Interesting review... It does leave something to be desired when it talks about population. Neither the Romans, the Mayans, nor the Chacoans had any form of birth control. Modern society does.
Babies are a positive externality. The benefits are diffuse to the rest of society.
In modern rich societies, babies are underprovided because the costs of them are localized to the families.
So, if you want more young people, you either subsidize babies by making free childcare, child cash bonuses, cheap housing, free education.... or you attract immigrants.
Those are your choices. Immigrants or paying families. There is no third option.
Also, I think the growing to a good place and stopping does have an example. The Nordic countries seem to have nailed their public policy.
Every society has birth control. The "wrong hole" method works great. The pullout method is less reliable, but good enough to reduce the birth rate.
Generally speaking, preindustrial farmers have a high birth rate (because kids can start doing chores early), and preindustrial cities have a low birth rate (because it takes much longer for kids to start pulling their weight). The modern birth rate is low because we're more urban, and because farming has changed.
I think this only explains a part of why fertility has declined, and the record for subsidizing children isn't that good, particularly over the long term (often such policies can have a temporary effect that fades in a few years).
Also, I can think of other options that might work besides the two you name, they're just not available to Western democracies at present, as they would require either constraining rights or modifying the flow of information in ways that would be perceived as reactionary and outside the Overton Window.
> Those are your choices. Immigrants or paying families. There is no third option.
Isn't a cultural shift a third option? Some cultures have a higher emphasis on children than others (e.g. there's a high correlation between fertility and religion, even correcting for economic factors).
If societies tried to push a cultural shift towards having children, e.g. the same way that they pushed things like environmentalism or anti-smoking campaigns, I think that could have a meaningful impact on birth rates. Sure, maybe "just pay them" might be more effective, but it's not like government subsidy is the only way to shift a culture.
I think it *is* a factor, but not the dominant factor since culture is only partially heritable (e.g. religious couples will have more children, but not all of them will inherit their parents religious views) - and there things like large changes in cultural zeitgeist and immigration that are going to have bigger effects on 'cultural demographics' than relative fertility between the different culture groups.
Yes, but in a century or two, if these trends continue, they will be the single largest group. I once worked these figures out and it's true. Also, unlike other formerly fast-growing groups, they have not slowed down at all.
Paying doesn't work, as a look at the European countries which have all those things demonstrates. One problem is that raising children is expensive in ways which can't be solved with any reasonable amount of money; the parents have an essentially non-delegable responsibility for their children. Which means they must take personal charge, and take it seriously. Which is a task which takes the better part of 20 years (another change from earlier times with earlier ages of independence)
Not true. The Scandinavian birth rate is well below even the USA's birth rate. And once you start comparing outside the developed world, forget about it.
There is a third option. Remove women from the (full-time) workforce. Not very politically palatable, but China seems to be trying it currently, if informally so. Obviously not endorsing this.
I'd encourage you to check out the general trend. If you exclude poor African countries, there seems to be a clear trend. Why do I feel it's fair to exclude African countries? It's probably safe to say that the sort of work women are doing there is not so fulfilling that it would compel them not to get pregnant. No one is worried about putting their career on hold when it's backbreaking subsistence farming, etc.
Relatedly, employment is not the "problem" per se — it's a fulfilling career that induces many women to substitute away from having children. So you could hypothetically have extremely developed countries like Japan with low(er) female employment rates that have low fertility rates nonetheless because many of the women who are working are doing jobs they enjoy.
I did read something recently on how female employment works different in Africa. It seems geared to be compatible with having large families. Unfortunately I can't find it now.
This is very speculative, but I've read that a lot of Africa is matriarchal and women do most of the productive work. So, it would make sense for them to have less of a problem keeping fertility high as their female population transitions into a modern careerist workforce. FWIW, I think we will basically converge with them, although we will take a bit of a circuitous route.
In the West, careers were designed for men willing to work very long and hard in order to secure same-class mates with high standards for partner single-earners. This led to a system where the middle class worked very hard, and the upper middle worked even harder (think of the insane hours lawyers and bankers [used to] put in); obviously, such a system is incredibly productive.
This level of dedication is in no way compatible with fertility if adopted by women, which is largely what is happening now. Something will have to give.
My guess is that what will "give" is how worker productivity. Women obviously work very hard. But men can work harder as a group, due to simple evolutionary psychology. If you're a woman, you only have to work as hard as it takes to provide for yourself and your family. If you're a man, you have to work as hard as it takes to attract a good mate and hold onto her. In the West, this level was generally much higher than the "provide for family" level, at least in the productive upper classes. As long as you have an environment where women demand yachts, men will do crazy things like work 18 hour days.
With the entrance of women into the workforce, the first thing you see is male performance go down, as they have less sexual motivation to work extremely hard; sexual motivation is probably most of the motivation men experience, so when women's earning standards go down due to less dependence, you expect men to be less excited about working 16 hour days. Empirically, you are seeing this — men are declining in educational attainment and on-the-job performance.
The second thing you'd expect is a gradual decrease in productivity. It's not that women are lazy, it's just they don't have a good (psychological) reason to work *extremely* hard. They're not excited about working 16 hour days either. So, you see pushes for "work-life balance" and "mental health awareness" and the like. Pretty soon, no one is working at the level that used to be the standard, and in the long-run you'd expect productivity to drastically decrease to roughly whatever is socially deemed "subsistence-level" (house in suburbs yes, yachts no). At this point, work standards will be so low for both genders that society will converge to the African/Soviet model where women work while maintaining fertility.
The only potential problem is that Western societies depend on a level of productivity that is much higher than that seen in Africa and the historic USSR. So, there might be a risk of societal collapse. More likely though, living standards will just plummet to those enjoyed in future Africa / past USSR with everyone blaming their favorite scapegoat for that development.
I thought the rule of thumb for bankers in the (mercilessly regulated) Bad Old Days was "3/6/3"? As in "pay 3% on deposits, lend out at 6% and hit the links no later than 3 PM".
And int he Bad Old Days, law firm partners kept bankers' hours.
An interesting example : The German Democratic Republic managed to have both a much higher birth rate, and a much higher labor force participation of woman than west germany.
Both the Romans and the Greeks practised infanticide for unwanted children. Plus, many newborns did not live very long - so much so that the Romans did not name them until around 9 days after birth.
That said, this has no bearing on modern, civilized societies (which is what I'm assuming the author means by the 'complex' euphemism).
Almost every pre-industrial society had more or less tolerated ways of infanticide.
To mention one especially interesting example: In 18. century france and britain, infanticide was a crime, but mothers could offload their babies to foundling homes, which then would do the killing for them ( these foundling homes had horrible death rates).
"Those are your choices. Immigrants or paying families. There is no third option."
One choice is to re-establish a clan-society. If raising the young is the duty of a whole extended family, people choose to have more kids, as I've heard. Because if you decide to have only two kids and your cousins all have four then you'll spend more of your money and free time on raising the cousins' children than your own.
That seems like it has the "if everyone would just..." problem. IE, everyone will not just anything. Increased immigration and increased benefits for parents/families are both things that are conceivable as government policies, whereas even if you somehow managed to elect a government that wanted to reestablish clans I'm not sure how they would do it.
Oh, it's super easy. Just look at the private sector conveniences / laws / government benefits that managed to turn formerly clannish populations into presently non-clannish ones, then just tax / don't enforce / do a worse job providing those.
Heck, it's hard to fail at this really. Even if you fuck up SO badly that you accidentally collapse your own civilization, you will still have succeeded in reestablishing clans.
I guess if you're willing to fully reverse industrialization that might do it but holding power long enough to see that through to its conclusion would be challenging.
Another way to spin this same thought is that in a nuclear family society, the parents have to be able to take care of the kids alone at all but the very highest needs times, while in an extended family society, you can have help whenever your kids are being most demanding, as long as you help your relatives whenever their kids are being most demanding. Having an extra kid might be more manageable if you know that you never have to deal with the worst alone, and you might gladly help someone else deal with their worst moments in exchange.
Romans had plenty of birth control. Besides the ones Bullseye mentioned, there was also infanticide, which was widespread in most ancient societies and less controversial than abortion is today.
I think you're ignoring the role that culture plays in fertility (Mormons vs Atheists), but I admit that's difficult for the state to influence in non-dystopian ways.
I was just looking at the data and it does seem like subsidising child-rearing actually works, Northern/Western European countries seem to be doing better than Southern/Eastern European countries, and while culture and immigration probably plays a role, the fact that the economy and social welfare programs are doing better in those parts of Europe must be making some contribution.
The premise sounds reasonable (states or whatever get bigger and more complex until they get too unwieldy and expensive to maintain, then they fall apart) but the conclusions sound a bit off.
"The agriculture section is full of graphs that look like this, showing the more work you put in the less yield you get per unit of work"
Yeah, I really want to know more about this, because it sounds on the face of it like "the idyllic hunter-gatherer lifestyle where you just strolled around for a few hours, picked berries, shot a deer, and that's your day's work done, then some idiot invented agriculture where now mud-bedaubed peasants toiled for hours in the muck for a few bowls of gruel", and I don't know how well that holds up.
If there is productivity loss, that may be down to: (1) all the good land is now being farmed, so new farmers have to move onto the marginal land, which is less productive (2) population increase means that instead of 100 kg of rice for 10 families, now you have 100 kg of rice for 20 families which is less return on labour.
And would you call, for instance, the collapse of Egyptian civilisation under the onslaught of the Sea Peoples due to complexity?
Tainter's weak answer to "why don't societies just stop at the perfect size?" may be the only one available; there is no "perfect size" or if there is, we don't know it at the time, we can only look back afterwards and say "yeah, when the population was six million in that city, that was the best time to stop there". If you try keeping the population of the city down to six million at the time, well - look at current debates on immigration. And unless the city turns into Detroit so that there is natural shrinkage due to economic collapse, you are not going to be able to reduce surplus population (indeed, you get the opposite problem - too great a loss). *At the time*, there is no easy way (maybe no way at all?) to say to everyone who wants to move to the city "Thanks, but we have calculated the perfect number for our top population range and we've hit it, so you can't come in".
>"Thanks, but we have calculated the perfect number for our top population range and we've hit it, so you can't come in". That's the way the Soviets did it, by the way. You couldn't just move to the city - and industrial plant or some other entity had to "sponsor" your move so that you could get a "propiska" (permit to change residence area). The terms were "limita" for people who were recently hired and immigrated to the city this way. Think the current quotas for foreign labor in most western countries.
It exists but it has become much, much less restrictive. The one thing that it does affect is your ability to draw local benefits - think being able to access some England-specific tax credit if you're originally from Scotland. You need to first establish your residence in the relevant region via some process.
Propiska itself no longer exists, but residential registration with the Ministry of the Interior is mandatory, and needed for access to municipal services.
If you own your own place, then no problem: register yourself. If you rent, then in principle, your landlord is obligated to provide registration for you, but in practice, many landlords are reluctant, because they don’t understand that laws have become simpler, or because they don’t want to get caught not paying taxes on rental income. The easiest solution to this problem is to have a friend who owns his own place virtually register you, which can be done by filling out a form at the post office.
This last option is not however available to foreigners like me, so you have to jump through an additional hoop, such as getting a friend to go with you to the local office of the Ministry of the Interior.
I think you have to break production and demand out into different processes in order to see this agricultural phenomenon in present society rather than talk about it in terms of a person's return on agricultural effort
On the production side, industrialized countries certainly aren't seeing a decline in calories-per-farmer-hour, but they are seeing an accellerating burden of complexity to sustain growth in crop outputs.
On the consumption side, there is a population to be fed which was bred up rather quickly to feed the demands of Fordist industry a large fraction of whom have little to contribute to growing and sustaining that complexity.
"And would you call, for instance, the collapse of Egyptian civilisation under the onslaught of the Sea Peoples due to complexity?"
I believe Tainter would suggest that the Sea People are the proximate cause, but that the ultimate cause is declining marginal returns. Eventually they become sclerotic, overextended and unresponsive to threats and opportunities. A similar answer could be applied to the various theories on the Mayan collapse.
In my review I brought up the Mycenean Greek civilization collapsing with the invasion of the Dorian Greeks. It's been years since I read Tainter, but I think he discussed that example but still insisted it followed his rule for collapses.
If you put in more work to extract resources *by the same method*, you get diminishing returns. But modern agriculture makes us fat with only a small percentage of the population producing food; this used to be completely impossible.
This looks to me like a major hole in Tainter's theory, though I haven't read the book myself.
I think that the implicit assumption is that things like the "green revolution" are finite and will one day be exhausted, leading to diminishing returns and unsustainable complexity
I don't think it's a hole - I think these sort of technology improvements shift the graph *up* (by increasing complexity), but it still has the same shape: increased effort leads to declining yields.
You can produce much more food with a fraction of the effort with modern techniques, but eventually they still hit diminishing returns when you have to start cultivating inferior land or whatever.
Which is why it's an overall race between our ability to increase complexity and push the graph up in order to keep up with the needs. I think currently (contra malthus), we've got a really good lead in that race, so I don't think it's a super pressing concern , but the fundamental dynamic sounds correct to me.
To be fair to him, Tainter does mention that science (and therefore engineering) also has diminishing returns, so that each agricultural advance has lower ROI than the next.
> Yeah, I really want to know more about this, because it sounds on the face of it like "the idyllic hunter-gatherer lifestyle where you just strolled around for a few hours, picked berries, shot a deer, and that's your day's work done, then some idiot invented agriculture where now mud-bedaubed peasants toiled for hours in the muck for a few bowls of gruel", and I don't know how well that holds up.
The point is *within a given system* more effort leads to diminishing returns.
Like you pick a few berries, that's great and easy. But very quickly the nearby berries are exhausted and you have to traveling farther and farther to find more berries. You need a change in system (an increase in complexity) to get more yields.
> And would you call, for instance, the collapse of Egyptian civilisation under the onslaught of the Sea Peoples due to complexity?
> In the specific context of the Middle East, a variety of factors – including population growth, soil degradation, drought, cast bronze weapon and iron production technologies – could have combined to push the relative price of weaponry (compared to arable land) to a level unsustainable for traditional warrior aristocracies. In complex societies that were increasingly fragile and less resilient, the combination of factors may have contributed to the collapse.
The answer to why societies don't just stop at the optimum size is that the optimum size is dynamic, you can be at optimum size with policies to hold at optimum size and then the big drought comes and diminishes your ability to collect energy via photosynthesis and thus the entire basis of your society is undermined. In modern society this can be other limiting factors causing fluctuations in required inputs, fossil fuel energy returns in energy invested, or imports, or supply chain disruptions etc ...
"The USA’s reliance on immigration to supply young people who can keep the economy running seems a little alarming just because it can’t be kept up as the USA’s relative desirability declines, but it doesn’t have declining marginal returns, exactly."
And
"The US national debt is growing. That seems obviously concerning to me..."
The USA doesn't need young people to keep the economy running, it needs young people to keep the ponzi scheme running. Debt is a word that implies repayment. There is no debt, for the national debt will never - and you can quote me - ever be repaid. Or at least not in the way that you think. Maybe it will be repaid someday, and that will be suddenly and climactically, all at once, with interest on interest accrued and due with all interest and principal and from each citizen, immediately. The declining marginal return (or diminishing return) is absolutely apparent once you realize that the importation of more immigrants to keep the house of cards intact becomes less efficacious when you import more people while simultaneously spending the new imports' futures as well, and at a new logarithmic rate. I wonder if immigrants realize that their futures are also being sold out from underneath them and at a faster rate than the native citizens they're being brought in to vouchsafe pensions and retirements for.
This is a frequent topic of discussion at Marginal Revolution in the comments section. My position on this subject as been quite developed and unchanging for some time.
Something that cannot go on forever won't. You don't have to look at the USA or China for signs of a coming demographic calamity, and not everyone on planet earth will be able to made US citizens, most certainly not without someone with the intelligence and money to make a bet against the US economy and dollar noticing such an opportunity. The US is currently paying to import foreign fertility because foreign fertility is currently cheaper, for now.
Does the IQ shredder actually hold up in terms of conditional fertility?
That is, while (fertility|Singaporean resident) < (fertility|origin state resident), for the IQ-shredder argument to hold what you need is (fertility|IQ150 Singaporean nth-generation immigrant) < (fertility|IQ150 origin state resident); has the latter been demonstrated?
(If the latter is not true, then either Singapore's native population just suck and are successfully replacing themselves with fertile, clever immigrants *or* IQ is actually self-shredding.)
My understanding is that conditional fertility is indeed lower in big cities (which were also notorious population sinks prior to modern sanitation systems). Fertility also seems to be more affected by years of education rather than IQ.
"Fertility also seems to be more affected by years of education rather than IQ."
This. I've seen this stated elsewhere. The world over most women with an undergraduate degree (or its equivalent) are not going to get into anything serious with a high-school graduate (its equivalent or less), or a woman with a post-grad getting serious with an undergrad.
I don't think it's a simple matter of mate choice. Instead there's a sort of extended adolescence and delay of childbearing during the most fertile years. There's an idea of marriage & childbearing as a "capstone" event once you've got everything else in order. That contrasts with the mothers profiled in "Promises I Can Keep".
Yes I think that's part of it. I was also speaking more broadly that statistics seem to show that rising education in women globally seems to portend lower fertility rates. For US, Canadian, and European women I think the 'capstone' idea holds, but there are other places where the cultural phenomenon isn't present, yet higher rates of female education produce the same results.
Yeah, but those are basically irrelevant to the IQ-shredder hypothesis (someone who's capable of emigrating to Singapore is probably in a big city anyway, and is already educated).
It's relevant to the general pattern of 1st world countries with below replacement fertility importing people from countries with higher fertility rates. Often those people will have higher fertility rates after they immigrate than most of the population of their new country... but their children are another story. I haven't discussed the cultural aspect of fertility, but that's a big deal, and unless people adopt some form of cultural separatism (like Anabaptists or ultra-orthodox Jews) they will tend to be absorbed into the larger culture.
"Debt is a word that implies repayment. There is no debt, for the national debt will never - and you can quote me - ever be repaid. "
Debt is being repaid every day. Go to treasurydirect.gov and you'll notice that every bond has a maturity date on which their face value is repaid to the bondholder. The US has never, in its entire history, missed a payment. It is true that the US will always have some outstanding debt, but that's no more concerning than you always having a balance on our credit card.
The balance on the credit card (being a revolver) is A) not a fiscally sound policy and B) is not feasible if the cost of borrowing that money changes (and it will someday) and C) doesn't really = "being repaid" if you keep adding to the balance, and the interest, every single year.
The US has never, in its entire history, missed a payment but it will someday, and that day will be will be, putting it lightly, quite memorable.
Yes, national debt is a problem if it's large, keeps on growing faster than GDP, and the cost of borrowing increases. Granted, those will probably be true for the USA for the forseeable future. But it's a growing debt burden that's the problem, not the fact that the debt will never be repaid. A small debt that's never repaid is no problem whatsoever.
For the sake of argument, let's say $1 billion. There is no problem whatsoever with the USA having a debt of $1 billion that's never repaid (or rather, continuously renewed).
Since the Federal Reserve can create US dollars at will, the United States cannot default on its debt, in the sense of "can't pay, won't pay." (If the Fed balks, Congress can amend or terminate the Federal Reserve Act as it sees fit.)
What is more likely is a devaluation, aka a "soft default".
"Tax rates go up more often than they go down." is a failed prediction. And one that required him to dismiss the Reagan tax cuts as being insignificant.
If he wrote this today, I wonder if he would say something like: 'It is easier to lower taxes than to raise taxes, so the the budget deficit will continue to grow.' Your theory should not be able to predict both increasing and decreasing taxes.
I think this is a bit glib. The Reagan tax cuts, like the Bush tax cuts and the Trump tax cuts, were fueled by increased debt, which isn't sustainable over the long term. From a 10,000 foot view, it's basically correct to say that the trend in the US in the 20th century was towards a bigger government that spent more and more.
This is an excellent review of a book which I read quite a few years ago, but just recently re-read earlier this year (oddly, a lot of the ACT book reviews are of books I have read recently). I agree completely with the summary, and I am impressed by the thoroughness of the review. I usually type up a brief recap of everything I read, but I am certainly going to add a link to this for future reference.
I too was frustrated by the authors choice of the three in-depth collapse examples. Rome was great, but the others were unfamiliar to me, and both lacked any written records to substantiate or at least "flesh out" the theory.
My take on the matter is that Tainter, like Olson before him, is highlighting a central tendency for organizations to suffer from declining marginal returns, rent seeking, bureaucratic growth and sclerosis. This makes them less adaptive to shocks which are inevitable over the long haul. Eventually they are probably going to collapse.
The only part of the review which lost me was in the final two paragraphs on the negotiating position or power of developed countries. I have no idea what this is referencing.
Why would an increase in complexity in a civilization/country lead to it being less adaptive to shocks? From what I could gather from the review, an increase in complexity is itself a demonstration of superior adaptation skills (governments increasing taxes, which leads to corporations to find clever ways to evade taxes, which leads the government to formulate clever-er ways to collect said tax, ad infinitum). Perhaps being less adaptive is a result of too much complexity, which leads subunits that are in competition to finally "give up" trying to adapt to change?
Also, I don't know if this is super relevant to the topic at hand, but I'd be interested in book recommendations that you might have for the summer :)
There is only so much potentially available surplus. If there is a ratchet effect for some interventions, then over a long enough time horizon you'll run out of headroom and you'll hit a shock that you can't overcome.
It isn’t complexity that gets them, it is declining marginal returns. This stretches them to the limits of growth (and internal rent seeking) and then a shock such as climate change, internal revolutions, barbarians or whatever hit and they are unable to respond. It’s like a balloon with everyone in society trying to gain as much space as they can until they reach the limits and it all pops.
For suggested books, I think a great place to start is the list of non finalist book reviews that Scott published a month or so ago. Some of my recent favorites are Heinrich's last two books. For an older one that I don’t remember being reviewed here I might suggest Beinhocker's Origins of Wealth.
Perhaps the phenomenon that we're trying to capture can also be expressed with Game Theory, in which two competing subunits try to outmaneuver the other, increasing the complexity of the game, and thereby blinding each other to the actual enemy, whether it be an enemy at the gates or climate change. Of course the competing subunits could cooperate and mutually agree to withdraw, reducing complexity and friction (like the government deciding to freeze tax rates and the corporations agreeing to honor them). However, what ends up happening in reality is closer to an iterated Prisoner's Dilemma, in which a useful working assumption is that the other side is only trying to outsmart you and that you should try your hardest to best them.
Part of the ratchet effect with territorial acquisition, I suspect, is the desire of state elites to signal that the state can't be made to retreat by rebels. "No, we will bear INFINITE costs to beat you!" Of course that's not literally true, they'll leave after some point, but it seems clear to me that e.g. Soviet and American forces both stayed in Afghanistan longer than any economic benefit from Afghanistan *directly* would justify. But you can quite rationally write that off as a loss just to prevent rebels in other provinces from getting the wrong idea. The modern nation-state's sacralization of territory is of course a way of making even more of a credible commitment to this.
This just kicks the can back a step, though, because presumably you want to ask "why do states overacquire territory?" and I presume at least part of the answer is that in any effective imperial state state elites are involved in status competition to Serve The Glory Of The Empire, and acquiring new territory is the juciest way to do that. Military entrepreneurs have an incentive to misrepresent just how valuable the latest acquisition will be, for obvious reasons.
Something relevant to the living standards racket seems to be that modern populations who are used to living standards regularly increasing will see stagnation as an attack on their rights. (It's easy enough to see how this could be rational for the population, especially under situations of distrust.) The Mayan case didn't have to deal with this but it's easy to see how it can become a problem for modern economies.
Re: living standards, another ratchet that seems to be in place is standards for how often one should be working. We've seen some reduction of working hours in some countries, but less than I think would be optimal (a dangerous way to cast judgment, but I'm certainly not the only person with that intuition), especially if you buy some of the concerns raised here. Perhaps reduction of working hours for the average person would be a useful way to reduce the amount of effort spent on the highest-hanging fruit, while managing "real" decline in a way that allows people to feel like significant parts of their lives were still improving (because they would be!) But you'd need a different distribution of power for that, and I suspect power lock-in is another one of these things producing these smooth declining curves (absent favorable shocks.)
Yeah, most of this rings true. Successful states probably have a tendency to be expansionistic and militaristic for a variety of reasons and to reward individuals who excel in those areas. And up to a point, that allows them to outcompete less expansionistic states because they can bring more resources to bear in competition with those states. But it's tough to turn back that expansionistic dial right on schedule. Rome did turn it off eventually, but it waited too long to do it.
Besides its commitments to other territories though, I think you miss one additional reason why unprofitable territories can be tough to give up, and it's that they will often have settlers or colonists from the imperial core in them who expect they will lose their homes if the empire abandons them, and they can easily represent an interest group that is a LOT more motivated to keep the colony in the empire than anyone else is motivated to kick it out. For a good modern example, look at French Algeria. Not a profitable colony by any measure, but it was tough to find a way to get rid of it because a lot of French lived there. When De Gaulle finally pulled the trigger, he was nearly assassinated for his trouble.
Also, depending on the geography of the empire, financially unprofitable territories might act as valuable buffer regions for more profitable areas. Part of the reason why the Roman Mediterranean became so wealthy was that the barbarians were kept hundreds of miles away, so people didn't have to worry about enemy warbands coming along and destroying all their stuff; that would no longer have held true if the Romans had abandoned their German and Balkan provinces, bringing the frontier right next to their productive lands.
For the companies, but not for the management deciding the merge. It makes the power pyramid grow, which is the best way to increase wages (in the broad sense including all benefits) at the top. Each layer can not easily benefit more that a certain factor compared to the layer immediately under. So it's in the top layers advantage to multiply the layers, and for this growing is one of the solution (so is increased complexity). Thinking of it, it may be one strong reason for increased complexity leading to less global efficiency. Growth (size and complexity) is there to increase inequality amount actors and make the top positions more and more rewarding...
Tainter's empirical evidence is consistently the first large, centralized civilization in that region. So it's unsurprising that we should see it struggle to figure out how to manage centralization and complexity: a first draft is rarely the last draft.
On the other hand, Tainter claims that his results are universal enough to apply to both Mayan city-states and modern America. So I would like to see Tainter present some temporal continuity: after Rome has undergone the growth-and-collapse cycle, its successor states should as well. After the Mayan city-states underwent the growth-and-collapse cycle, so should the Aztecs. (Obviously, the conquistadors screwed up that natural experiment.) Something something Chinese dynastic succession (the dynastic change is usually presented in a way that confirms Tainter's hypothesis, which makes it all the more concerning that Tainter instead ignores China). Without some sort of evidence that successor states find the same sorts of complexity-management problems intractable, I remain unconvinced that Tainter can tell us anything about modernity.
"most of the time complexity works. It is a fundamental problem-solving tool. In its early phases, complexity can generate positive feedback and increasing returns. Confronted with challenges, we often respond by strategies such as developing more complex technologies, adding more elements to an institution (specialists, bureaucratic levels, controls, etc.), increasing organization or regulation of transactions, or gathering and processing more information. Each such action represents increasing complexity. Their effectiveness comes in part because changes in these dimensions can be enacted rapidly. While humans may be complexity averse when we personally bear the cost, our problem-solving institutions can be powerful complexity generators. All that is needed for growth of complexity is a problem that requires it. Since problems always arise, complexity seems to grow inexorably."
Complexity is not the right word for the initial process which is really a simplification that transforms the illegible environment into a legible system (legible in the James Scott sense). But this is a quibble.
A useful analogy is that of ‘software entropy’ - “A computer program that is used will be modified. When a program is modified, its complexity will increase, provided that one does not actively work against this.” You can substitute “program” in this definition with almost any “legible” construct (database, process, organisation structure, society etc) and the conclusion still holds. Control starts with simplifying an illegible, complex environment into something legible. This works for a while but the changes required to maintain the efficiency, productivity and stability of the system eventually makes the system complex and illegible again. The natural complexity is replaced in the long run with the complexity of the control process itself.
However, the result of too much complexity is just as likely to be stagnation/sclerosis as it is collapse.
I haven't finished reading yet, but wanted to offer thoughts on standing armies vs professional armies.
Overall, the distinction between a standing army and a professional army is more of a spectrum than a binary choice. A city with a small hinterland, for instance early Rome or a classical Greek city, can field a highly skilled citizen army that is NOT a "standing" army. In fact, having a standing army is silly if most of your battles take place within a reasonable distance of home - it's obviously better to have your soldiers support themselves through farming rather than having to constantly confiscate resources. Farmers will also usually be in good physical condition for fighting, and group training can happen during the long parts of the year during which farmers have lots of free time.
What a standing army gets you, as a state, is not necessarily a superior fighting force, but rather the ability to project that fighting force at a greater distance for longer periods of time. Rome developed a standing army because the size of its empire made it less and less practical to recall its armies home after each campaign. A professional standing army means you can march out on campaign, overwinter in camp, and immediately resume fighting in spring. Later in Roman history though, it became common for large non-professional barbarian armies to smash professional Roman armies. Those large barbarian armies were mostly composed of farmers, with a small core of truly professional fighters at the center. But they were generally able to hold against a Roman legion man-for-man, because being a farmer still leaves you plenty of time to train to fight.
Those are good thoughts -- I've never heard that statement about the barbarian armies. For all my reading on the classical world, one thing I realize I don't have a good sense of is how the lifestyle and social organization of the average inhabitant of, say, present-day Germany differed from 400 BC to 1 AD to 400 AD. We tend to present this "Here Be Dragons" model of European lands in pre-medieval times that were never conquered by Rome.
<i>Later in Roman history though, it became common for large non-professional barbarian armies to smash professional Roman armies.</i>
I don't think it was quite like that. *On the battlefield*, the Roman army tended to outperform its opponents until pretty much the end of the Empire's history. The problem was that, by the late fourth century, Rome had become too corrupt and inefficient to support a big enough army to properly defend itself.
I think a closer look at the historical record suggests that the Roman qualitative edge was really a factor in the explosive expansion of the late republic/early imperial period. After that point, Rome dominated its frontiers through brute numerical superiority.
The Roman army of the fourth century was the largest of any point in Roman history, but it ultimately wasn't enough. As state development proceeded in the "barbarian" lands beyond the frontier, barbarian kings started to achieve local military superiority in their zones of the frontier, and once this happened the professional Roman army was mostly powerless to stop them from crashing through and doing whatever they wanted. The decay of the Roman army came AFTER huge losses were inflicted upon them and key revenue-generating territories were either captured or pillaged.
We don't actually have enough info to say that "The Roman army of the fourth century was the largest of any point in Roman history". There are historians who think that, but equally there are historians who think that the actual number of effective soldiers (as opposed to the paper strength of the army) was lower, or at least not noticeably bigger, than it had been in previous centuries. Personally I lean towards this position: if you read narratives of campaigns in the late fourth and fifth centuries, Romans commanders seem to have generally avoided pitched battles in favour of raiding, blockades, and other comparatively low-risk strategies, which makes more sense if they were short on manpower than if they had waves upon waves of men to throw at their enemies. Plus, there aren't that many records of big Roman defeats at the hands of barbarians: there was Adrianople, and the Vandals managed to win a (naval) battle against the Romans when the latter tried to retake Africa, but other than that, not a lot. I suppose the Huns did manage to win a few times against the Eastern Roman Empire, although they were exceptionally good at fighting (they managed to defeat many of Rome's Germanic enemies as well), so they're not a very representative sample.
> There’s a tax/regulation avoidance <-> compliance enforcement chain that only ever moves the taxation arm of the government in the direction of bloat.
Historically, this is clearly untrue.
My favorite example is a Chinese emperor who announced a simultaneous tax increase and tax reform. Taxes were collected in kind (grain) and delivered to the capital by water transport. Things had gotten into such a state that the boats mostly arrived empty (!), with spoilage and rats having taken most of the taxes.
The reform had two major prongs:
- Collect more taxes in metal, less in kind.
- Get the boats going faster.
And this was so successful that the tax increase was canceled.
But more broadly, a common feature of historical taxation is wide-ranging permanent (generally inheritable) exemptions.
I mean just in my own lifetime I've witnessed at least two times that (income) taxes have been cut significantly in the United States. Feels like we have this ongoing dance where Republicans cut taxes once every decade or so and then Democrats raise them back. General complexity has probably gone up but without digging into the stats directly it fees like taxes themselves kind of oscillate.
Total government receipts (state, local, and other fed taxes) stick to 20% to 30% of GDP since WWII. Income tax is salient because you have to do it all at once and its a source of either a refund or large bill. But other federal taxes + state and local taxes seem to wash it out.
I'm not quite clear on whether you meant to contradict or support my point, looking at the chart, I see this:
2000: 31%
2015: 20%
(It currently sits at 21% as of 2018)
That's a 35% drop, which doesn't feel like a minor perturbation to me, and seems totally in line with what I just said -- a drop in government receipts as a % of GDP by more than a third over 15 years. Although increases/decreases don't seem to line up super closely with Republicans=down, Democrats=up.
Were you trying to say something about the increases/decreases themselves, or just income tax in specific as a driver of that? This chart doesn't seem to support that there haven't been significant fluctuations in the tax rate, but it does support the case that flashy changes to income tax rates mask how much money the government actually takes in as taxes.
It's not saying that taxes only go up, it's saying that the inefficiency of taxes always goes up (i.e. the amount of tax collectors and tax lawyers goes up). Sweeping reforms can negate this, but those are hard and especially hard when the need for them becomes obvious (i.e. when a government is on the ropes because it doesn't get enough tax income, it no longer has the slack to put in major reforms without overstretching and suffering rebellions).
This is true even to some extent for a modern democracy, as while the state has immense power to suppress a violent rebellion, all individual parties are incentivised against reforms with a payback longer than the election cycle.
I read this around 15 years ago and it dominated my thinking about the world for some time. Couldn't get past seeing this increasing complexity and decreasing marginal returns for a while. I now think there are some interesting grains of truth but in general it's all over the place.
One thing that I realized since that I didn't grasp at all when I read it - the ever-increasing Mayan monuments (and Egyptian pyramids for that matter) don't have to be explained by the rules trying to achieve any sort of rational gain. They can also be explained by growing corruption / "regulatory capture" in a society. The first ruler builds a very small monument - this creates a demand for artisans. The artisans might then bribe an advisor to pitch a greater monument to the next ruler. Soon enough there is a whole caste of artisans, their suppliers, their servants, priests etc, and everyone is aligned to redirect more and more resources towards whatever they're doing, and it doesn't matter if it makes any rational sense at all from an abstract "society" perspective. This can cascade until you get to crazy situations where 30000 people are continuously building ever-larger pyramid-shaped tombs instead of doing something useful (Egypt). This also explains a lot of stuff in the modern world, e.g. some aspects of US defense spending, or why NASA stagnated building extremely expensive single-use rockets for 40 years with suppliers distributed across most US states (hint: they weren't actually exploring space, the whole space thing was a byproduct of capturing a resource stream).
I think that it also explains the increase in the number of universities, which increases the number of degree seeking students, which further increases the number of universities ad infinitum. Although this too is not necessarily a rational change, I think that both this and your example are instances of "increase in complexity"
The sad thing is, the number of universities hasn’t increased much in the USA. We just subsidize the demand like crazy without letting supply increase much.
I also wondered why he assumed marginal returns must diminish to the point of being negative rather than stopping at the equilibrium of 0. As you might guess, I tend to read more economics than anthropology.
A thought o Rome relinquishing land. The territorial complexity seems be linked to the social one. The conquered land was used partly to supply veterans with a parcel of their own and I guess land to be exploited by the nobility that funded the war. So some benefits were privatised and now influential people had an interest in Rome keeping the province instead of releasing it.
The Roman Empire did relinquish conquered lands when it thought that it couldn't hold on to it for longer durations. For example, Rome had captured large parts of England. However, it was seemingly difficult to rule, and the locals kept on attacking. Rome subsequently abandoned the whole island and sailed back to the mainland.
While there is some debate about the timing, it occurred on or after 410 AD. The Goths sacked Rome in that year. The collapse of the western empire had already begun by that point.
Possibly a better example is the Roman Empire conquering lands on the other side of the Rhine, and then abandoning them because they would be difficult to rule. I am blanking on the dates, but I think that this was considerably before the decline of the empire
I learned something today! Agri Decumates was the part of the empire north of the Danube and West of the Rhine in modern Germany. It was abandoned, like Dacia Felix (in modern Romania) during the crisis of the third century.
I'd argue that there was a great deal of complexity shedding during that time period (reversion to in kind taxation, decrease in population due to war and plague, much reduced trade and other economic measures due to the debasement of the currency as well as the above) but also some additions of complexity (a much larger government by roman standards).
I've read the original book, and I thought Tanner was at his best when telling various stories about historical societies (and their inevitably collapse). However, IMO he fails to pull all these stories into a coherent model.
I think the problem with his central thesis is that it is not *societies* that drive complexity, but physics. If you want to feed 10 people, you could just roam around picking berries and maybe shooting the occasional antelope. If you want to feed 1M people, you need a complex system of agriculture and trade; you cannot just linearly scale up the hunter-gatherer system.
Similarly, if you want to learn the basic facts of handling fire, you just need a natural fire and some wood chips. If you want to learn about what the surface of the Moon looks like close up, you need a complex system of science and technology that builds increasingly complex instruments. If you want to learn about gravity, you need to build multiple 20-mile-long interferometers.
One possible answer is, "yes, but no one truly *needs* to feed 1M people or learn about gravity", but the hidden price you pay there is fragility. Your merry band of 10 hunter-gatherers may live amazingly happy lives... until they all get wiped out by a flood, a disease, or an angry bear. A nation of 10,000 people may live arguably happy lives, yet get wiped out by a volcanic eruption, a disease, a series of catastrophic floods... though probably not a single bear. A complex technological civilization can pretty much be only wiped out by a planet-killer event (which, admittedly, could be self-inflicted).
There are other hidden costs to an idyllic pastoral existence, of course; for example, hunter-gatherers cannot build or maintain MRI machines; nor can they write books about sociopolitical trends and discuss them on a global telecommunication network. In any case, until the Singularity comes along or the more traditional Rapture whisks us all away or whatever, we have to pay the costs in complexity in order to gain the benefits of power -- the kind of power that even the Romans would find nearly unimaginable.
I don't know how relevant this is to the spirit of the review, but I think the increase in complexity -> splitting phenomenon is perhaps even more true for romantic relationships than countries/civilizations
Interesting thought. Are you talking about any kind of romantic relationships in particular?
To riff on this theme: Getting married definitely increased the complexity in both my and my wife's lives temporarily, for many reasons. We have recognized that this complexity is not sustainable if we want our marriage to last long enough to benefit our kids. It's taken us a few years to create new ways of communicating and working with one another that are now reducing the complexity. OTOH, we're still having kids and that increases complexity in other ways.
Oh nice. I'd be interested in knowing what ways you and your wife and come up with. Was it just attempts at clearer communication or something along those lines?
Well, I guess it's kind of a long story. Every unhappy family is unhappy in its own way and all that.
Our first issue was that we were not aware of our emotional states enough to disentangle them from all the problems that naturally arise for everyone every day. This would lead to silly emotional arguments based on things we were sure the other had said but the other actually didn't say because we were in the Red Zone emotionally. (Bit of dog training jargon there with Red/Yellow/Green zones. And just like dogs, humans are much better at sensing the emotional states of others than they are their own emotional state, particularly when we're already in the Red zone).
The solution to this was for each of us to become aware of our own emotional states enough to calm down our bodies and leave the Red zone. Then we could talk out solutions to other problems.
But blah blah blah, I'm sure that was all very abstract and inapplicable. What are your romantic issues?
I've had bad experiences with what I now understand to be complexity in my previous relationships. I'm just trying to "change myself" and "become better" so that this doesn't affect my current relationship or when I get married. Thanks for sharing your story!
Makes sense. I'm happy to discuss more if you want to post in the next Hidden OT. In the meantime, here are 2 more random thoughts for you:
1. By nature, I value and train being adaptable. The main hard thing for me to realize once I married was: Sometimes you really do need to "change yourself", but other times effectively the other person needs to change in order for a relationship to work. Knowing the difference between these situations, and learning how to (functionally) *cause* my wife to change when it was really needed, were two skills I needed to learn to *stay* married.
2. The main site that helped me with self-change... don't laugh... was chasegirls.com . If you are looking to become better in the same ways I was, that site may resonate. But not the newer overly-monetized articles by the, just the generic PUAs who use too much hair gel. Just the early ones by Chase about spezzatura and all that. I think he wrote an early pdf guide that was helpful for self-work too.
Happy to continue this convo in the next Hidden OT if you might find it helpful.
> Why can states not simply grow to their optimal size and stay there?
I've been thinking about this a bit lately. Banks, brokers, insurance companies, investment funds, F500 companies, your pension, money market funds, and even sovereign debt all have a similar assumption built into the foundation of their existence: that there will be a return on invested capital. This ROIC takes on many forms, but it is incorporated into almost every aspect of every institution in our lives. Without growth, there is little incentive for investment, so it all falls apart. These institutions can't exist without it.
An expectation of growth also allows for leverage in all of the systems above. It varies by industry, but it can hit 20X. And we've seen what happens when systems like these are deleveraged. They don't gracefully fall back to a steady state, they explode.
I don't know if this applies at all to Rome or the Chacoan civilization.
1. "Adjusted for inflation, each dollar invested in energy production in 1960 yielded approximately 2,250,000 BTUs. By 1970 this had declined to 2,168,000 BTUs, while in 1976 the same dollar could produce only 1,845,000 BTUs."
Since the book was published in 1988, and this sort of data is released annually, Tainter was irresponsible to stop his time series in 1976.
2. The decline in per-capita STEM progress looks even worse when you contrast it with the explosion in postsecondary degrees that allegedly enable people to do STEM. Perhaps this is some combination of low hanging fruit exhaustion, institutional rot, and dysgenics.
This has been mostly rising ever since 1981 (after declining in 74-81). Once again it looks like the Tainter is using a trend that ended 7 years before his book was published.
D. The wikipedia article on dependency theory states: "This theory was officially developed in the late 1960s following World War II, as scholars searched for the root issue in the lack of development in Latin America."
As Laplace apocryphally said to Napoleon: "I had no need of that hypothesis."
So that Psychology Today article is referring to a meta-analysis by Ritchie showing gains in test scores from schooling. But the same Ritchie wrote another paper showing that the gains in test scores from schooling are on specific abilities and not on g (just like the Flynn effect). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4445388/
There is a much better patents-per-capita graph here, which goes up to the year 2010: https://www.brookings.edu/wp-content/uploads/2016/06/patenting-prosperity-rothwell.pdf. As you can see, the graph starts to go back up after 1950 and really shoots back up after about 1985 (the book was written in 1988). This updated graph does not support the book's narrative that innovation is going down. Rather, it looks like the book was published during an anomalous period in terms of patents.
Wow. If that's an accurate snapshot of the author's theories regarding ancient Rome, then they're just completely inadequate.
Almost all scholarship these days is agreed that Rome's downfall was multi-faceted and probably intrinsic - i.e. the very nature of Roman society anticipated its ending. Certainly this was the case for the Republic, which fell prey to a variety of issues that were exacerbated in the late 2nd century but had been apparent from its earliest beginnings, such as its unique political system, education of elites, rising populism that invoked earlier rebellion against kingship, etc.
For the empire, its rise due to superficially populistic (but really self-serving) policies made detractors inevitable. Now reaching beyond political arenas, in-fighting and civil conflict had become rampant by the time the western portions fell. Corruption, rebellious legions, barbarian attacks, lack of societal unity, the embracing of mystery cult ideology in the form of Christianity... all of these and more played a role.
The economic factors were certainly there as well, without a doubt. Assigning them such central prominence because they happen to fit a theory, however, requires quite a lot of academic evidence - probably enough to make this two books.
I read this book came to the same questions as you and asked sent a mail to Joseph, I'll reproduce the exchange below:
I wanted to ask you a few questions still lingering after reading the book, would you let me know your thoughts?
1) In a capitalist society, if the state consistently reduces its size when it becomes too big, will the collapse be avoided? I’m trying to understand if your marginal rate of productivity argument applies to private and public entities of a civilisation together or only to the state. Private entities differ from the state as they are allowed to perish regularly (creative destruction)
2) Logically population control should solve the problem. So ideas like one child per family policy which China Implemented should work?
3) What is the way to avoid a collapse? It seemed like you had a very pessimistic / nihilistic view of this phenomenon. I feel like a gradual decrease in complexity till the ideal level should be possible.
Thank you for your time.
Joseph replies :
Thank you very much for your email and your comments on my work. I appreciate both.
Attached please find a couple of older papers that contain the answers to your questions. I am also sending a more recent paper on innovation. I consider this my most important work since the collapse book.
I am lazy and don't want to read those. You have very good questions; what were his answers to them?
Reading this thread, a very common critique of this book is something like "why doesn't society just avoid this by maintaining an optimal population size"? This critique to me seems silly on its face, although it would work well when I play Civilization 4 if that game had such a mechanic. Who is this "society" and why should anyone assume it has free will or can in any way make conscious decisions like an individual human can?
I feel that an easy answer to that question (for SSC readers at least) is that society doesn't stop growing beyond an optimal size because society is Moloch. There is no one person in charge. It is just a conglomeration of competing subunits playing an iterated game of Prisoner's Dilemma. Of course this cannot be the complete answer as stretching the analogy too far leads to monarchy or Fascism
Yes, I have drank the Moloch Kool-Aid as well. Now I'm looking for a way out of that headspace. It's very difficult to unsee Moloch once you've heard of them.
No, this cannot be the complete answer. But can the complete answer come when we add in the concept of faith? Eg, the people's faith that Moloch does not exist and the devil is not behind the curtain?
IOW, maybe the game is reducible to (1) iterated Prisoner's Dilemma, PLUS (2) Keynesian Beauty Contest, ie Common Knowledge Game.
H/t for this idea to epsilontheory.com ; certainly not an endorsement by me of the whole site, as I still don't understand exactly what they're trying to do over there.
Was there a time when you haven't heard of Moloch, when you thought there's a single guiding force behind society instead of lots of competing subunits? If so, why did you think so, and who did you think the guiding force was?
I figured there were lots of competing subunits. It did not occur to me that the force we call Moloch, which does not need to be a "who" but which is both patterned and nebulous, has such seemingly vast explanatory power.
As a humanist, I hope Moloch is just a shared hallucination we have and there are other forces with more power that have more benign interpretations to us.
As far as national debt is concerned, if the economy can grow quickly enough compared to the cost of borrowing, then you can eternally pay off whatever amount of debt is currently due using that growth. It is extremely cheap for the US government to borrow money--the rate of return on bonds is ~2% and in normal times inflation is ~2%, so the cost to borrow in real terms is 0. I do think there's a hidden risk here, namely that if the US government ever looks shaky enough to be forced to pay out more for bonds, we could be in a lot of trouble. Its debts are all denominated in dollars, so if necessary it could inflate its way out of a serious debt problem, although that would probably ruin its credit rating for a long time.
-----
Some notes about the Mayan collapse:
-The statues from later times are full of gibberish. It appears that they forgot how to write, and copied individual symbols from earlier statues, but couldn't actually create sentences.
-The Yucatun is horrible for farming. The soil is thin and rocky, and the rock is full of holes, so in spite of getting lots of rain, you have to store water in order to grow crops.
-Tainter's theory looks very good in some respects: One of the things that hurt the Maya was being forced to farm land higher in the mountains as they ran out of room in the valley, which was even more marginal, and which caused rain to wash chemicals from the ground into the valley, making that land even less useful.
-On the other hand, it looks slightly less good in others; at the time of collapse, the area was going a once-per-7,000-year drought.
For more information, I highly recommend the Fall of Civilizations podcast, which has an episode on the Maya. Actually, thinking about some of the episodes, I would say that Tainter's theory looks quite good in a number of cases, often combined with some natural disaster or exterior pressure that pops the over-inflated balloon.
"As far as national debt is concerned, if the economy can grow quickly enough compared to the cost of borrowing, then you can eternally pay off whatever amount of debt is currently due using that growth."
I've been thinking about this a lot. My understanding is that the major example of such successful growth in the US was in the 1950s. This allowed us to pay back ~all the debt incurred during FDR's administration.
The critique I have heard for our current situation is that our economy is not really growing in a meaningful sense. Yes, we are expanding our GDP, but that's a lot of rentier trickery. Mostly we're just cub scouts selling the same candy bars back and forth to each other that we don't even own.
So, assuming the above paragraph is true, then it is not really economic growth that matters, but people's *faith* that economic growth will continue. This to me is a very scary thought.
I'm not sure where you've seen the critique, so it would depend on the details. That take sounds unusually cynical at best. We're definitely dealing with issues of waste and inefficiency, but I wouldn't say we have literally 0 growth, especially with positive population growth.
Sorry, I agree that as I presented this argument it's definitely too stream of consciousness to make sense of.
I cribbed most of it from Yves Smith's book _Econned_, which you should probably go read if you're interested as she's a lot more eloquent than I. I will try to summarize in another comment later on if I can remember :)
The argument does not require 0 real growth, however you define real growth. It just requires our real ability to pay back debts, however you define that, to grow more slowly than the debts themselves.
It helps to imagine that you are watching from space as an anthropologist. You don't really see "GDP" or a lot of the white-collar work (directly), but you can see people being added, roads being laid down, buildings being built.
I think the problem with these and other similar analysis is the complete lack of consideration of the people, their "virtue" or character in the empire etc. I think anthropologists avoid this because it's not easy to quantify rather than thinking it as unimportant. While I first used to agree with the books thesis, later when I read the book Ceasar: A Sketch I realised Ceasar saved the empire from itself and gave it 400 more years. Rome would have collapsed much earlier if not for him ending the republic and bringing needed reforms into the polity. Tacitus too had a very similar view of Roman History, Rome was failing because since Augustus, the emperor's lost their "virtu" and no one could seem to bring it out of its spiral. The only way it could be saved if again similar to Ceasar the entire ruling class was changed with a new set of highly performing individuals. Rome possibly didn't have such individuals at the time (during Ceasars time it was the army), or they didn't take initiative unlike Ceasar and so Rome Failed. In Ceasar: a sketch I realised the importance on Pompei, a character I never understood well before. You have these pirates that made their own quasi nation and constantly sacked Rome destroying trade and terrorizing the populace for over 5 years. The Roman senate of course seemed completely incapable of handling it any form for 5 years. Then they take Pompei an aristocratic army general and declare him king of the sea and within 3 months he gets rid of the pirates. So we have a failing institution, the senate and an excellently effective institution -- The army and Ceasar basically brings the army to replace senate control of Rome. This circulation of elites or rulers I think explains the collapse of societies better than socio-economic explanations that have a part but are not the driving force.
>Whatever is causing increasing costs, though, Tainter considers the kind of thing that causes societies to collapse.
This makes more sense in reference to the sticky-prices factors considered by Goldstone and Turchin (I think previous reviewed by Scott on SSC): *regardless* of what is causing chronic changes in the price level, expectations about coordination between social subunits are indexed to price expectations, and when expectations are violated you get some mix of {indexed behavior goes haywire}+{people get into escalating conflicts about ever-accelerating adjustments in nominal level of taxes and incomes}.
>Tainter briefly addresses the possibility of a sustainable society that hits the point of zero marginal return on complexity and freezes into place.
Quite obviously none of the social sub-units has any way of knowing whether they have collectively hit the point of zero marginal return! The point of Tainter's argument is that once you overshoot even by a seemingly small amount, the resulting deficit in real resources means that you can no longer sustain the amount of complex infrastructure you had at the peak. (The real question, in fact, is why you can't oscillate back and forth near the point of ZMR, and here Tainter's reference to the flexibility of hunter-gatherer's makes good sense: *if you can agree on who will eat the losses* you can recover from overshooting ZMR without collapse, but if no one will back down and consumption stays at super-ZMR levels then the deficit between infrastructure spending in the last period and what is available for infrastructure spending in this period will just keep growing.)
As to why states do not stop expanding, one possible explanation for this is the Red Queen effect as described in acoup's recent blog post "Collections: Teaching Paradox, Europa Universalis IV, Part II: Red Queens". Past a certain level, continued expansion results in excessive complexity on certain dimensions such as corruption, bureaucratic efficiency, and so forth. But the size-complexity curve scales better for military organization than civilian organization, so the better managed state only succeeds in making themselves more attractive as plunder. At the same time, it also makes the state vulnerable to sudden shocks (natural disasters, revolt, civil war) which result in decay. But this model also suggests that in times of peace, such as the Pax Romana or Pax Americana, small and well governed states should become more common. Is this the case?
Interesting review. I sometimes work professionally with questions related to these, and am embarrassed I had not even heard about Joseph Tainter. A benefit of following ACT!
…That said, reading the review I can understand why he is not a household name in my academic tribe.
The review’er state as Tainter’s central thesis that “the collapse of a society is a response to declining marginal returns on investment in complexity.”
Yeah...right…so we’re dealing with a mono-causal social scientist here…
Ok, to go along plus give a Reader’s Digest review of the review: “degrees of complexity” is the independent variable, and “collapse/non-collapse” is the dependent variable (outcome-variable). In pre-modern times the outcome-variable always turned into a constant sooner or later (it always ends in collapse). While in modern times (after the breakthrough of Modernity) the outcome is always non-collapse, but usually implying variations on stagnation.
hmmm…
To illustrate why pre-modern societies always ends in collapse, Tainter picks the three case-studies Roman, Mayan and Chacoan.
Stop a minute: How did Tainter construct the universe of pre-modern societies from which he drew this sample of three? Any information on this in his book? Probably not…
So what about the premodern Chinese civilization? It just sinified wave after wave of Mongol barbarians who conquered the throne and hummed along nicely, right? Arguably, the Chinese civilisation has survived the bumpy transition to Modernity as well...Today’s rulers seem to have the self-understanding that after a brief 500 year interlude by those talented but annoying round-eyed barbarians from that far-off wind-swept little system of peninsulas and islands at the far end of the Eurasian mainland, the Middle Kingdom (almost with a new King these days) is finally reclaiming its natural place in the world. That’s Popperian falsification right there.
Predictions gone wrong: If I read the review correctly, Tainter believes that break-up into smaller states will become more difficult. Exactly the opposite is happening! After the fall of the Soviet Union (itself an example of state breakup), other states can afford the “luxury” to split up, since there is – at least at the moment – less concern that splitting up will expose you to the risk of being invaded by someone bigger.
And they do. Cf. the old Jugoslav joke from pre-1992 days: “In the future there will only be eight European countries: The EU, Russia and six independent former Jugoslav states”. Except from the fact that EU is the opposite of becoming a country, that’s perceptive.
… Plus, the global demographic transition (to below-reproduction fertility) seems to have been totally under the radar for Tainter. Fertility has been in global decline for decades (even in Sub-Saharan Africa, although the absolute level there is still high). Indicating that humans will be able to handle the population explosion ourselves, without Nature having to do it for us. Sustainability, here we come.
Finally, the graph on fewer patents. I notice the x axis reads “Technical Workers” or “Scientist and Engineers”. If Tainter had instead used “STEM scientists” on the x axis, I would have been more convinced. As it now stands, it is possible that what the graph really tells, is only that Engineers and/or non-science-working technicians have expanded faster than the people (STEM scientists) who really are the ones making new stuff to patent.
I should really start my real work soon, so let me end by following up the last point with a brief foray into the world of literature. In 1884 the playwright Henrik Ibsen published the play The Wild Duck. Its main character, Hjalmar Ekdal, is a failed-everything guy who none the less keeps his head above the water by telling himself “I am a scientist”. (A belief steadily reinstalled in him by the likable old cynic Doctor Relling, a self-declared “Doctor of the Soul”, based on his deep insight: “If you take the life-lie away from an ordinary human being, you take his life-joy away at the same time”.) However, at some point Ekdal hits the wall, and bursts out: “The problem is that most of the things that can be invented, has already been invented”.
…so Tainter’s belief that there are soon no low-hanging fruits left, and we are facing the inevitable decrease in the marginal utility of scientific research, is not exactly new.
It is interesting to compare Tainter's book with Jared Diamond's book, which is similarly titled "Collapse". It's been a while since I read both books, but as I recall, Diamond focuses on environmental degradation, and his argument is essentially Malthusian. Societies grow in population until they exceed the carrying capacity of the land. They can then survive for a while longer by using overly-intensive, non-sustainable extraction, but this degrades the land in the process, by e.g. deforestation, overfishing, desertification, etc. Eventually they can't keep it going, and there's a major population die-off: a collapse.
Diamond's explanation works great for the Mayans, Easter Island, or modern day Haiti, but not so much for Rome, which was depopulated. Tainter's explanation is most interesting in the case of Rome, but does not give enough emphasis to simpler, Malthusian/environmental explanations for other societies. However, I think that marrying these two lines of reasoning provides a better explanation than either. So here's my version:
Societies tend to increase in both population and complexity as long as resources are abundant. (Think the settling of the American west.) The two are linked -- greater population requires greater administrative complexity to manage. As population/complexity increases, diminishing marginal returns come into play. Resources per person are less abundant, and it requires more effort to extract them. Eventually population growth declines and comes to a halt; the society has reached the maximum population/complexity that the land can support with current technology. I am not sure that I buy Tainter's argument that there are forces which necessarily drive complexity beyond this point. I tend to think that need is the mother of invention.
So in my version, simply being at peak size does not necessarily cause collapse. However, a society at its peak is very vulnerable to external shocks. Due to diminishing marginal returns, there is no slack in the system; every unexploited resource has a negative return on investment; it costs more to extract than it provides in benefit. Thus, a foreign invasion, or a drought, or over-extraction, or civil war, or anything else that decreases carrying capacity of the land can trigger a collapse.
Is the modern world vulnerable? According to both Tainter and Diamond, yes. There is ample evidence of decreasing marginal returns, and we are clearly extracting resources in a non-sustainable manner -- fossil fuels, deforestation, fishery collapse, etc. are all major warning signs, and climate change could cause a reduction in capacity. However, IMO, there are also major differences with past civilizations.
First, women's rights and birth control have lead to negative population growth in highly complex, developed economies (Japan, Europe, U.S. w/out immigration.) This has never happened before.
Second, technology now moves faster than population dynamics. Previously, technologies which increase carrying capacity were developed slowly, over multiple generations, so it was easy for population growth to run ahead of them. Now, the opposite is true.
Third, the world is highly interconnected, with superb transport and communication, and yet also highly diversified, which gives global civilization a lot of resilience and adaptability. Individual countries may still collapse -- (USSR, Syria) -- but a global collapse seems less likely, and countries which are still prospering can pick up some of the slack for those that aren't.
Wrt. to the national debt, it is a mistake to think of U.S. national debt, which is issued in a fiat currency that the nation itself controls, as being in any way similar to personal debt. Personal debt must be paid off in the future to avoid bankruptcy, so a person (or corporation) must be careful not to take on more debt than can be paid off in the future by expected future income.
A nation which issues its own fiat currency doesn't have to pay off the debt, ever, unless it wants to. Even if GDP growth ground to a halt, it is always possible to simply print money to "pay off" the debt. Printing money supposedly causes inflation, so the limit on debt is really a limit on how much future inflation you can tolerate, rather than a limit on future earnings. Countries get into trouble only when they have debt in a currency that they don't control, such as U.S. dollar denominated debt in Latin America, or euro-denominated debt in Greece. Greece could have easily dealt with its economic crisis if was able to devalue its currency. Nor is hyperinflation a concern. Hyperinflation is caused by economic collapse, not by debt per-se.
However, the situation is even more complicated than that! In fact, it is actually unclear whether issuing debt is any different from printing money, even at the time it is issued. U.S. government debt is regarded as 100% safe, and is thus held as an asset by banks. Under fractional reserve banking, banks have the power to issue cash (in the form of loans) against their assets. So when the U.S. government issues a Treasury Bond, it is not really "borrowing" money that has to be paid back in the future. It simply creates a bond, that bond is more-or-less equivalent to cash, and it is treated as such by all major institutions. A $100 bill and a $100 treasury note are entirely exchangeable. The only real difference between government bonds and cash, especially with near-zero interest rates, is that bonds have an expiry date -- they get removed from circulation after 30 years, when they are "paid off." So the government has to keep issuing new bonds each year, simply to avoid a contraction of the money supply. (In the late 1990s, when the debt was being paid down, lack of liquidity caused by too few bonds was a real concern that people worried about.)
It is also not clear that issuing too much debt, or printing money, will even cause inflation. Japan is a case study. It's debt is outrageously high, and yet interest rates have remained at rock-bottom, with near zero inflation. In fact, the problem that the Bank of Japan has is that it has tried to engineer higher inflation by printing money/debt, and has been unable to do so. Population decline and stagnant GDP have kept inflation low, even while debt continues to rise.
This is why most economists don't worry too much about the debt.
> Printing money supposedly causes inflation, so the limit on debt is really a limit on how much future inflation you can tolerate,
And that's what Rome did, constantly de-valuing its currency. OP is questioning why they are different, and you are telling him not to worry because they are the same.
> Countries get into trouble only when they have debt in a currency that they don't control
Okay. How is this a useful distinction between Rome and the current day?
I'm not a strict budget hawk, in that I think that the US can maintain deficits indefinitely. But "in own currency I can devalue == no problem" failed for Rome so it's not a good explanation for why the US isn't in any trouble at all.
Also, the US has bondholders, not trade creditors.
A trade creditor can raise prices to reflect debased coinage. A bondholder has already given up the goodies,.so can either accept bumwad or nothing at all.
> “why can’t states just grow to a good size and stop”
Never go full Kelly. (https://thezvi.wordpress.com/2018/10/15/the-kelly-criterion/) "Stop when you reach the point of diminishing returns" is not a strategy that can be implemented, because you don't know where that point is with precision. If you stop when you hit what you think is that point, you have probably gone well past it.
1. To what extent does The Iron Law of Oligarchy explain collapse?
2. Speaking of collapse and its roots, it has been written than a democracy cannot survive as a permanent form of government, because eventually the voters will figure out that they can vote themselves money and other benefits. That may or may not be true, but it's not as if the princes of the Italian Renaissance or Egyptian Pharaohs were known for their frugality and insistence that treasury funds be spent for strictly public purposes, either.
"(1) There’s a tax/regulation avoidance <-> compliance enforcement chain that only ever moves the taxation arm of the government in the direction of bloat."
In the early 1950s, when Wall Street was abandoned for lack of capital and Job Creators were Going Galt on a daily basis, the top US marginal income tax rate was 92%. As in, after a certain point, a individual taxpayer would pay 92% of every dollar earned as income tax.
This was lowered to a much more reasonable 91% in 1954. Generous, even.
Admittedly, the average frustrated US taxpayer didn't pay anything like the top marginal rate. Of course, the tax code of that time also contained a lot fewer loopholes, so there were a lot fewer opportunities for legal tax avoidance for those who were subject to the higher rates.
So it's not as if tax burdens ever always only go up as a function of time.
I actually think that the collapse of the Soviet Union matches Tainters model. What collapsed was the Soviet System not the nation. And the Soviet System was a complete self contained global system, distinct from the Western System (not perfectly distinct but mostly so). Particularly from a financial perspective, there was little to no interaction between the systems and only small amount of material transactions. Glasnot was an attempt to increase the connections between Western and Soviet Systems, probably because the Soviet leadership saw the collapse coming. But the collapse happen, fast and completely and globally, leaving the pieces of the system to go their own ways as best they could.
Regarding gains through trade when trading with developing countries, I think most economic historians would sorta agree with you, but would describe things a bit differently. The terms of trade of most developing countries, the relative price of their exports versus their imports, has been going down for decades, because developing countries mostly export raw materials or goods which see little increased demand / prices as incomes / economies grow. For example, Argentina used to be one of the richest countries on Earth, mostly by exporting beef and other agricultural products to US / Europe. Today they are very poor, and a big reason for that is that these products are simply not all that important now as they were 100-200 years before.
I have a couple points of confusion/concern here (with the book's conclusions as portrayed, the review is lovely):
1) The idea that everything after hunter-gathering forward has made food production less labor-efficient seems implausible on its face to me. A plow made of iron rather than wood, that tills the earth more effectively, thus increasing crop yield for the same plowing process...makes things worse? When I look at crop yields per unit area, they have gone up massively over time, while the percentage of the population doing agricultural labor has gone down enormously. What am I missing that squares this circle?
2) Rome feels like a singularly odd choice for his argument to me, because it lasted a *long* time, in human terms, and survived several significant restructurings. It went from republic to empire, the empire suffered a major civil war, but mostly recovered, and that's before we consider that the Eastern Roman Empire lasted *1000 more years* after the fall of the Western Empire. When Cicero was complaining about yields from territories, western Rome hadn't yet reached its economic peak, and was going to kick around for ~500 more years. Rome's series of largely-successful reorganizations/improvements over time would seem to complicate the "complexity drives inevitable collapse" theory, not reinforce it.
Diminishing marginal returns is a pretty well-established point. However, let's talk about food.
At the moment, returns to food production are obviously diminishing (cow udders can only get so big, etc.) But I see two tech shifts coming that will explosively improve the productivity of investments in food production. One is to move many crops indoors, where inputs (water, fertilizer, pesticide, energy, land, labor) can be dramatically reduced at the cost of increased expertise and machine costs. The other is to move from animal husbandry to synthetic meat/protein, which once scaled up, should have even more dramatic productivity benefits.
Entertaining reading. Overall the whole review can be reformulated as the inefficiency of scale problem. That makes me curious to look into this direction more (probably should be a lot in the business field).
But also very speculative without measure. Literally. It sounds like everyone agrees on the definition of the "collapse" and some generalisation of a "society" what size (or other features) should it have to be admitted into this model? Is it a selection bias? Do societies actually die? Most importantly while at one side it measures investments there is no a single word how to measure the "complexity". the whole construction depends heavily on what do we measure. Is it combinatorial complexity?-) Or the cost to maintain it? And if the latter do they correlate at least?
This isn't serious analysis of the content here, but I was reminded of the premise of Vinge's "A Deepness in the Sky":
"Namqem was a triumph of distributed automation. And every decade it became a little better.
Every decade the flexibility of the governance responded to the pressures to optimize resource allocation, and the margins of safety shrank. The downward spiral was far more subtle than the Dawn Age pessimism of Karl Marx or Han Su, and only vaguely related to the insights of Mancur Olson. The governance did not attempt direct management. Free enterprise and individual planning were much more effective. But if you avoid all the classic traps of corruption and central planning and mad invention, still — 'In the end there will be failures. The governance will have to take a direct hand.' If you avoided all other threats, the complexity of your own successes would eventually get you.
'Okay, I know.' Sammy looked away, and Pham synched his huds to follow what the younger man was seeing: Tarelsk and Marest, the two largest moons. Two billion people on each. They were gleaming disks of city lights as they slid across the face of their mother world — which itself was the largest park in Human Space. When the end finally came to Namqem, it would be a steep, swift collapse. Namqem solar system was not as naturally desolate as the pure asteroidal colonies of the early days of the Space Age . . . but the megalopolis moons required high technology to sustain their billions. Large failures there could easily spread into a system-wide war. It was the sort of debacle that had sterilized more than one of Humankind’s homes."
Planetary civilizations can grow tall and great, but are doomed to collapse from inevitable overoptimization. Interstellar ramscoop traders aren't really a civilization by comparison, but they can avoid the dark age collapses by sleeping through most of their relativistic trade missions and then restocking at golden age planetary civilizations. (By their laws of nature, technological singularities and other clarketech are impossible in known space, so civilizations pick all the possible scientific fruit and peak at that tech level.)
Isn't the cost of offsetting carbon superlinear? IE the price increase the more carbon you need to offset? For instance if you offset by planting more trees, it gets harder and harder to find room (or you need to make sure that the carbon stocked in the trees do not go back in the air when the trees decompose).
In this video, https://youtu.be/G0R09YzyuCI, Tainter says that “collapse is a radical simplification of an overly complex system.” Sums it up rather nicely. In effect, when you are good at solving problems, the consequence is that you have to expand the administrative state to sustain the solution. He also says that his research has not identified a society in history that has voluntarily simplified their system. The question then what then does the average person do to prepare for this seemly inevitable collapse.
Like a few other reviews of the seventeen, this one made me roll my eyes a little at the explain-all-of-history thesis. The treatment is fine but not that deep or engaging, and the tone is more casual than I'd like. Glad to have read it, but I won't be voting for it.
Seems like the key question is whether new technologies can create new "low-hanging fruit" fast enough to counteract the decreasing returns on marginal investment.
Computing and the Internet created an enormous value unlock that lowered the barriers to entry of a bunch of new and meaningful projects (disclaimer: I'm a venture capitalist so I have a certain perspective on this). What's interesting is that the book uses patents as a way to measure scientific progress, and talks about diminishing returns on additional years of education, but at least in some industries this is an outdated way to measure progress and represents an "old" way of thinking about innovation.
Most of the pitch presentations I receive from 50- and 60-somethings lean heavily on patents and the academic pedigree of the team (number of PhDs in management, e.g.). In the software industry at least, these tend to be really bad pitches. Most of the "good" pitches involve 20- and 30- and 40-somethings who stopped their education at the undergrad level, stayed at a couple of jobs long enough to acquire some skills and identify an important problem space, and then got to work building software. Not trying to be ageist but this is the reality for most Internet businesses.
To me this is an example of new technology creating new "low-hanging fruit" and pushing the marginal returns curve out and to the right -- what used to take an army of PhDs is now accessible to, or even better suited for, a team of young hungry people with internet connections and low burn rates. No patents required. Yes, I'm not talking about cancer drugs or moon rockets, but these projects are not without value. I'm talking about the kind of infrastructure that lets you read these words I'm writing.
As for our own civilization, I'm not suggesting that we build a bunch more Airbnbs and Slacks as the world burns around us like Rome, but since technology is the key factor in setting the marginal returns curve (and for forecasting GDP, as you learn in Econ 101), don't all these collapse narratives only hold as long as there's not some new technological unlock to "save" us? Put another way, won't the world look like it's following these collapse trends all the way up until the day we master nuclear fusion, at which point *everything* will be different?
It's public choice theory all the way down.
Interesting review... It does leave something to be desired when it talks about population. Neither the Romans, the Mayans, nor the Chacoans had any form of birth control. Modern society does.
Babies are a positive externality. The benefits are diffuse to the rest of society.
In modern rich societies, babies are underprovided because the costs of them are localized to the families.
So, if you want more young people, you either subsidize babies by making free childcare, child cash bonuses, cheap housing, free education.... or you attract immigrants.
Those are your choices. Immigrants or paying families. There is no third option.
Also, I think the growing to a good place and stopping does have an example. The Nordic countries seem to have nailed their public policy.
Every society has birth control. The "wrong hole" method works great. The pullout method is less reliable, but good enough to reduce the birth rate.
Generally speaking, preindustrial farmers have a high birth rate (because kids can start doing chores early), and preindustrial cities have a low birth rate (because it takes much longer for kids to start pulling their weight). The modern birth rate is low because we're more urban, and because farming has changed.
I am unfamiliar with the "wrong hole" method. Can you elaborate?
If you want to make a baby, you need to put a certain thing into a certain hole. But if you use a different hole, it doesn't work.
I think this only explains a part of why fertility has declined, and the record for subsidizing children isn't that good, particularly over the long term (often such policies can have a temporary effect that fades in a few years).
Also, I can think of other options that might work besides the two you name, they're just not available to Western democracies at present, as they would require either constraining rights or modifying the flow of information in ways that would be perceived as reactionary and outside the Overton Window.
Late empire Roman aristocrats were worried about low birth rates and men not being interested in marrying.
> Those are your choices. Immigrants or paying families. There is no third option.
Isn't a cultural shift a third option? Some cultures have a higher emphasis on children than others (e.g. there's a high correlation between fertility and religion, even correcting for economic factors).
If societies tried to push a cultural shift towards having children, e.g. the same way that they pushed things like environmentalism or anti-smoking campaigns, I think that could have a meaningful impact on birth rates. Sure, maybe "just pay them" might be more effective, but it's not like government subsidy is the only way to shift a culture.
If there's cultural variation within the society, you might expect that shift to take place every generation.
I think it *is* a factor, but not the dominant factor since culture is only partially heritable (e.g. religious couples will have more children, but not all of them will inherit their parents religious views) - and there things like large changes in cultural zeitgeist and immigration that are going to have bigger effects on 'cultural demographics' than relative fertility between the different culture groups.
The retention rate for the Amish seems to be increasing over time. Selection for plainness.
Yeah, and the Amish culture is growing as rapidly as you'd expect given those circumstances: https://en.wikipedia.org/wiki/List_of_U.S._states_by_Amish_population
But there were only ~5000 of them in 1900, and so even with them doubling basically every generation, there's still only about a quarter million.
Yes, but in a century or two, if these trends continue, they will be the single largest group. I once worked these figures out and it's true. Also, unlike other formerly fast-growing groups, they have not slowed down at all.
My impression is that attempts to deliberately change the culture from the top down like that have not had very significant effects in the past.
Paying doesn't work, as a look at the European countries which have all those things demonstrates. One problem is that raising children is expensive in ways which can't be solved with any reasonable amount of money; the parents have an essentially non-delegable responsibility for their children. Which means they must take personal charge, and take it seriously. Which is a task which takes the better part of 20 years (another change from earlier times with earlier ages of independence)
Robin Hanson advocates changing the balance between parents & children to help boost fertility back up:
https://www.overcomingbias.com/2020/10/what-are-parents-owed.html
Countries with better public healthcare, like the nordic countries, DO have a higher birth rate.
Not true. The Scandinavian birth rate is well below even the USA's birth rate. And once you start comparing outside the developed world, forget about it.
There is a third option. Remove women from the (full-time) workforce. Not very politically palatable, but China seems to be trying it currently, if informally so. Obviously not endorsing this.
I was under the impression Japan had both low rates of both fertility & female employment (although the latter had been increasing more recently).
I'd encourage you to check out the general trend. If you exclude poor African countries, there seems to be a clear trend. Why do I feel it's fair to exclude African countries? It's probably safe to say that the sort of work women are doing there is not so fulfilling that it would compel them not to get pregnant. No one is worried about putting their career on hold when it's backbreaking subsistence farming, etc.
Relatedly, employment is not the "problem" per se — it's a fulfilling career that induces many women to substitute away from having children. So you could hypothetically have extremely developed countries like Japan with low(er) female employment rates that have low fertility rates nonetheless because many of the women who are working are doing jobs they enjoy.
https://ourworldindata.org/grapher/fertility-and-female-labor-force-participation
I did read something recently on how female employment works different in Africa. It seems geared to be compatible with having large families. Unfortunately I can't find it now.
This is very speculative, but I've read that a lot of Africa is matriarchal and women do most of the productive work. So, it would make sense for them to have less of a problem keeping fertility high as their female population transitions into a modern careerist workforce. FWIW, I think we will basically converge with them, although we will take a bit of a circuitous route.
In the West, careers were designed for men willing to work very long and hard in order to secure same-class mates with high standards for partner single-earners. This led to a system where the middle class worked very hard, and the upper middle worked even harder (think of the insane hours lawyers and bankers [used to] put in); obviously, such a system is incredibly productive.
This level of dedication is in no way compatible with fertility if adopted by women, which is largely what is happening now. Something will have to give.
My guess is that what will "give" is how worker productivity. Women obviously work very hard. But men can work harder as a group, due to simple evolutionary psychology. If you're a woman, you only have to work as hard as it takes to provide for yourself and your family. If you're a man, you have to work as hard as it takes to attract a good mate and hold onto her. In the West, this level was generally much higher than the "provide for family" level, at least in the productive upper classes. As long as you have an environment where women demand yachts, men will do crazy things like work 18 hour days.
With the entrance of women into the workforce, the first thing you see is male performance go down, as they have less sexual motivation to work extremely hard; sexual motivation is probably most of the motivation men experience, so when women's earning standards go down due to less dependence, you expect men to be less excited about working 16 hour days. Empirically, you are seeing this — men are declining in educational attainment and on-the-job performance.
The second thing you'd expect is a gradual decrease in productivity. It's not that women are lazy, it's just they don't have a good (psychological) reason to work *extremely* hard. They're not excited about working 16 hour days either. So, you see pushes for "work-life balance" and "mental health awareness" and the like. Pretty soon, no one is working at the level that used to be the standard, and in the long-run you'd expect productivity to drastically decrease to roughly whatever is socially deemed "subsistence-level" (house in suburbs yes, yachts no). At this point, work standards will be so low for both genders that society will converge to the African/Soviet model where women work while maintaining fertility.
The only potential problem is that Western societies depend on a level of productivity that is much higher than that seen in Africa and the historic USSR. So, there might be a risk of societal collapse. More likely though, living standards will just plummet to those enjoyed in future Africa / past USSR with everyone blaming their favorite scapegoat for that development.
There is a historical difference between societies where men plow fields vs ones where women can farm using hoes:
https://www.gnxp.com/WordPress/2010/12/02/men-at-work-hoes-ploughs-and-steel/
I thought the rule of thumb for bankers in the (mercilessly regulated) Bad Old Days was "3/6/3"? As in "pay 3% on deposits, lend out at 6% and hit the links no later than 3 PM".
And int he Bad Old Days, law firm partners kept bankers' hours.
An interesting example : The German Democratic Republic managed to have both a much higher birth rate, and a much higher labor force participation of woman than west germany.
Both the Romans and the Greeks practised infanticide for unwanted children. Plus, many newborns did not live very long - so much so that the Romans did not name them until around 9 days after birth.
That said, this has no bearing on modern, civilized societies (which is what I'm assuming the author means by the 'complex' euphemism).
Almost every pre-industrial society had more or less tolerated ways of infanticide.
To mention one especially interesting example: In 18. century france and britain, infanticide was a crime, but mothers could offload their babies to foundling homes, which then would do the killing for them ( these foundling homes had horrible death rates).
"Those are your choices. Immigrants or paying families. There is no third option."
One choice is to re-establish a clan-society. If raising the young is the duty of a whole extended family, people choose to have more kids, as I've heard. Because if you decide to have only two kids and your cousins all have four then you'll spend more of your money and free time on raising the cousins' children than your own.
A weird reverse Tragedy of the Commons solution.
That seems like it has the "if everyone would just..." problem. IE, everyone will not just anything. Increased immigration and increased benefits for parents/families are both things that are conceivable as government policies, whereas even if you somehow managed to elect a government that wanted to reestablish clans I'm not sure how they would do it.
Oh, it's super easy. Just look at the private sector conveniences / laws / government benefits that managed to turn formerly clannish populations into presently non-clannish ones, then just tax / don't enforce / do a worse job providing those.
Heck, it's hard to fail at this really. Even if you fuck up SO badly that you accidentally collapse your own civilization, you will still have succeeded in reestablishing clans.
I guess if you're willing to fully reverse industrialization that might do it but holding power long enough to see that through to its conclusion would be challenging.
Another way to spin this same thought is that in a nuclear family society, the parents have to be able to take care of the kids alone at all but the very highest needs times, while in an extended family society, you can have help whenever your kids are being most demanding, as long as you help your relatives whenever their kids are being most demanding. Having an extra kid might be more manageable if you know that you never have to deal with the worst alone, and you might gladly help someone else deal with their worst moments in exchange.
Romans had plenty of birth control. Besides the ones Bullseye mentioned, there was also infanticide, which was widespread in most ancient societies and less controversial than abortion is today.
And Silphium.
https://en.wikipedia.org/wiki/Silphium
I think you're ignoring the role that culture plays in fertility (Mormons vs Atheists), but I admit that's difficult for the state to influence in non-dystopian ways.
I was just looking at the data and it does seem like subsidising child-rearing actually works, Northern/Western European countries seem to be doing better than Southern/Eastern European countries, and while culture and immigration probably plays a role, the fact that the economy and social welfare programs are doing better in those parts of Europe must be making some contribution.
The premise sounds reasonable (states or whatever get bigger and more complex until they get too unwieldy and expensive to maintain, then they fall apart) but the conclusions sound a bit off.
"The agriculture section is full of graphs that look like this, showing the more work you put in the less yield you get per unit of work"
Yeah, I really want to know more about this, because it sounds on the face of it like "the idyllic hunter-gatherer lifestyle where you just strolled around for a few hours, picked berries, shot a deer, and that's your day's work done, then some idiot invented agriculture where now mud-bedaubed peasants toiled for hours in the muck for a few bowls of gruel", and I don't know how well that holds up.
If there is productivity loss, that may be down to: (1) all the good land is now being farmed, so new farmers have to move onto the marginal land, which is less productive (2) population increase means that instead of 100 kg of rice for 10 families, now you have 100 kg of rice for 20 families which is less return on labour.
Looking up the collapse of the Maya, there seem to be a host of theories as to why: https://en.wikipedia.org/wiki/Classic_Maya_collapse
And would you call, for instance, the collapse of Egyptian civilisation under the onslaught of the Sea Peoples due to complexity?
Tainter's weak answer to "why don't societies just stop at the perfect size?" may be the only one available; there is no "perfect size" or if there is, we don't know it at the time, we can only look back afterwards and say "yeah, when the population was six million in that city, that was the best time to stop there". If you try keeping the population of the city down to six million at the time, well - look at current debates on immigration. And unless the city turns into Detroit so that there is natural shrinkage due to economic collapse, you are not going to be able to reduce surplus population (indeed, you get the opposite problem - too great a loss). *At the time*, there is no easy way (maybe no way at all?) to say to everyone who wants to move to the city "Thanks, but we have calculated the perfect number for our top population range and we've hit it, so you can't come in".
>"Thanks, but we have calculated the perfect number for our top population range and we've hit it, so you can't come in". That's the way the Soviets did it, by the way. You couldn't just move to the city - and industrial plant or some other entity had to "sponsor" your move so that you could get a "propiska" (permit to change residence area). The terms were "limita" for people who were recently hired and immigrated to the city this way. Think the current quotas for foreign labor in most western countries.
Somewhat incredibly, from what I hear, propiska is still around.
It exists but it has become much, much less restrictive. The one thing that it does affect is your ability to draw local benefits - think being able to access some England-specific tax credit if you're originally from Scotland. You need to first establish your residence in the relevant region via some process.
Propiska itself no longer exists, but residential registration with the Ministry of the Interior is mandatory, and needed for access to municipal services.
If you own your own place, then no problem: register yourself. If you rent, then in principle, your landlord is obligated to provide registration for you, but in practice, many landlords are reluctant, because they don’t understand that laws have become simpler, or because they don’t want to get caught not paying taxes on rental income. The easiest solution to this problem is to have a friend who owns his own place virtually register you, which can be done by filling out a form at the post office.
This last option is not however available to foreigners like me, so you have to jump through an additional hoop, such as getting a friend to go with you to the local office of the Ministry of the Interior.
In China the Hukou system is based on the Soviets and is still alive and restrictive to stop overpopulation of tier 1 and 2 cities
I think you have to break production and demand out into different processes in order to see this agricultural phenomenon in present society rather than talk about it in terms of a person's return on agricultural effort
On the production side, industrialized countries certainly aren't seeing a decline in calories-per-farmer-hour, but they are seeing an accellerating burden of complexity to sustain growth in crop outputs.
On the consumption side, there is a population to be fed which was bred up rather quickly to feed the demands of Fordist industry a large fraction of whom have little to contribute to growing and sustaining that complexity.
"And would you call, for instance, the collapse of Egyptian civilisation under the onslaught of the Sea Peoples due to complexity?"
I believe Tainter would suggest that the Sea People are the proximate cause, but that the ultimate cause is declining marginal returns. Eventually they become sclerotic, overextended and unresponsive to threats and opportunities. A similar answer could be applied to the various theories on the Mayan collapse.
In my review I brought up the Mycenean Greek civilization collapsing with the invasion of the Dorian Greeks. It's been years since I read Tainter, but I think he discussed that example but still insisted it followed his rule for collapses.
If you put in more work to extract resources *by the same method*, you get diminishing returns. But modern agriculture makes us fat with only a small percentage of the population producing food; this used to be completely impossible.
This looks to me like a major hole in Tainter's theory, though I haven't read the book myself.
I think that the implicit assumption is that things like the "green revolution" are finite and will one day be exhausted, leading to diminishing returns and unsustainable complexity
I don't think it's a hole - I think these sort of technology improvements shift the graph *up* (by increasing complexity), but it still has the same shape: increased effort leads to declining yields.
You can produce much more food with a fraction of the effort with modern techniques, but eventually they still hit diminishing returns when you have to start cultivating inferior land or whatever.
Which is why it's an overall race between our ability to increase complexity and push the graph up in order to keep up with the needs. I think currently (contra malthus), we've got a really good lead in that race, so I don't think it's a super pressing concern , but the fundamental dynamic sounds correct to me.
To be fair to him, Tainter does mention that science (and therefore engineering) also has diminishing returns, so that each agricultural advance has lower ROI than the next.
This seems untrue. The Green Revolution didn't have lower ROI than previous revolutions in insecticides or pesticides.
I mean no one would make the claim that each is strictly lower. What he probably means is that returns to tech progress are a supermartingale
> Yeah, I really want to know more about this, because it sounds on the face of it like "the idyllic hunter-gatherer lifestyle where you just strolled around for a few hours, picked berries, shot a deer, and that's your day's work done, then some idiot invented agriculture where now mud-bedaubed peasants toiled for hours in the muck for a few bowls of gruel", and I don't know how well that holds up.
The point is *within a given system* more effort leads to diminishing returns.
Like you pick a few berries, that's great and easy. But very quickly the nearby berries are exhausted and you have to traveling farther and farther to find more berries. You need a change in system (an increase in complexity) to get more yields.
> And would you call, for instance, the collapse of Egyptian civilisation under the onslaught of the Sea Peoples due to complexity?
Yeah, he does. I went to refresh my memory on the Bronze Age Collapse, and the Wikipedia page actually cites Tainter: https://en.wikipedia.org/wiki/Late_Bronze_Age_collapse#General_systems_collapse
> In the specific context of the Middle East, a variety of factors – including population growth, soil degradation, drought, cast bronze weapon and iron production technologies – could have combined to push the relative price of weaponry (compared to arable land) to a level unsustainable for traditional warrior aristocracies. In complex societies that were increasingly fragile and less resilient, the combination of factors may have contributed to the collapse.
I also don't quite understand it. Don't we have lots of food now, and very few people working on producing food? https://ourworldindata.org/employment-in-agriculture
The answer to why societies don't just stop at the optimum size is that the optimum size is dynamic, you can be at optimum size with policies to hold at optimum size and then the big drought comes and diminishes your ability to collect energy via photosynthesis and thus the entire basis of your society is undermined. In modern society this can be other limiting factors causing fluctuations in required inputs, fossil fuel energy returns in energy invested, or imports, or supply chain disruptions etc ...
> the collapse of a society is a response to declining marginal returns on investment in complexity.
I once knew a guy who used to go on about this...
Does he write?
This guy does :
https://cassandralegacy.blogspot.com
"The USA’s reliance on immigration to supply young people who can keep the economy running seems a little alarming just because it can’t be kept up as the USA’s relative desirability declines, but it doesn’t have declining marginal returns, exactly."
And
"The US national debt is growing. That seems obviously concerning to me..."
The USA doesn't need young people to keep the economy running, it needs young people to keep the ponzi scheme running. Debt is a word that implies repayment. There is no debt, for the national debt will never - and you can quote me - ever be repaid. Or at least not in the way that you think. Maybe it will be repaid someday, and that will be suddenly and climactically, all at once, with interest on interest accrued and due with all interest and principal and from each citizen, immediately. The declining marginal return (or diminishing return) is absolutely apparent once you realize that the importation of more immigrants to keep the house of cards intact becomes less efficacious when you import more people while simultaneously spending the new imports' futures as well, and at a new logarithmic rate. I wonder if immigrants realize that their futures are also being sold out from underneath them and at a faster rate than the native citizens they're being brought in to vouchsafe pensions and retirements for.
Like the IQ shredder argument. Michael Brendan Dougherty has characterized it similarly, although just in terms of fertility rather than IQ.
This is a frequent topic of discussion at Marginal Revolution in the comments section. My position on this subject as been quite developed and unchanging for some time.
Something that cannot go on forever won't. You don't have to look at the USA or China for signs of a coming demographic calamity, and not everyone on planet earth will be able to made US citizens, most certainly not without someone with the intelligence and money to make a bet against the US economy and dollar noticing such an opportunity. The US is currently paying to import foreign fertility because foreign fertility is currently cheaper, for now.
https://en.wikipedia.org/wiki/One_Billion_Americans
I wish substack had the clap emoji so I could do this properly: One. Billion. Americans.
I was not aware of this book. The premise seems unfeasible and self-centered, but thank you.
👏 is just Unicode.
You can type it and it 👏 just 👏 shows 👏 up.
Bless U+1F647
Does the IQ shredder actually hold up in terms of conditional fertility?
That is, while (fertility|Singaporean resident) < (fertility|origin state resident), for the IQ-shredder argument to hold what you need is (fertility|IQ150 Singaporean nth-generation immigrant) < (fertility|IQ150 origin state resident); has the latter been demonstrated?
(If the latter is not true, then either Singapore's native population just suck and are successfully replacing themselves with fertile, clever immigrants *or* IQ is actually self-shredding.)
"Just suck" was probably not the best choice of words. Suck at reproduction is what I meant.
My understanding is that conditional fertility is indeed lower in big cities (which were also notorious population sinks prior to modern sanitation systems). Fertility also seems to be more affected by years of education rather than IQ.
"Fertility also seems to be more affected by years of education rather than IQ."
This. I've seen this stated elsewhere. The world over most women with an undergraduate degree (or its equivalent) are not going to get into anything serious with a high-school graduate (its equivalent or less), or a woman with a post-grad getting serious with an undergrad.
I don't think it's a simple matter of mate choice. Instead there's a sort of extended adolescence and delay of childbearing during the most fertile years. There's an idea of marriage & childbearing as a "capstone" event once you've got everything else in order. That contrasts with the mothers profiled in "Promises I Can Keep".
https://www.manhattan-institute.org/our-conjugal-class-divide
Yes I think that's part of it. I was also speaking more broadly that statistics seem to show that rising education in women globally seems to portend lower fertility rates. For US, Canadian, and European women I think the 'capstone' idea holds, but there are other places where the cultural phenomenon isn't present, yet higher rates of female education produce the same results.
Hypergamy, for any reason, does not select down.
Yeah, but those are basically irrelevant to the IQ-shredder hypothesis (someone who's capable of emigrating to Singapore is probably in a big city anyway, and is already educated).
It's relevant to the general pattern of 1st world countries with below replacement fertility importing people from countries with higher fertility rates. Often those people will have higher fertility rates after they immigrate than most of the population of their new country... but their children are another story. I haven't discussed the cultural aspect of fertility, but that's a big deal, and unless people adopt some form of cultural separatism (like Anabaptists or ultra-orthodox Jews) they will tend to be absorbed into the larger culture.
"Debt is a word that implies repayment. There is no debt, for the national debt will never - and you can quote me - ever be repaid. "
Debt is being repaid every day. Go to treasurydirect.gov and you'll notice that every bond has a maturity date on which their face value is repaid to the bondholder. The US has never, in its entire history, missed a payment. It is true that the US will always have some outstanding debt, but that's no more concerning than you always having a balance on our credit card.
The balance on the credit card (being a revolver) is A) not a fiscally sound policy and B) is not feasible if the cost of borrowing that money changes (and it will someday) and C) doesn't really = "being repaid" if you keep adding to the balance, and the interest, every single year.
The US has never, in its entire history, missed a payment but it will someday, and that day will be will be, putting it lightly, quite memorable.
Yes, national debt is a problem if it's large, keeps on growing faster than GDP, and the cost of borrowing increases. Granted, those will probably be true for the USA for the forseeable future. But it's a growing debt burden that's the problem, not the fact that the debt will never be repaid. A small debt that's never repaid is no problem whatsoever.
What do you define as a "small debt"?
For the sake of argument, let's say $1 billion. There is no problem whatsoever with the USA having a debt of $1 billion that's never repaid (or rather, continuously renewed).
Such an amount is basically fractions of the change in the country's pocket, euphemistically speaking.
Let me put it another way, do you think that a national debt = to one FY's full GDP (roughly $20 trillion) is a small debt?
Since the Federal Reserve can create US dollars at will, the United States cannot default on its debt, in the sense of "can't pay, won't pay." (If the Fed balks, Congress can amend or terminate the Federal Reserve Act as it sees fit.)
What is more likely is a devaluation, aka a "soft default".
Yes, a devaluation is in a sense what I was referring to. However, there will be nothing soft about the USA's "soft default".
We've raised and lowered tax rates all the time in the US: https://bradfordtaxinstitute.com/Free_Resources/Federal-Income-Tax-Rates.aspx
we've raised and lowered military size and budget several times:
https://www.nytimes.com/interactive/2017/03/22/us/is-americas-military-big-enough.html
"Tax rates go up more often than they go down." is a failed prediction. And one that required him to dismiss the Reagan tax cuts as being insignificant.
If he wrote this today, I wonder if he would say something like: 'It is easier to lower taxes than to raise taxes, so the the budget deficit will continue to grow.' Your theory should not be able to predict both increasing and decreasing taxes.
I think this is a bit glib. The Reagan tax cuts, like the Bush tax cuts and the Trump tax cuts, were fueled by increased debt, which isn't sustainable over the long term. From a 10,000 foot view, it's basically correct to say that the trend in the US in the 20th century was towards a bigger government that spent more and more.
This is an excellent review of a book which I read quite a few years ago, but just recently re-read earlier this year (oddly, a lot of the ACT book reviews are of books I have read recently). I agree completely with the summary, and I am impressed by the thoroughness of the review. I usually type up a brief recap of everything I read, but I am certainly going to add a link to this for future reference.
I too was frustrated by the authors choice of the three in-depth collapse examples. Rome was great, but the others were unfamiliar to me, and both lacked any written records to substantiate or at least "flesh out" the theory.
My take on the matter is that Tainter, like Olson before him, is highlighting a central tendency for organizations to suffer from declining marginal returns, rent seeking, bureaucratic growth and sclerosis. This makes them less adaptive to shocks which are inevitable over the long haul. Eventually they are probably going to collapse.
The only part of the review which lost me was in the final two paragraphs on the negotiating position or power of developed countries. I have no idea what this is referencing.
Why would an increase in complexity in a civilization/country lead to it being less adaptive to shocks? From what I could gather from the review, an increase in complexity is itself a demonstration of superior adaptation skills (governments increasing taxes, which leads to corporations to find clever ways to evade taxes, which leads the government to formulate clever-er ways to collect said tax, ad infinitum). Perhaps being less adaptive is a result of too much complexity, which leads subunits that are in competition to finally "give up" trying to adapt to change?
Also, I don't know if this is super relevant to the topic at hand, but I'd be interested in book recommendations that you might have for the summer :)
There is only so much potentially available surplus. If there is a ratchet effect for some interventions, then over a long enough time horizon you'll run out of headroom and you'll hit a shock that you can't overcome.
Yes, the ratchet effect seems to capture the phenomenon of increasing complexity quite well.
It isn’t complexity that gets them, it is declining marginal returns. This stretches them to the limits of growth (and internal rent seeking) and then a shock such as climate change, internal revolutions, barbarians or whatever hit and they are unable to respond. It’s like a balloon with everyone in society trying to gain as much space as they can until they reach the limits and it all pops.
For suggested books, I think a great place to start is the list of non finalist book reviews that Scott published a month or so ago. Some of my recent favorites are Heinrich's last two books. For an older one that I don’t remember being reviewed here I might suggest Beinhocker's Origins of Wealth.
Perhaps the phenomenon that we're trying to capture can also be expressed with Game Theory, in which two competing subunits try to outmaneuver the other, increasing the complexity of the game, and thereby blinding each other to the actual enemy, whether it be an enemy at the gates or climate change. Of course the competing subunits could cooperate and mutually agree to withdraw, reducing complexity and friction (like the government deciding to freeze tax rates and the corporations agreeing to honor them). However, what ends up happening in reality is closer to an iterated Prisoner's Dilemma, in which a useful working assumption is that the other side is only trying to outsmart you and that you should try your hardest to best them.
Thanks for the suggestions!
Part of the ratchet effect with territorial acquisition, I suspect, is the desire of state elites to signal that the state can't be made to retreat by rebels. "No, we will bear INFINITE costs to beat you!" Of course that's not literally true, they'll leave after some point, but it seems clear to me that e.g. Soviet and American forces both stayed in Afghanistan longer than any economic benefit from Afghanistan *directly* would justify. But you can quite rationally write that off as a loss just to prevent rebels in other provinces from getting the wrong idea. The modern nation-state's sacralization of territory is of course a way of making even more of a credible commitment to this.
This just kicks the can back a step, though, because presumably you want to ask "why do states overacquire territory?" and I presume at least part of the answer is that in any effective imperial state state elites are involved in status competition to Serve The Glory Of The Empire, and acquiring new territory is the juciest way to do that. Military entrepreneurs have an incentive to misrepresent just how valuable the latest acquisition will be, for obvious reasons.
Something relevant to the living standards racket seems to be that modern populations who are used to living standards regularly increasing will see stagnation as an attack on their rights. (It's easy enough to see how this could be rational for the population, especially under situations of distrust.) The Mayan case didn't have to deal with this but it's easy to see how it can become a problem for modern economies.
Re: living standards, another ratchet that seems to be in place is standards for how often one should be working. We've seen some reduction of working hours in some countries, but less than I think would be optimal (a dangerous way to cast judgment, but I'm certainly not the only person with that intuition), especially if you buy some of the concerns raised here. Perhaps reduction of working hours for the average person would be a useful way to reduce the amount of effort spent on the highest-hanging fruit, while managing "real" decline in a way that allows people to feel like significant parts of their lives were still improving (because they would be!) But you'd need a different distribution of power for that, and I suspect power lock-in is another one of these things producing these smooth declining curves (absent favorable shocks.)
Yeah, most of this rings true. Successful states probably have a tendency to be expansionistic and militaristic for a variety of reasons and to reward individuals who excel in those areas. And up to a point, that allows them to outcompete less expansionistic states because they can bring more resources to bear in competition with those states. But it's tough to turn back that expansionistic dial right on schedule. Rome did turn it off eventually, but it waited too long to do it.
Besides its commitments to other territories though, I think you miss one additional reason why unprofitable territories can be tough to give up, and it's that they will often have settlers or colonists from the imperial core in them who expect they will lose their homes if the empire abandons them, and they can easily represent an interest group that is a LOT more motivated to keep the colony in the empire than anyone else is motivated to kick it out. For a good modern example, look at French Algeria. Not a profitable colony by any measure, but it was tough to find a way to get rid of it because a lot of French lived there. When De Gaulle finally pulled the trigger, he was nearly assassinated for his trouble.
Also, depending on the geography of the empire, financially unprofitable territories might act as valuable buffer regions for more profitable areas. Part of the reason why the Roman Mediterranean became so wealthy was that the barbarians were kept hundreds of miles away, so people didn't have to worry about enemy warbands coming along and destroying all their stuff; that would no longer have held true if the Romans had abandoned their German and Balkan provinces, bringing the frontier right next to their productive lands.
On a related note, most corporate mergers appear to be economically inefficient:
https://www.overcomingbias.com/2010/03/hard-facts-mergers.html
For the companies, but not for the management deciding the merge. It makes the power pyramid grow, which is the best way to increase wages (in the broad sense including all benefits) at the top. Each layer can not easily benefit more that a certain factor compared to the layer immediately under. So it's in the top layers advantage to multiply the layers, and for this growing is one of the solution (so is increased complexity). Thinking of it, it may be one strong reason for increased complexity leading to less global efficiency. Growth (size and complexity) is there to increase inequality amount actors and make the top positions more and more rewarding...
Tainter's empirical evidence is consistently the first large, centralized civilization in that region. So it's unsurprising that we should see it struggle to figure out how to manage centralization and complexity: a first draft is rarely the last draft.
On the other hand, Tainter claims that his results are universal enough to apply to both Mayan city-states and modern America. So I would like to see Tainter present some temporal continuity: after Rome has undergone the growth-and-collapse cycle, its successor states should as well. After the Mayan city-states underwent the growth-and-collapse cycle, so should the Aztecs. (Obviously, the conquistadors screwed up that natural experiment.) Something something Chinese dynastic succession (the dynastic change is usually presented in a way that confirms Tainter's hypothesis, which makes it all the more concerning that Tainter instead ignores China). Without some sort of evidence that successor states find the same sorts of complexity-management problems intractable, I remain unconvinced that Tainter can tell us anything about modernity.
For those that want a shorter version of Tainter's thesis, his paper 'Problem Solving: Complexity, History, Sustainability' (http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.372.816&rep=rep1&type=pdf) touches on most of the same points. To quote a relevant paragraph:
"most of the time complexity works. It is a fundamental problem-solving tool. In its early phases, complexity can generate positive feedback and increasing returns. Confronted with challenges, we often respond by strategies such as developing more complex technologies, adding more elements to an institution (specialists, bureaucratic levels, controls, etc.), increasing organization or regulation of transactions, or gathering and processing more information. Each such action represents increasing complexity. Their effectiveness comes in part because changes in these dimensions can be enacted rapidly. While humans may be complexity averse when we personally bear the cost, our problem-solving institutions can be powerful complexity generators. All that is needed for growth of complexity is a problem that requires it. Since problems always arise, complexity seems to grow inexorably."
Complexity is not the right word for the initial process which is really a simplification that transforms the illegible environment into a legible system (legible in the James Scott sense). But this is a quibble.
A useful analogy is that of ‘software entropy’ - “A computer program that is used will be modified. When a program is modified, its complexity will increase, provided that one does not actively work against this.” You can substitute “program” in this definition with almost any “legible” construct (database, process, organisation structure, society etc) and the conclusion still holds. Control starts with simplifying an illegible, complex environment into something legible. This works for a while but the changes required to maintain the efficiency, productivity and stability of the system eventually makes the system complex and illegible again. The natural complexity is replaced in the long run with the complexity of the control process itself.
However, the result of too much complexity is just as likely to be stagnation/sclerosis as it is collapse.
I haven't finished reading yet, but wanted to offer thoughts on standing armies vs professional armies.
Overall, the distinction between a standing army and a professional army is more of a spectrum than a binary choice. A city with a small hinterland, for instance early Rome or a classical Greek city, can field a highly skilled citizen army that is NOT a "standing" army. In fact, having a standing army is silly if most of your battles take place within a reasonable distance of home - it's obviously better to have your soldiers support themselves through farming rather than having to constantly confiscate resources. Farmers will also usually be in good physical condition for fighting, and group training can happen during the long parts of the year during which farmers have lots of free time.
What a standing army gets you, as a state, is not necessarily a superior fighting force, but rather the ability to project that fighting force at a greater distance for longer periods of time. Rome developed a standing army because the size of its empire made it less and less practical to recall its armies home after each campaign. A professional standing army means you can march out on campaign, overwinter in camp, and immediately resume fighting in spring. Later in Roman history though, it became common for large non-professional barbarian armies to smash professional Roman armies. Those large barbarian armies were mostly composed of farmers, with a small core of truly professional fighters at the center. But they were generally able to hold against a Roman legion man-for-man, because being a farmer still leaves you plenty of time to train to fight.
Those are good thoughts -- I've never heard that statement about the barbarian armies. For all my reading on the classical world, one thing I realize I don't have a good sense of is how the lifestyle and social organization of the average inhabitant of, say, present-day Germany differed from 400 BC to 1 AD to 400 AD. We tend to present this "Here Be Dragons" model of European lands in pre-medieval times that were never conquered by Rome.
<i>Later in Roman history though, it became common for large non-professional barbarian armies to smash professional Roman armies.</i>
I don't think it was quite like that. *On the battlefield*, the Roman army tended to outperform its opponents until pretty much the end of the Empire's history. The problem was that, by the late fourth century, Rome had become too corrupt and inefficient to support a big enough army to properly defend itself.
I think a closer look at the historical record suggests that the Roman qualitative edge was really a factor in the explosive expansion of the late republic/early imperial period. After that point, Rome dominated its frontiers through brute numerical superiority.
The Roman army of the fourth century was the largest of any point in Roman history, but it ultimately wasn't enough. As state development proceeded in the "barbarian" lands beyond the frontier, barbarian kings started to achieve local military superiority in their zones of the frontier, and once this happened the professional Roman army was mostly powerless to stop them from crashing through and doing whatever they wanted. The decay of the Roman army came AFTER huge losses were inflicted upon them and key revenue-generating territories were either captured or pillaged.
We don't actually have enough info to say that "The Roman army of the fourth century was the largest of any point in Roman history". There are historians who think that, but equally there are historians who think that the actual number of effective soldiers (as opposed to the paper strength of the army) was lower, or at least not noticeably bigger, than it had been in previous centuries. Personally I lean towards this position: if you read narratives of campaigns in the late fourth and fifth centuries, Romans commanders seem to have generally avoided pitched battles in favour of raiding, blockades, and other comparatively low-risk strategies, which makes more sense if they were short on manpower than if they had waves upon waves of men to throw at their enemies. Plus, there aren't that many records of big Roman defeats at the hands of barbarians: there was Adrianople, and the Vandals managed to win a (naval) battle against the Romans when the latter tried to retake Africa, but other than that, not a lot. I suppose the Huns did manage to win a few times against the Eastern Roman Empire, although they were exceptionally good at fighting (they managed to defeat many of Rome's Germanic enemies as well), so they're not a very representative sample.
Didn't Rome increasingly rely upon (barbarian) mercenaries?
Stilicho is the first name to come to mind.
> There’s a tax/regulation avoidance <-> compliance enforcement chain that only ever moves the taxation arm of the government in the direction of bloat.
Historically, this is clearly untrue.
My favorite example is a Chinese emperor who announced a simultaneous tax increase and tax reform. Taxes were collected in kind (grain) and delivered to the capital by water transport. Things had gotten into such a state that the boats mostly arrived empty (!), with spoilage and rats having taken most of the taxes.
The reform had two major prongs:
- Collect more taxes in metal, less in kind.
- Get the boats going faster.
And this was so successful that the tax increase was canceled.
But more broadly, a common feature of historical taxation is wide-ranging permanent (generally inheritable) exemptions.
I mean just in my own lifetime I've witnessed at least two times that (income) taxes have been cut significantly in the United States. Feels like we have this ongoing dance where Republicans cut taxes once every decade or so and then Democrats raise them back. General complexity has probably gone up but without digging into the stats directly it fees like taxes themselves kind of oscillate.
Total government receipts (state, local, and other fed taxes) stick to 20% to 30% of GDP since WWII. Income tax is salient because you have to do it all at once and its a source of either a refund or large bill. But other federal taxes + state and local taxes seem to wash it out.
https://www.taxpolicycenter.org/statistics/government-current-receipts-and-expenditures
Thanks for the chart!
I'm not quite clear on whether you meant to contradict or support my point, looking at the chart, I see this:
2000: 31%
2015: 20%
(It currently sits at 21% as of 2018)
That's a 35% drop, which doesn't feel like a minor perturbation to me, and seems totally in line with what I just said -- a drop in government receipts as a % of GDP by more than a third over 15 years. Although increases/decreases don't seem to line up super closely with Republicans=down, Democrats=up.
Were you trying to say something about the increases/decreases themselves, or just income tax in specific as a driver of that? This chart doesn't seem to support that there haven't been significant fluctuations in the tax rate, but it does support the case that flashy changes to income tax rates mask how much money the government actually takes in as taxes.
It's not saying that taxes only go up, it's saying that the inefficiency of taxes always goes up (i.e. the amount of tax collectors and tax lawyers goes up). Sweeping reforms can negate this, but those are hard and especially hard when the need for them becomes obvious (i.e. when a government is on the ropes because it doesn't get enough tax income, it no longer has the slack to put in major reforms without overstretching and suffering rebellions).
This is true even to some extent for a modern democracy, as while the state has immense power to suppress a violent rebellion, all individual parties are incentivised against reforms with a payback longer than the election cycle.
I read this around 15 years ago and it dominated my thinking about the world for some time. Couldn't get past seeing this increasing complexity and decreasing marginal returns for a while. I now think there are some interesting grains of truth but in general it's all over the place.
One thing that I realized since that I didn't grasp at all when I read it - the ever-increasing Mayan monuments (and Egyptian pyramids for that matter) don't have to be explained by the rules trying to achieve any sort of rational gain. They can also be explained by growing corruption / "regulatory capture" in a society. The first ruler builds a very small monument - this creates a demand for artisans. The artisans might then bribe an advisor to pitch a greater monument to the next ruler. Soon enough there is a whole caste of artisans, their suppliers, their servants, priests etc, and everyone is aligned to redirect more and more resources towards whatever they're doing, and it doesn't matter if it makes any rational sense at all from an abstract "society" perspective. This can cascade until you get to crazy situations where 30000 people are continuously building ever-larger pyramid-shaped tombs instead of doing something useful (Egypt). This also explains a lot of stuff in the modern world, e.g. some aspects of US defense spending, or why NASA stagnated building extremely expensive single-use rockets for 40 years with suppliers distributed across most US states (hint: they weren't actually exploring space, the whole space thing was a byproduct of capturing a resource stream).
I think that it also explains the increase in the number of universities, which increases the number of degree seeking students, which further increases the number of universities ad infinitum. Although this too is not necessarily a rational change, I think that both this and your example are instances of "increase in complexity"
The sad thing is, the number of universities hasn’t increased much in the USA. We just subsidize the demand like crazy without letting supply increase much.
I have a notion that rules have constituencies. Just knowing the rules can give an advantage, so any effort to simplify rules will meet resistance.
Rules also directly distribute power. We should expect "is powerful" and "is in favor of the current ruleset" to be correlated.
I also wondered why he assumed marginal returns must diminish to the point of being negative rather than stopping at the equilibrium of 0. As you might guess, I tend to read more economics than anthropology.
https://entitledtoanopinion.wordpress.com/2012/12/29/the-collapse-of-complex-societies/
"Standing armies rarely get smaller. "
That seems trivially untrue.
The absolute number of people in the british armed forces has dropped steadily since WW2
As a fraction of population this seems to be the general case:
https://ourworldindata.org/grapher/armed-forces-personnel-percent
The British empire has also shrunk dramatically during that time period, no?
But, as Tainter noted in his book (and I quoted in my review) the staff of their colonial office increased!
A thought o Rome relinquishing land. The territorial complexity seems be linked to the social one. The conquered land was used partly to supply veterans with a parcel of their own and I guess land to be exploited by the nobility that funded the war. So some benefits were privatised and now influential people had an interest in Rome keeping the province instead of releasing it.
The Roman Empire did relinquish conquered lands when it thought that it couldn't hold on to it for longer durations. For example, Rome had captured large parts of England. However, it was seemingly difficult to rule, and the locals kept on attacking. Rome subsequently abandoned the whole island and sailed back to the mainland.
While there is some debate about the timing, it occurred on or after 410 AD. The Goths sacked Rome in that year. The collapse of the western empire had already begun by that point.
Possibly a better example is the Roman Empire conquering lands on the other side of the Rhine, and then abandoning them because they would be difficult to rule. I am blanking on the dates, but I think that this was considerably before the decline of the empire
I learned something today! Agri Decumates was the part of the empire north of the Danube and West of the Rhine in modern Germany. It was abandoned, like Dacia Felix (in modern Romania) during the crisis of the third century.
I'd argue that there was a great deal of complexity shedding during that time period (reversion to in kind taxation, decrease in population due to war and plague, much reduced trade and other economic measures due to the debasement of the currency as well as the above) but also some additions of complexity (a much larger government by roman standards).
East of the Rhine*
Oh that's interesting. Thanks for writing about this!
I've read the original book, and I thought Tanner was at his best when telling various stories about historical societies (and their inevitably collapse). However, IMO he fails to pull all these stories into a coherent model.
I think the problem with his central thesis is that it is not *societies* that drive complexity, but physics. If you want to feed 10 people, you could just roam around picking berries and maybe shooting the occasional antelope. If you want to feed 1M people, you need a complex system of agriculture and trade; you cannot just linearly scale up the hunter-gatherer system.
Similarly, if you want to learn the basic facts of handling fire, you just need a natural fire and some wood chips. If you want to learn about what the surface of the Moon looks like close up, you need a complex system of science and technology that builds increasingly complex instruments. If you want to learn about gravity, you need to build multiple 20-mile-long interferometers.
One possible answer is, "yes, but no one truly *needs* to feed 1M people or learn about gravity", but the hidden price you pay there is fragility. Your merry band of 10 hunter-gatherers may live amazingly happy lives... until they all get wiped out by a flood, a disease, or an angry bear. A nation of 10,000 people may live arguably happy lives, yet get wiped out by a volcanic eruption, a disease, a series of catastrophic floods... though probably not a single bear. A complex technological civilization can pretty much be only wiped out by a planet-killer event (which, admittedly, could be self-inflicted).
There are other hidden costs to an idyllic pastoral existence, of course; for example, hunter-gatherers cannot build or maintain MRI machines; nor can they write books about sociopolitical trends and discuss them on a global telecommunication network. In any case, until the Singularity comes along or the more traditional Rapture whisks us all away or whatever, we have to pay the costs in complexity in order to gain the benefits of power -- the kind of power that even the Romans would find nearly unimaginable.
* I meant "Tainter", not "Tanner", sorry
Tainter's view of societies as a means of harnessing energy actually seems unusually physics-focused.
I don't know how relevant this is to the spirit of the review, but I think the increase in complexity -> splitting phenomenon is perhaps even more true for romantic relationships than countries/civilizations
Interesting thought. Are you talking about any kind of romantic relationships in particular?
To riff on this theme: Getting married definitely increased the complexity in both my and my wife's lives temporarily, for many reasons. We have recognized that this complexity is not sustainable if we want our marriage to last long enough to benefit our kids. It's taken us a few years to create new ways of communicating and working with one another that are now reducing the complexity. OTOH, we're still having kids and that increases complexity in other ways.
Oh nice. I'd be interested in knowing what ways you and your wife and come up with. Was it just attempts at clearer communication or something along those lines?
Well, I guess it's kind of a long story. Every unhappy family is unhappy in its own way and all that.
Our first issue was that we were not aware of our emotional states enough to disentangle them from all the problems that naturally arise for everyone every day. This would lead to silly emotional arguments based on things we were sure the other had said but the other actually didn't say because we were in the Red Zone emotionally. (Bit of dog training jargon there with Red/Yellow/Green zones. And just like dogs, humans are much better at sensing the emotional states of others than they are their own emotional state, particularly when we're already in the Red zone).
The solution to this was for each of us to become aware of our own emotional states enough to calm down our bodies and leave the Red zone. Then we could talk out solutions to other problems.
But blah blah blah, I'm sure that was all very abstract and inapplicable. What are your romantic issues?
I've had bad experiences with what I now understand to be complexity in my previous relationships. I'm just trying to "change myself" and "become better" so that this doesn't affect my current relationship or when I get married. Thanks for sharing your story!
Makes sense. I'm happy to discuss more if you want to post in the next Hidden OT. In the meantime, here are 2 more random thoughts for you:
1. By nature, I value and train being adaptable. The main hard thing for me to realize once I married was: Sometimes you really do need to "change yourself", but other times effectively the other person needs to change in order for a relationship to work. Knowing the difference between these situations, and learning how to (functionally) *cause* my wife to change when it was really needed, were two skills I needed to learn to *stay* married.
2. The main site that helped me with self-change... don't laugh... was chasegirls.com . If you are looking to become better in the same ways I was, that site may resonate. But not the newer overly-monetized articles by the, just the generic PUAs who use too much hair gel. Just the early ones by Chase about spezzatura and all that. I think he wrote an early pdf guide that was helpful for self-work too.
Happy to continue this convo in the next Hidden OT if you might find it helpful.
Thanks! This was really helpful. Sure, I look forward to talking in the next hidden OT
> Why can states not simply grow to their optimal size and stay there?
I've been thinking about this a bit lately. Banks, brokers, insurance companies, investment funds, F500 companies, your pension, money market funds, and even sovereign debt all have a similar assumption built into the foundation of their existence: that there will be a return on invested capital. This ROIC takes on many forms, but it is incorporated into almost every aspect of every institution in our lives. Without growth, there is little incentive for investment, so it all falls apart. These institutions can't exist without it.
An expectation of growth also allows for leverage in all of the systems above. It varies by industry, but it can hit 20X. And we've seen what happens when systems like these are deleveraged. They don't gracefully fall back to a steady state, they explode.
I don't know if this applies at all to Rome or the Chacoan civilization.
1. "Adjusted for inflation, each dollar invested in energy production in 1960 yielded approximately 2,250,000 BTUs. By 1970 this had declined to 2,168,000 BTUs, while in 1976 the same dollar could produce only 1,845,000 BTUs."
This shows the energy price index was flat from 1981-1999: https://fred.stlouisfed.org/series/CPIENGSL
Since the book was published in 1988, and this sort of data is released annually, Tainter was irresponsible to stop his time series in 1976.
2. The decline in per-capita STEM progress looks even worse when you contrast it with the explosion in postsecondary degrees that allegedly enable people to do STEM. Perhaps this is some combination of low hanging fruit exhaustion, institutional rot, and dysgenics.
5. One way to measure productivity changes over time is to look at Real Median Personal Income: https://fred.stlouisfed.org/series/MEPAINUSA672N
This has been mostly rising ever since 1981 (after declining in 74-81). Once again it looks like the Tainter is using a trend that ended 7 years before his book was published.
D. The wikipedia article on dependency theory states: "This theory was officially developed in the late 1960s following World War II, as scholars searched for the root issue in the lack of development in Latin America."
As Laplace apocryphally said to Napoleon: "I had no need of that hypothesis."
Latin American countries have average IQs in the 80s (https://commons.wikimedia.org/wiki/File:World-iq-map-lynn-2006.svg) and lowish economic freedom (https://en.wikipedia.org/wiki/Index_of_Economic_Freedom). IQ, economic freedom, free trade, and oil can explain the vast majority of the international variation in GDP per capita. The graph is so good you don't even need to do statistics on it (https://xkcd.com/2400/)(https://www.nextbigfuture.com/2011/03/per-capita-gdp-prediction-with-97.html)
The wealth->IQ direction of causality seems dubious because nobody can find any specific environmental interventions that raise IQ a lot besides "don't be severely malnourished". Adoptive parent SES has negligible effect. Large racial gaps on test scores (Asians>Whites>Hispanics>Blacks) persist after controlling for parental income (http://theunsilencedscience.blogspot.com/2013/10/black-suits-gowns-skin-sat-scores-by.html). Oil-rich middle eastern countries score lower on pisa than much poorer northeast European countries like Ukraine (https://www.unz.com/isteve/the-new-2018-pisa-school-test-scores-usa-usa/)
I thought it was settled that education increases IQ. See here, for example: https://www.psychologytoday.com/us/blog/brainstorm/201806/how-much-does-education-really-boost-intelligence
So that Psychology Today article is referring to a meta-analysis by Ritchie showing gains in test scores from schooling. But the same Ritchie wrote another paper showing that the gains in test scores from schooling are on specific abilities and not on g (just like the Flynn effect). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4445388/
There is a much better patents-per-capita graph here, which goes up to the year 2010: https://www.brookings.edu/wp-content/uploads/2016/06/patenting-prosperity-rothwell.pdf. As you can see, the graph starts to go back up after 1950 and really shoots back up after about 1985 (the book was written in 1988). This updated graph does not support the book's narrative that innovation is going down. Rather, it looks like the book was published during an anomalous period in terms of patents.
Wow. If that's an accurate snapshot of the author's theories regarding ancient Rome, then they're just completely inadequate.
Almost all scholarship these days is agreed that Rome's downfall was multi-faceted and probably intrinsic - i.e. the very nature of Roman society anticipated its ending. Certainly this was the case for the Republic, which fell prey to a variety of issues that were exacerbated in the late 2nd century but had been apparent from its earliest beginnings, such as its unique political system, education of elites, rising populism that invoked earlier rebellion against kingship, etc.
For the empire, its rise due to superficially populistic (but really self-serving) policies made detractors inevitable. Now reaching beyond political arenas, in-fighting and civil conflict had become rampant by the time the western portions fell. Corruption, rebellious legions, barbarian attacks, lack of societal unity, the embracing of mystery cult ideology in the form of Christianity... all of these and more played a role.
The economic factors were certainly there as well, without a doubt. Assigning them such central prominence because they happen to fit a theory, however, requires quite a lot of academic evidence - probably enough to make this two books.
I read this book came to the same questions as you and asked sent a mail to Joseph, I'll reproduce the exchange below:
I wanted to ask you a few questions still lingering after reading the book, would you let me know your thoughts?
1) In a capitalist society, if the state consistently reduces its size when it becomes too big, will the collapse be avoided? I’m trying to understand if your marginal rate of productivity argument applies to private and public entities of a civilisation together or only to the state. Private entities differ from the state as they are allowed to perish regularly (creative destruction)
2) Logically population control should solve the problem. So ideas like one child per family policy which China Implemented should work?
3) What is the way to avoid a collapse? It seemed like you had a very pessimistic / nihilistic view of this phenomenon. I feel like a gradual decrease in complexity till the ideal level should be possible.
Thank you for your time.
Joseph replies :
Thank you very much for your email and your comments on my work. I appreciate both.
Attached please find a couple of older papers that contain the answers to your questions. I am also sending a more recent paper on innovation. I consider this my most important work since the collapse book.
Attached files:
https://www.sciencedirect.com/science/article/abs/pii/S221042241000002X
https://www.sciencedirect.com/science/article/abs/pii/S1476945X0600002X
https://www.sciencedirect.com/science/article/abs/pii/S0360544215002558
I am lazy and don't want to read those. You have very good questions; what were his answers to them?
Reading this thread, a very common critique of this book is something like "why doesn't society just avoid this by maintaining an optimal population size"? This critique to me seems silly on its face, although it would work well when I play Civilization 4 if that game had such a mechanic. Who is this "society" and why should anyone assume it has free will or can in any way make conscious decisions like an individual human can?
I feel that an easy answer to that question (for SSC readers at least) is that society doesn't stop growing beyond an optimal size because society is Moloch. There is no one person in charge. It is just a conglomeration of competing subunits playing an iterated game of Prisoner's Dilemma. Of course this cannot be the complete answer as stretching the analogy too far leads to monarchy or Fascism
Yes, I have drank the Moloch Kool-Aid as well. Now I'm looking for a way out of that headspace. It's very difficult to unsee Moloch once you've heard of them.
No, this cannot be the complete answer. But can the complete answer come when we add in the concept of faith? Eg, the people's faith that Moloch does not exist and the devil is not behind the curtain?
IOW, maybe the game is reducible to (1) iterated Prisoner's Dilemma, PLUS (2) Keynesian Beauty Contest, ie Common Knowledge Game.
H/t for this idea to epsilontheory.com ; certainly not an endorsement by me of the whole site, as I still don't understand exactly what they're trying to do over there.
Was there a time when you haven't heard of Moloch, when you thought there's a single guiding force behind society instead of lots of competing subunits? If so, why did you think so, and who did you think the guiding force was?
I figured there were lots of competing subunits. It did not occur to me that the force we call Moloch, which does not need to be a "who" but which is both patterned and nebulous, has such seemingly vast explanatory power.
As a humanist, I hope Moloch is just a shared hallucination we have and there are other forces with more power that have more benign interpretations to us.
As far as national debt is concerned, if the economy can grow quickly enough compared to the cost of borrowing, then you can eternally pay off whatever amount of debt is currently due using that growth. It is extremely cheap for the US government to borrow money--the rate of return on bonds is ~2% and in normal times inflation is ~2%, so the cost to borrow in real terms is 0. I do think there's a hidden risk here, namely that if the US government ever looks shaky enough to be forced to pay out more for bonds, we could be in a lot of trouble. Its debts are all denominated in dollars, so if necessary it could inflate its way out of a serious debt problem, although that would probably ruin its credit rating for a long time.
-----
Some notes about the Mayan collapse:
-The statues from later times are full of gibberish. It appears that they forgot how to write, and copied individual symbols from earlier statues, but couldn't actually create sentences.
-The Yucatun is horrible for farming. The soil is thin and rocky, and the rock is full of holes, so in spite of getting lots of rain, you have to store water in order to grow crops.
-Tainter's theory looks very good in some respects: One of the things that hurt the Maya was being forced to farm land higher in the mountains as they ran out of room in the valley, which was even more marginal, and which caused rain to wash chemicals from the ground into the valley, making that land even less useful.
-On the other hand, it looks slightly less good in others; at the time of collapse, the area was going a once-per-7,000-year drought.
For more information, I highly recommend the Fall of Civilizations podcast, which has an episode on the Maya. Actually, thinking about some of the episodes, I would say that Tainter's theory looks quite good in a number of cases, often combined with some natural disaster or exterior pressure that pops the over-inflated balloon.
"As far as national debt is concerned, if the economy can grow quickly enough compared to the cost of borrowing, then you can eternally pay off whatever amount of debt is currently due using that growth."
I've been thinking about this a lot. My understanding is that the major example of such successful growth in the US was in the 1950s. This allowed us to pay back ~all the debt incurred during FDR's administration.
The critique I have heard for our current situation is that our economy is not really growing in a meaningful sense. Yes, we are expanding our GDP, but that's a lot of rentier trickery. Mostly we're just cub scouts selling the same candy bars back and forth to each other that we don't even own.
So, assuming the above paragraph is true, then it is not really economic growth that matters, but people's *faith* that economic growth will continue. This to me is a very scary thought.
Where is the above argument wrong?
I'm not sure where you've seen the critique, so it would depend on the details. That take sounds unusually cynical at best. We're definitely dealing with issues of waste and inefficiency, but I wouldn't say we have literally 0 growth, especially with positive population growth.
Sorry, I agree that as I presented this argument it's definitely too stream of consciousness to make sense of.
I cribbed most of it from Yves Smith's book _Econned_, which you should probably go read if you're interested as she's a lot more eloquent than I. I will try to summarize in another comment later on if I can remember :)
The argument does not require 0 real growth, however you define real growth. It just requires our real ability to pay back debts, however you define that, to grow more slowly than the debts themselves.
It helps to imagine that you are watching from space as an anthropologist. You don't really see "GDP" or a lot of the white-collar work (directly), but you can see people being added, roads being laid down, buildings being built.
The claim that later Mayan statues are full of gibberish looks false. Some good discussion here: https://www.reddit.com/r/AskHistorians/comments/2pzn1s/is_it_true_that_later_mayan_writing_is_gibberish/
Seems comparable to Western people today wearing shirts with nonsensical Asian-language symbols, or the reverse trend in Asian countries.
Reminds me of something I read about Norse runestones. Remote communities had stones with runes that didn't spell anything, or even fake runes.
I think the problem with these and other similar analysis is the complete lack of consideration of the people, their "virtue" or character in the empire etc. I think anthropologists avoid this because it's not easy to quantify rather than thinking it as unimportant. While I first used to agree with the books thesis, later when I read the book Ceasar: A Sketch I realised Ceasar saved the empire from itself and gave it 400 more years. Rome would have collapsed much earlier if not for him ending the republic and bringing needed reforms into the polity. Tacitus too had a very similar view of Roman History, Rome was failing because since Augustus, the emperor's lost their "virtu" and no one could seem to bring it out of its spiral. The only way it could be saved if again similar to Ceasar the entire ruling class was changed with a new set of highly performing individuals. Rome possibly didn't have such individuals at the time (during Ceasars time it was the army), or they didn't take initiative unlike Ceasar and so Rome Failed. In Ceasar: a sketch I realised the importance on Pompei, a character I never understood well before. You have these pirates that made their own quasi nation and constantly sacked Rome destroying trade and terrorizing the populace for over 5 years. The Roman senate of course seemed completely incapable of handling it any form for 5 years. Then they take Pompei an aristocratic army general and declare him king of the sea and within 3 months he gets rid of the pirates. So we have a failing institution, the senate and an excellently effective institution -- The army and Ceasar basically brings the army to replace senate control of Rome. This circulation of elites or rulers I think explains the collapse of societies better than socio-economic explanations that have a part but are not the driving force.
>Whatever is causing increasing costs, though, Tainter considers the kind of thing that causes societies to collapse.
This makes more sense in reference to the sticky-prices factors considered by Goldstone and Turchin (I think previous reviewed by Scott on SSC): *regardless* of what is causing chronic changes in the price level, expectations about coordination between social subunits are indexed to price expectations, and when expectations are violated you get some mix of {indexed behavior goes haywire}+{people get into escalating conflicts about ever-accelerating adjustments in nominal level of taxes and incomes}.
Also,
>Tainter briefly addresses the possibility of a sustainable society that hits the point of zero marginal return on complexity and freezes into place.
Quite obviously none of the social sub-units has any way of knowing whether they have collectively hit the point of zero marginal return! The point of Tainter's argument is that once you overshoot even by a seemingly small amount, the resulting deficit in real resources means that you can no longer sustain the amount of complex infrastructure you had at the peak. (The real question, in fact, is why you can't oscillate back and forth near the point of ZMR, and here Tainter's reference to the flexibility of hunter-gatherer's makes good sense: *if you can agree on who will eat the losses* you can recover from overshooting ZMR without collapse, but if no one will back down and consumption stays at super-ZMR levels then the deficit between infrastructure spending in the last period and what is available for infrastructure spending in this period will just keep growing.)
As to why states do not stop expanding, one possible explanation for this is the Red Queen effect as described in acoup's recent blog post "Collections: Teaching Paradox, Europa Universalis IV, Part II: Red Queens". Past a certain level, continued expansion results in excessive complexity on certain dimensions such as corruption, bureaucratic efficiency, and so forth. But the size-complexity curve scales better for military organization than civilian organization, so the better managed state only succeeds in making themselves more attractive as plunder. At the same time, it also makes the state vulnerable to sudden shocks (natural disasters, revolt, civil war) which result in decay. But this model also suggests that in times of peace, such as the Pax Romana or Pax Americana, small and well governed states should become more common. Is this the case?
Interesting review. I sometimes work professionally with questions related to these, and am embarrassed I had not even heard about Joseph Tainter. A benefit of following ACT!
…That said, reading the review I can understand why he is not a household name in my academic tribe.
The review’er state as Tainter’s central thesis that “the collapse of a society is a response to declining marginal returns on investment in complexity.”
Yeah...right…so we’re dealing with a mono-causal social scientist here…
Ok, to go along plus give a Reader’s Digest review of the review: “degrees of complexity” is the independent variable, and “collapse/non-collapse” is the dependent variable (outcome-variable). In pre-modern times the outcome-variable always turned into a constant sooner or later (it always ends in collapse). While in modern times (after the breakthrough of Modernity) the outcome is always non-collapse, but usually implying variations on stagnation.
hmmm…
To illustrate why pre-modern societies always ends in collapse, Tainter picks the three case-studies Roman, Mayan and Chacoan.
Stop a minute: How did Tainter construct the universe of pre-modern societies from which he drew this sample of three? Any information on this in his book? Probably not…
So what about the premodern Chinese civilization? It just sinified wave after wave of Mongol barbarians who conquered the throne and hummed along nicely, right? Arguably, the Chinese civilisation has survived the bumpy transition to Modernity as well...Today’s rulers seem to have the self-understanding that after a brief 500 year interlude by those talented but annoying round-eyed barbarians from that far-off wind-swept little system of peninsulas and islands at the far end of the Eurasian mainland, the Middle Kingdom (almost with a new King these days) is finally reclaiming its natural place in the world. That’s Popperian falsification right there.
Predictions gone wrong: If I read the review correctly, Tainter believes that break-up into smaller states will become more difficult. Exactly the opposite is happening! After the fall of the Soviet Union (itself an example of state breakup), other states can afford the “luxury” to split up, since there is – at least at the moment – less concern that splitting up will expose you to the risk of being invaded by someone bigger.
And they do. Cf. the old Jugoslav joke from pre-1992 days: “In the future there will only be eight European countries: The EU, Russia and six independent former Jugoslav states”. Except from the fact that EU is the opposite of becoming a country, that’s perceptive.
… Plus, the global demographic transition (to below-reproduction fertility) seems to have been totally under the radar for Tainter. Fertility has been in global decline for decades (even in Sub-Saharan Africa, although the absolute level there is still high). Indicating that humans will be able to handle the population explosion ourselves, without Nature having to do it for us. Sustainability, here we come.
Finally, the graph on fewer patents. I notice the x axis reads “Technical Workers” or “Scientist and Engineers”. If Tainter had instead used “STEM scientists” on the x axis, I would have been more convinced. As it now stands, it is possible that what the graph really tells, is only that Engineers and/or non-science-working technicians have expanded faster than the people (STEM scientists) who really are the ones making new stuff to patent.
I should really start my real work soon, so let me end by following up the last point with a brief foray into the world of literature. In 1884 the playwright Henrik Ibsen published the play The Wild Duck. Its main character, Hjalmar Ekdal, is a failed-everything guy who none the less keeps his head above the water by telling himself “I am a scientist”. (A belief steadily reinstalled in him by the likable old cynic Doctor Relling, a self-declared “Doctor of the Soul”, based on his deep insight: “If you take the life-lie away from an ordinary human being, you take his life-joy away at the same time”.) However, at some point Ekdal hits the wall, and bursts out: “The problem is that most of the things that can be invented, has already been invented”.
…so Tainter’s belief that there are soon no low-hanging fruits left, and we are facing the inevitable decrease in the marginal utility of scientific research, is not exactly new.
It is interesting to compare Tainter's book with Jared Diamond's book, which is similarly titled "Collapse". It's been a while since I read both books, but as I recall, Diamond focuses on environmental degradation, and his argument is essentially Malthusian. Societies grow in population until they exceed the carrying capacity of the land. They can then survive for a while longer by using overly-intensive, non-sustainable extraction, but this degrades the land in the process, by e.g. deforestation, overfishing, desertification, etc. Eventually they can't keep it going, and there's a major population die-off: a collapse.
Diamond's explanation works great for the Mayans, Easter Island, or modern day Haiti, but not so much for Rome, which was depopulated. Tainter's explanation is most interesting in the case of Rome, but does not give enough emphasis to simpler, Malthusian/environmental explanations for other societies. However, I think that marrying these two lines of reasoning provides a better explanation than either. So here's my version:
Societies tend to increase in both population and complexity as long as resources are abundant. (Think the settling of the American west.) The two are linked -- greater population requires greater administrative complexity to manage. As population/complexity increases, diminishing marginal returns come into play. Resources per person are less abundant, and it requires more effort to extract them. Eventually population growth declines and comes to a halt; the society has reached the maximum population/complexity that the land can support with current technology. I am not sure that I buy Tainter's argument that there are forces which necessarily drive complexity beyond this point. I tend to think that need is the mother of invention.
So in my version, simply being at peak size does not necessarily cause collapse. However, a society at its peak is very vulnerable to external shocks. Due to diminishing marginal returns, there is no slack in the system; every unexploited resource has a negative return on investment; it costs more to extract than it provides in benefit. Thus, a foreign invasion, or a drought, or over-extraction, or civil war, or anything else that decreases carrying capacity of the land can trigger a collapse.
Is the modern world vulnerable? According to both Tainter and Diamond, yes. There is ample evidence of decreasing marginal returns, and we are clearly extracting resources in a non-sustainable manner -- fossil fuels, deforestation, fishery collapse, etc. are all major warning signs, and climate change could cause a reduction in capacity. However, IMO, there are also major differences with past civilizations.
First, women's rights and birth control have lead to negative population growth in highly complex, developed economies (Japan, Europe, U.S. w/out immigration.) This has never happened before.
Second, technology now moves faster than population dynamics. Previously, technologies which increase carrying capacity were developed slowly, over multiple generations, so it was easy for population growth to run ahead of them. Now, the opposite is true.
Third, the world is highly interconnected, with superb transport and communication, and yet also highly diversified, which gives global civilization a lot of resilience and adaptability. Individual countries may still collapse -- (USSR, Syria) -- but a global collapse seems less likely, and countries which are still prospering can pick up some of the slack for those that aren't.
Wrt. to the national debt, it is a mistake to think of U.S. national debt, which is issued in a fiat currency that the nation itself controls, as being in any way similar to personal debt. Personal debt must be paid off in the future to avoid bankruptcy, so a person (or corporation) must be careful not to take on more debt than can be paid off in the future by expected future income.
A nation which issues its own fiat currency doesn't have to pay off the debt, ever, unless it wants to. Even if GDP growth ground to a halt, it is always possible to simply print money to "pay off" the debt. Printing money supposedly causes inflation, so the limit on debt is really a limit on how much future inflation you can tolerate, rather than a limit on future earnings. Countries get into trouble only when they have debt in a currency that they don't control, such as U.S. dollar denominated debt in Latin America, or euro-denominated debt in Greece. Greece could have easily dealt with its economic crisis if was able to devalue its currency. Nor is hyperinflation a concern. Hyperinflation is caused by economic collapse, not by debt per-se.
However, the situation is even more complicated than that! In fact, it is actually unclear whether issuing debt is any different from printing money, even at the time it is issued. U.S. government debt is regarded as 100% safe, and is thus held as an asset by banks. Under fractional reserve banking, banks have the power to issue cash (in the form of loans) against their assets. So when the U.S. government issues a Treasury Bond, it is not really "borrowing" money that has to be paid back in the future. It simply creates a bond, that bond is more-or-less equivalent to cash, and it is treated as such by all major institutions. A $100 bill and a $100 treasury note are entirely exchangeable. The only real difference between government bonds and cash, especially with near-zero interest rates, is that bonds have an expiry date -- they get removed from circulation after 30 years, when they are "paid off." So the government has to keep issuing new bonds each year, simply to avoid a contraction of the money supply. (In the late 1990s, when the debt was being paid down, lack of liquidity caused by too few bonds was a real concern that people worried about.)
It is also not clear that issuing too much debt, or printing money, will even cause inflation. Japan is a case study. It's debt is outrageously high, and yet interest rates have remained at rock-bottom, with near zero inflation. In fact, the problem that the Bank of Japan has is that it has tried to engineer higher inflation by printing money/debt, and has been unable to do so. Population decline and stagnant GDP have kept inflation low, even while debt continues to rise.
This is why most economists don't worry too much about the debt.
> Printing money supposedly causes inflation, so the limit on debt is really a limit on how much future inflation you can tolerate,
And that's what Rome did, constantly de-valuing its currency. OP is questioning why they are different, and you are telling him not to worry because they are the same.
> Countries get into trouble only when they have debt in a currency that they don't control
Weimar Republic. Zimbabwe.
Rome didn't issue a fiat currency.
They debased their currency. A gold coin was watered down to be 90% gold and declared (by fiat, one might say) to still be a gold coin.
Fine, but that's not the same as fiat.
Okay. How is this a useful distinction between Rome and the current day?
I'm not a strict budget hawk, in that I think that the US can maintain deficits indefinitely. But "in own currency I can devalue == no problem" failed for Rome so it's not a good explanation for why the US isn't in any trouble at all.
Starters, Rome faces a de facto fixed exchange rate. Say's Law made manifest.
Also, the US has bondholders, not trade creditors.
A trade creditor can raise prices to reflect debased coinage. A bondholder has already given up the goodies,.so can either accept bumwad or nothing at all.
> “why can’t states just grow to a good size and stop”
Never go full Kelly. (https://thezvi.wordpress.com/2018/10/15/the-kelly-criterion/) "Stop when you reach the point of diminishing returns" is not a strategy that can be implemented, because you don't know where that point is with precision. If you stop when you hit what you think is that point, you have probably gone well past it.
1. To what extent does The Iron Law of Oligarchy explain collapse?
2. Speaking of collapse and its roots, it has been written than a democracy cannot survive as a permanent form of government, because eventually the voters will figure out that they can vote themselves money and other benefits. That may or may not be true, but it's not as if the princes of the Italian Renaissance or Egyptian Pharaohs were known for their frugality and insistence that treasury funds be spent for strictly public purposes, either.
"(1) There’s a tax/regulation avoidance <-> compliance enforcement chain that only ever moves the taxation arm of the government in the direction of bloat."
In the early 1950s, when Wall Street was abandoned for lack of capital and Job Creators were Going Galt on a daily basis, the top US marginal income tax rate was 92%. As in, after a certain point, a individual taxpayer would pay 92% of every dollar earned as income tax.
This was lowered to a much more reasonable 91% in 1954. Generous, even.
https://en.wikipedia.org/wiki/History_of_taxation_in_the_United_States#:~:text=For%20tax%20years%201944%20through,tax%20years%201965%20through%201981
Admittedly, the average frustrated US taxpayer didn't pay anything like the top marginal rate. Of course, the tax code of that time also contained a lot fewer loopholes, so there were a lot fewer opportunities for legal tax avoidance for those who were subject to the higher rates.
So it's not as if tax burdens ever always only go up as a function of time.
I actually think that the collapse of the Soviet Union matches Tainters model. What collapsed was the Soviet System not the nation. And the Soviet System was a complete self contained global system, distinct from the Western System (not perfectly distinct but mostly so). Particularly from a financial perspective, there was little to no interaction between the systems and only small amount of material transactions. Glasnot was an attempt to increase the connections between Western and Soviet Systems, probably because the Soviet leadership saw the collapse coming. But the collapse happen, fast and completely and globally, leaving the pieces of the system to go their own ways as best they could.
Regarding gains through trade when trading with developing countries, I think most economic historians would sorta agree with you, but would describe things a bit differently. The terms of trade of most developing countries, the relative price of their exports versus their imports, has been going down for decades, because developing countries mostly export raw materials or goods which see little increased demand / prices as incomes / economies grow. For example, Argentina used to be one of the richest countries on Earth, mostly by exporting beef and other agricultural products to US / Europe. Today they are very poor, and a big reason for that is that these products are simply not all that important now as they were 100-200 years before.
I have a couple points of confusion/concern here (with the book's conclusions as portrayed, the review is lovely):
1) The idea that everything after hunter-gathering forward has made food production less labor-efficient seems implausible on its face to me. A plow made of iron rather than wood, that tills the earth more effectively, thus increasing crop yield for the same plowing process...makes things worse? When I look at crop yields per unit area, they have gone up massively over time, while the percentage of the population doing agricultural labor has gone down enormously. What am I missing that squares this circle?
2) Rome feels like a singularly odd choice for his argument to me, because it lasted a *long* time, in human terms, and survived several significant restructurings. It went from republic to empire, the empire suffered a major civil war, but mostly recovered, and that's before we consider that the Eastern Roman Empire lasted *1000 more years* after the fall of the Western Empire. When Cicero was complaining about yields from territories, western Rome hadn't yet reached its economic peak, and was going to kick around for ~500 more years. Rome's series of largely-successful reorganizations/improvements over time would seem to complicate the "complexity drives inevitable collapse" theory, not reinforce it.
Diminishing marginal returns is a pretty well-established point. However, let's talk about food.
At the moment, returns to food production are obviously diminishing (cow udders can only get so big, etc.) But I see two tech shifts coming that will explosively improve the productivity of investments in food production. One is to move many crops indoors, where inputs (water, fertilizer, pesticide, energy, land, labor) can be dramatically reduced at the cost of increased expertise and machine costs. The other is to move from animal husbandry to synthetic meat/protein, which once scaled up, should have even more dramatic productivity benefits.
Entertaining reading. Overall the whole review can be reformulated as the inefficiency of scale problem. That makes me curious to look into this direction more (probably should be a lot in the business field).
But also very speculative without measure. Literally. It sounds like everyone agrees on the definition of the "collapse" and some generalisation of a "society" what size (or other features) should it have to be admitted into this model? Is it a selection bias? Do societies actually die? Most importantly while at one side it measures investments there is no a single word how to measure the "complexity". the whole construction depends heavily on what do we measure. Is it combinatorial complexity?-) Or the cost to maintain it? And if the latter do they correlate at least?
This isn't serious analysis of the content here, but I was reminded of the premise of Vinge's "A Deepness in the Sky":
"Namqem was a triumph of distributed automation. And every decade it became a little better.
Every decade the flexibility of the governance responded to the pressures to optimize resource allocation, and the margins of safety shrank. The downward spiral was far more subtle than the Dawn Age pessimism of Karl Marx or Han Su, and only vaguely related to the insights of Mancur Olson. The governance did not attempt direct management. Free enterprise and individual planning were much more effective. But if you avoid all the classic traps of corruption and central planning and mad invention, still — 'In the end there will be failures. The governance will have to take a direct hand.' If you avoided all other threats, the complexity of your own successes would eventually get you.
'Okay, I know.' Sammy looked away, and Pham synched his huds to follow what the younger man was seeing: Tarelsk and Marest, the two largest moons. Two billion people on each. They were gleaming disks of city lights as they slid across the face of their mother world — which itself was the largest park in Human Space. When the end finally came to Namqem, it would be a steep, swift collapse. Namqem solar system was not as naturally desolate as the pure asteroidal colonies of the early days of the Space Age . . . but the megalopolis moons required high technology to sustain their billions. Large failures there could easily spread into a system-wide war. It was the sort of debacle that had sterilized more than one of Humankind’s homes."
Planetary civilizations can grow tall and great, but are doomed to collapse from inevitable overoptimization. Interstellar ramscoop traders aren't really a civilization by comparison, but they can avoid the dark age collapses by sleeping through most of their relativistic trade missions and then restocking at golden age planetary civilizations. (By their laws of nature, technological singularities and other clarketech are impossible in known space, so civilizations pick all the possible scientific fruit and peak at that tech level.)
Isn't the cost of offsetting carbon superlinear? IE the price increase the more carbon you need to offset? For instance if you offset by planting more trees, it gets harder and harder to find room (or you need to make sure that the carbon stocked in the trees do not go back in the air when the trees decompose).
In this video, https://youtu.be/G0R09YzyuCI, Tainter says that “collapse is a radical simplification of an overly complex system.” Sums it up rather nicely. In effect, when you are good at solving problems, the consequence is that you have to expand the administrative state to sustain the solution. He also says that his research has not identified a society in history that has voluntarily simplified their system. The question then what then does the average person do to prepare for this seemly inevitable collapse.
Brief review-of-the-review:
Like a few other reviews of the seventeen, this one made me roll my eyes a little at the explain-all-of-history thesis. The treatment is fine but not that deep or engaging, and the tone is more casual than I'd like. Glad to have read it, but I won't be voting for it.
Seems like the key question is whether new technologies can create new "low-hanging fruit" fast enough to counteract the decreasing returns on marginal investment.
Computing and the Internet created an enormous value unlock that lowered the barriers to entry of a bunch of new and meaningful projects (disclaimer: I'm a venture capitalist so I have a certain perspective on this). What's interesting is that the book uses patents as a way to measure scientific progress, and talks about diminishing returns on additional years of education, but at least in some industries this is an outdated way to measure progress and represents an "old" way of thinking about innovation.
Most of the pitch presentations I receive from 50- and 60-somethings lean heavily on patents and the academic pedigree of the team (number of PhDs in management, e.g.). In the software industry at least, these tend to be really bad pitches. Most of the "good" pitches involve 20- and 30- and 40-somethings who stopped their education at the undergrad level, stayed at a couple of jobs long enough to acquire some skills and identify an important problem space, and then got to work building software. Not trying to be ageist but this is the reality for most Internet businesses.
To me this is an example of new technology creating new "low-hanging fruit" and pushing the marginal returns curve out and to the right -- what used to take an army of PhDs is now accessible to, or even better suited for, a team of young hungry people with internet connections and low burn rates. No patents required. Yes, I'm not talking about cancer drugs or moon rockets, but these projects are not without value. I'm talking about the kind of infrastructure that lets you read these words I'm writing.
As for our own civilization, I'm not suggesting that we build a bunch more Airbnbs and Slacks as the world burns around us like Rome, but since technology is the key factor in setting the marginal returns curve (and for forecasting GDP, as you learn in Econ 101), don't all these collapse narratives only hold as long as there's not some new technological unlock to "save" us? Put another way, won't the world look like it's following these collapse trends all the way up until the day we master nuclear fusion, at which point *everything* will be different?