicators are good and a pro active governance show how ground level implementation is satisfactory. New initiatives, encouraging sports are always on the agenda. (Among other positive activity) Civil society actors continue to play a role. There are pockets that need more focus, that is accepted. I have seen this myself, over the years.. from 1998 to 2001, with a gap in between and from 2015, 17, 19 and 2023 and now..
I agree and appreciate how Bano has said this can be avoided.. can save the essence of the Community Based Organizations.. the examples of Pakistan say it all.
Community is a strong pillar in the work of many organizations. The tripartite of Community, Corporate /Donor and Government are ideal for scale. And it is the toughest.. with challenges mentioned so well in the review..
While this is a reality, there is a lot more that makes the whole truth...
Besides the sharing in the review, I am also touched by Social Welfare professionals in Canada , Sydney or London (as this is what I have seen). The professionals who work with the Dept of State, do their work well.. they believe in what they do.. they are trained and carry the spirit, working for over 10 or 15 years.. Does the assurance of funds/job security make this happen ? They do not get a whole lot of perks.. my sample of friends and acquaintances may be small ..yet, I marvel their at their service, commitment and objectivity.
You have given strong reminders through this read.. Well done!
In my public health course I studied healthcare in Afghanistan and my impression was that foreign help was highly effective. The foreign sources provided about $10 per person per year and that was basically all of its healthcare budget for this country. Most of this money was used by NGOs who provided healthcare services. The outcomes showed great improvements in life expectancy, mortality, especially child mortality and in all other measures. There was a need for the country to become self-sufficient and become less dependent on foreign aid and some other issues but overall that was a big success story how much could be attained by so little money.
This study greatly influenced my view on healthcare spending in general, showing how great is the curve that marginal healthcare spending has diminishing returns (the case in most western countries).
I haven't followed up what is happening when Taliban has taken over but a quick glance shows the country continues to rely on foreign help.
The biggest problem is not how to properly spend the foreign aid in most efficient manner. This example shows that it can be done when it really matters despite many anecdotal examples that it fails. The problem is that without changing country's fundamentals, it is not going to fundamentally change citizens lives to the better. I don't really know if I should be glad for Afghans having healthier lives than in the past or be sad for them (especially women) living under Taliban regime.
For all practical purposes I am in favour for them receiving foreign help to run their healthcare and making sure this money is used efficiently, yet I am not happy with the outcome of the country in general. Sometimes doing right thing does not bring joy.
I also have a different view about hashar.
The expectation was that the government takes over and provides a system where people get paid for their work. Unfortunately the government is too weak and that causes whole range of problems that this particular issue is just one of the many.
If this hashar work could be disrupted so easily, it means that it was never strong in the first place. Probably, it was also quite inefficient and any paid work is more productive and efficient.
In my childhood my father used to borrow a horse from a neighbour to till the private plot. My parents had full time work in the Soviet collective farm but we were allowed to use small plots of lesser quality land that were not usable by tractors. This system was common in the Soviet times and while not substantial by area, it provided significant portion of food for the family. It is not dissimilar to allotments in the UK except that it was economically relevant because the Soviet system had no functioning market economy and was plagued by frequent shortages and long queues in the shops. Despite smaller areas, poor quality of soil and needed hard work, they provided better yields per acre than collective farms. Sometimes even provided excess food that was sold to the state supplementing family income.
It is a textbook example how private ownership makes everything better than collective ownership.
But it was a hard work involving not only family members but often other people from the neighbourhood on certain days as the process of planting and collecting potatoes was very manual and required many helpers. It was done without renumeration, often relying on reciprocal help when neighbours needed the same help. Overall, it was still less productive than collective farms using tractors and other tools at scale.
The Soviet country ceased to be and private entrepreneurship was allowed again, and initial turbulences have passed, now agriculture is done not by collective farms but private farmers with rather efficient processes regardless of the EU subsidies. It makes no sense to work on small private plots even though some people continue doing this for fun.
Or not only for fun. Some elderly people are so invested and continue to grow their own potatoes. The neighbours are no longer interested in helping, so they ask their children and other relatives to come on certain days to do this hard work of planting or collecting crops. Everybody hates it. When someone does physical job daily, another day is just a routine. If one works in the office, that day becomes a torture. It rarely snows in the UK but when it does, the government has to remind people not to shovel snow too hard as many people get heart attacks from not being used to excessive physical exertion.
It is no wonder that helpers argue that economically it makes no sense as they can easily buy enough potatoes and bring to their elderly relatives. But only few people openly refuse. Elderly people shame their relatives for being such weaklings, then they appeal that the potatoes they grow taste completely different. I don't know about this, of all things one can grow in the garden, potatoes are most irrelevant by taste differences.
But the real reason is that elderly people are too old to change. They have grown their own potatoes whole life and stopping growing them means they no longer know what to do. Their lives lose meaning. So, the tradition continues. The family ties in the post-Soviet countries appear stronger than hashars in Taliban controlled Afghanistan.
Foreign aid poured into South Korea after the Korean war, largely from the US, all got stolen by the Chaebols, and then South Korea had an economic miracle (depending on your point of view, despite the foreign aid or because of it). All the effective altruism endpoints got hit: drop in child mortality, no starvation, boost in income, qalys skyrocket. There are many such cases, but the book seems to pick countries that are disasters now, then look back to when foreign aid started pouring in, not grab a historical sample of foreign aid floods and then watch forwards. This makes conclusion drawing much harder!
I guess the NGOs were motivated by Asian and German economic miracles, but got confused why it doesn't follow the same track? Seems like there's no more aid miracle after cold war.
The damage channel Bano describes is quite specific: foreign money flowing *through local organizations* reshapes their incentives and destroys the trust signals (visible sacrifice, accountability to members) that made them work.
Worth noting that the largest financial flow from rich to poor countries doesn't touch that channel at all: remittances, roughly $650B/year against ~$220B of official aid. The money arrives person-to-person, carried by the most legitimate signal possible, a family member's labour. No grant proposals, no workshopias, no salaried intermediary layer. Nobody needs to demonstrate charitable skin in the game for a wage to be trusted.
The review even contains the supporting detail: the Edhi Foundation, Bano's healthiest case study, is funded mainly by the Pakistani diaspora. Migration created that donor base.
So "let in more migrant workers, if you want to help" seems to dodge nearly the entire failure mode documented here. Two caveats:
1. Brain drain of exactly the civically capable (Hirschman's exit vs. voice). AMP's tenant farmers organized partly because they couldn't leave. But this bites much less for low-skill labour migration of the Gulf or Singapore construction-work variety, and enforced rotation turns permanent drain into circular migration: remittances during, savings and skills after.
2. Rent capture at the gate. When slots are scarce, brokers in Dhaka harvest the surplus as recruitment fees, financed by debt at moneylender rates. It's the same "large rents flowing through whoever controls access" pathology as the fire hose, just one layer earlier. Korea's EPS showed that a state-run zero-fee channel can mostly kill this. So it's an argument about how to expand migration, not whether.
Migration also doesn't substitute for the narrow slice of aid the review defends (vaccines, malaria work), since those are public goods you can't buy with remittance income. But for the broader goal of transferring resources to poor people without hollowing out their institutions, letting them earn the money themselves beats routing it through NGOs.
> In Zimbabwe, there was the religious VO that had worked for a long time against rural poverty. Following a major drought, it accepted international emergency funds and worked intensively for two years on famine relief. It helped a lot of people get food, while its staff pushed themselves overtime to comply with the intense reporting, quality, and anti-fraud requirements of the EU and UK governments, with help from international “capacity-building” partners.
> And after the disaster, when the VO went back to local supporters who’d always previously given to its work, it got the cold shoulder: “You’re an NGO now. Why do you want our help?”
I notice I'm confused. An organization taking extra aid, temporarily, during a significant famine--drove away its local support? That seems like quite a bit of an overreaction. What did the locals want, for them to just let people starve so that they don't get too big for their britches?
Just guessing here, but maybe previously the local people had mental model "if we won't help to make it happen, no one will", and now their mental model is "there are people out there who have several magnitudes more money than me, so it does not make sense for me to contribute; the NGO should try to get some more money from those rich donors".
Basically, just knowing that there are other sources of money out there reduces your willingness to spend your own money.
This is a different situation, but in Europe, many people's plans to do something prosocial start with step 1 "try to get some EU grant for that" instead of thinking how they could use their own resources (and maybe try to minimize the cost, instead of dreaming big).
That's fair, although it sounds different from what the review is describing.
Most edifying.
1. Most NGOs are basically a jobs program.
2. Outside cash corrupts a VO, turning it into an NGO/jobs program. Dutch Disease (easy money drives out every other form of economic activity) is most instructive here.
The problem is that the pendulum swings both ways. Today it is “NGOs are spending too much time having paid staff write reports”, tomorrow it’s “there’s no accountability, look at the waste and theft!” Ironically, if you asked most donors the latter is their primary concern. You’ll see some conservatives reading your review as justification for axing USAID and killing hundreds of thousands of people, but if you presented your anecdote about the Islamic school in isolation those same people would freak out over money finding a religion they don’t like.
My half hearted solution: make the paid locals suffer. The catholic solution is to insist on the self mutilation of monks, or at least the pain of Mother Theresa. In the modern secular world: consider making the report writing even more painful and tedious, so that their community hears the anguish from every excel sheet.
Their community doesn't see the Excel sheets. It's too subtle a hairshirt for the purposes we're discussing. :)
The metaphor of GHD aid as a monocrop plantation that erodes the natural local ecosystem, and then languishes when the common goods created by that ecosystem have been depleted, is useful but I think it mischaracterizes things:
(1) The local ecosystem in LMICs is often dysfunctional — perhaps plagued by extractive patronage networks that consume the undergrowth like an aggressive local species, or perhaps just stuck in a stable equilibrium at an early stage of development like scrubland at the desert’s edge that’s never had the nitrogen or soil depth to become anything richer [I had to do some biology googling to make this strained metaphor work :P]
(2) GHD aid needn't be a monodrop plantation — at its best it could be permaculture. But admittedly, at its best today AFAIK it's more like a patchwork of targeted grafted planted mindfully not to undermine what’s already working in the ecosystem.
What's the closest to permaculture style GHD interventions (that might work) that you know of?
(Stuff that comes to mind for me is more on the growth side then the health side like Africa Jobs Fund or Mushtaq Khan style industrial policy)