The Strange Disappearance of Cooperation in America II

Peter Turchin


Join 36.9K other subscribers

Apart from sessions on space exploration, the other highlight of Sci Foo for me was, naturally, the discussion of the dynamics of cooperation in America (I have more to say about the search for extraterrestrial intelligence and will return to it in the next blog).

The session was well attended by folks with very different backgrounds. Most (or the ‘plurality’) were social scientists of various kinds, including political scientists and an economist. There were a few ‘techno-geeks,’ this being the Silicon Valley (at least one was a Google employee). And there were other folks such as journalists/media folks and an artist or two.

Such a mixture made for an interesting discussion. I started the discussion by showing a bunch of slides that support my contention that the level of cooperation in America has indeed been declining in the last 3-4 decades. You can see these graphs in the previous blog.

As soon as I showed the first slide (on average membership rate in voluntary organization) the social scientists around the table started questioning whether it was truly showing that cooperation was declining, or could be explained by… and then giving some specific reason why the trend has been declining.

To this I replied that, sure, you can explain this particular trend with that particular explanation, but wait and see until I show you more data. And I went to show the rest of my slides. A couple of other times people would bring up special explanations for one or another declining trend.

Once I showed all the slides, I argued that you can explain any single one away, but it would be a kind of ‘special pleading.’ Numerous indicators of different aspects of cooperation all experienced trend reversals around the 1970s. We can either try to explain each of them separately, by bringing up an ad hoc explanation (different for each trend), or we can admit that something fundamental is going on. I think this argument made an impression, because the discussion next turned to the possible causes of the decline in cooperation and what can be done about reversing it.

There were a lot of interesting points. At one point we were discussing whether it would be a good idea to devolve a lot of functions from the federal to the more local level (states and communities). I thought that was an excellent suggestion because that’s what the theory of cultural multilevel selection suggests. We should allow (indeed, promote) a diversity of approaches by different communities, let them try to make it work, and then other communities (or states) can imitate those solutions that showed they work best. This is one of the ways that cultural group selection can operate without ‘ethnocide’ (that, is only successful practices are borrowed, instead of replacing culture whole-sale).

The idea is, of course, not new. It’s something that the Supreme Court Justice Louis Brandeis used to call “laboratories of democracy.” Unfortunately, this idea was never implemented systematically and we now live in an overly centralized (unnecessarily so) state, in which the freedom of different states to experiment is severely limited by the center.



Another interesting discussion centered on whether technology will enable novel ways of cooperation (well, this should have been expected since we were hosted by Google). In fact, it may already be happening. People brought up two excellent examples: Wikipedia and Linux. I am more familiar with Wikipedia, and I must say that it is a true example of cooperation on a very large scale. It involves huge numbers of contributors donating selflessly their time, energy, expertise, or even money (you can make a donation). And the result is quite spectacular. Wikipedia is by far the most useful web place for looking up facts and data (yes, you have to check it, but most of the time it checks out well; studies show that Wikipedia gets more facts right than traditional encyclopedias).


The future will show whether new technologies will enable large groups of people to cooperate better. Note that most of activities facilitated by the social media are not cooperation in the strong sense, because no public goods are produced as a result of individuals contributing their time, money, or energy (unlike Wikipedia, Facebook and Twitter are for-profit corporations).

I’d like to end this blog on an upbeat note, but there is one area where gloom reigns. We may argue whether Americans are cooperating less or cooperating less in traditional ways, instead shifting their cooperation to new domains opened up by the Internet and computer technology. But at the national level the news are uniformly grim. Our political elites are increasingly polarized and not only between the two parties. The Republicans are now split into two parties: the traditional GOP and the Tea Party. The federal government, and many states, are increasingly dysfunctional, the political gridlock is getting worse, and I see no hope that new technology could somehow save us.

Notify of
Most Voted
Newest Oldest
Inline Feedbacks
View all comments
T. Greer

I recently reviewed the book America 3.0 over at my place. One of the book’s best arguments, IMHO, was on the importance of political decentralization. I summarized their argument as follows:

Key to the program is the goal to “push as many contentious issues as possible to the most basic local level as possible, and then reducing the transaction costs as low as possible (229).” In other words, let each community decide its own policy on social issues but make it as easy as possible for people to switch from one community to another. If state senators in Connecticut want to ban the ownership of assault rifles – let them! If a small town in Utah wants to require every teacher to carry a gun with them to school – let them! If you do not like the policies in your community, move to somewhere new. The end result will be drastic idealogical sorting, as people move to the communities who have the laws and services they want their government to have.

You call it cultural selection, Arnold Kling calls it competitive governance, I’ll just call it a good idea.

Peter Turchin

Agreed in general, but “voting with their feet” is not a terribly efficient mechanism of social selection. If some counties arm everybody, and others disarm all, and in one set homicide rates drop, while in the other go through the roof, then that would be a pretty solid basis for deciding which way to go.


Unfortunately, it will not be a clean experiment. Cities that already have tight gun restrictions (i. e. NYC) don’t have crime rates as low as they could because people constantly bring in guns bought elsewhere. To make group selection work well, you have to somehow restrict such things – not just gun trafficking, but issues like capital flight from unionized states and freeloader migration to places with good social services.


The “voting with one’s feet” argument (in economics, this is more or less the so called Tiebout Model) has some validity to it but it mostly applies only in cases where 1) moving is relatively cost-less (ex. Idaho to Montana) 2) the disparity between choice locations is very large (ex. East and West Berlin), which is really two different ways of saying the same thing, and 3) “some other stuff”.

While this kind of voting may produce “efficient” outcomes it may not produce “desirable” outcomes. I don’t want to throw out a bunch of Social Choice theory from economics here but basically an “efficient” outcome is one where you can’t make anyone better off without making someone worse off. But this is a very minimal criteria. Some of these kinds of outcomes can be strikingly … “unjust” or “unfair” or just plain sucky if you think about it. A standard example is the situation where you got two people and one person has all the money and the other one nothing. This is an “efficient” situation – since taking some money from the rich guy and giving it to the poor one would make one of them worse off – but not a good one. More specific to this context would be something like Thomas Schelling’s old work on the dynamics of racial segregation (which was sort of a “evolutionary economics” before that description caught on) which showed that even a very very very very slight preference for living next to a person of the same race could produce extremely segregated societies (the very very very small preference gets amplified as you move up from the individual to social level – Schelling supposedly used a checkerboard to illustrate the dynamics of the process).

On top of that the “voting with one’s feet” inherits all the potential pitfalls of just ye ol’ regular voting. You can get cycling equilibria a la Condorcet Paradox (some people from A move to B, this makes some people from B want to move to C, this makes people from C want to move A, this makes the people who moved from A to B move back to A, which makes the people who moved from B to C move back, etc). And if you go to higher dimensions (people choosing not just the policy on assault rifles but a mixed basket of policies which include ownership of assault rifles, education funding, provision of parks etc) a lot of the nice results with regard to voting fall apart.

This wouldn’t be that bad except it brings us back to the #1) above. Moving/”voting with one’s feet” isn’t costless. Otherwise – accepting the underlying logic – there’d be a lot more Mexicans (and Kenyans, and Moldovans, and Sri Lankans etc.) pouring to the United States. Yes, there is a lot of immigration already but that reflects the huge disparities in the standards of living between the destinations (in some cases as high as 50x or 60x+ fold) not that costs of migration are small. Borjas (mentioned on this blog previously), whom I don’t agree on a lot of things, at one point estimated that the implicit cost to a Mexican worker moving from Mexico to United States was several hundred thousand dollars, maybe as high as half a million. Think of it this way: you’re an American. You make, the median income of 50 k per year. You have an opportunity to move to … say, Bangladesh and you’re guaranteed a salary of of 100 k (purchasing power adjusted and all that) there for ever. You’ll be be part of the elite. No uncertainty. Would you do it? You don’t speak the language. You won’t relate to anyone until you invest some considerable resources into assimilating into local culture. Your family might not like it. Etc. Some people would. Most people wouldn’t.

Of course international migration provides a very dramatic example but to some extent even within-country migration has hidden substantial costs. I’ve known quite a number of people who have adamantly refused to leave the (US) state of their birth no matter what the economic rewards of relocation. The Idahoans are afraid that they won’t quite relate to the Wyomingians. Broke up a relationship over that kind of thing once (long long time ago).

So yes. The ability of people to “vote with their feet” can help. And it’s good if this kind of voting doesn’t cost much and if there is a wide menu of choices available. But once you consider the costs inherent in this kind of a process and the possibility of undesirable and unstable outcomes it’s pretty clear that this kind of “political decentralization” or “competitive governance” is not a panacea by far – maybe just a slightly positive nudge in the right direction- and by no means a substitute for actual meaningful improvements in how local communities make choices to satisfy the folks who already live in a given place.

Incidentally, I think that concurrent with the decline of cooperation in America that is the subject of the blog post, interstate mobility within US has also been on a decline. I think Paul Krugman had some blog posts about this recently (don’t quote me). I don’t know if the two phenomena are related or what the potential relationship is, but after I saw the first comment on this post that’s what popped into my mind.

T. Greer

I think part of the allure of ‘voting with your feet’ is that it allows the type of cultural selection you are talking about to actually take place.* Gridlock on a smaller level will be reduced because those who are unwilling to cooperate with each other will move away from each other, leaving behind those who will cooperate.

This type of demographic sorting is already happening. This has been portrayed as a bad thing – contributing to the lack of cooperation at the top – but it need not be. As the authors of the work suggest:

“We do not see sorting as necessarily bad. His concern about the impossibility of finding middle ground politically because people all live in a political and cultural echo chamber becomes less important as the role of national government is scaled back massively. Disagreements are tolerable where diversity is tolerated.” (p. 279)

*There is also a normative argument for allowing people to vote with their feet, but it seems less relevant to the content of this discussion.

Tim Tyler

Of course, the “something fundamental” could be cherry-picking by the researcher. How come there are no “trade” figures? Perhaps that’s because trade is up. How come there are no “violence” figures? Perhaps that’s because violence is down. Ngram is fun, but not too useful. “Cooperation” is down, but “help” is up. What does that prove? Not too much – besides that change in word frequencies happens.

John Lillburne

Francis Fukuyama on hi book on trust posits the hospital test for trust – will you help me and come and visit if I am sick in hospital. Cooperation on an intellectual level is a form of trade. I trst the writers to gain by my trade in facts but I will never expect him to visit me in hospital or help if I am being attacked.(trying to)Summaring the work of Paul Zak and Stephen Knack the list of factors that increase trust are-

Trust is a key component in business life and found to be key component in advancing investment and capital deepening
Trust is promoted by
Ethnically homogenous societies
Common religious frame work
Exogamous marriage customs
Income equality
Rule of law (government effectiveness)

The west as a whole has become more diverse on a ethnic,racial,educational income and religious bases, leading to an overall loss in cooperation on a personal basis. Rudyard Kipiling put it thus

The Stranger within my gates,
He may be evil or good,
But I cannot tell what powers control–
What reasons sway his mood;
Nor when the Gods of his far-off land
Shall repossess his blood.

John Lillburne

What other factors contributed to the decline in society wide reduction in trust.
A major factor is technology- Americans spend approximately 4hrs a day in front of a TV and then additional hours tweeting and face booking. Given you sleep 8hours eat for 1-2hrs, work for 8hours, it doesnt leave much time for much else. Technology and property ownership saves time but also consumes time . The power of the TV only really grew in the 1960’s
Another factor is that the leadership doesnt promote a sense of abasiyya. This is probably related to the changing structure of the US leadership. An example may be seen at the apex of society such as the IVY leagur …in the 50’s harvard was a wasp finishing school but by 1995 harvard consisted of 27%jewish 18% white gentile, 20%asian, 8%black, 7% Hispanic and a very large global constituency.
A changed elite needs a changed religion for control and the social message moved continuity and christianity to (sexual) freedom ,diversity and equality. Martriage became passe with resuling decline in social harmony and increased single parent failies. Fatherless housholds have the following social stats-
5 times more likely to commit suicide.
32 times more likely to run away.
20 times more likely to have behavioral disorders.
14 times more likely to commit rape.
9 times more likely to drop out of high school.
10 times more likely to abuse chemical substances.
9 times more likely to end up in a state-operated institution.
20 times more likely to end up in prison.

Peter Turchin

Everybody loves to blame technology for all ills afflicting cooperation. Yet consider two things. 1. Cooperation waxes and wanes in long centuries-long cycles – before the modern technology began its runaway growth. 2. Technology can also enable more effective cooperation. Just witness Wikipedia and Linux that I discuss in the blog. All things taken into the consideration, I am unconvinced that technological change is the root of the problem. Obviously, technological change occurs; the issue is how we deal with it.

T. Greer

Do the two theses have to be in conflict? In Bowling Alone Robert Putnam created a rough pie chart to describe how much each of the various variables discussed has contributed to declining social capital:

If Turchin’s model is generational (and he said in the comments to the last post that it is) then it accounts for at least 60% of the decline, and perhaps more. TV just compounds this broader problem.

(Putnam also talks about what kind of activities TV is disrupting – he calls them schmoozer activities. TV has replaced things like playing card games, having dinner parties, recreational group activities, and other forms of entertainment/hobbies that used to require and/or build social capital. The more civic activities – joining caucuses or associations, voting, and the like – were not as affected by television. This latter group seems to be the type of activities Peter is talking about when he speaks of the ‘disappearance of cooperation”).

John Lillburne

James II of britain once said you cant rule without the church. The church acted as a creator of abasiya and also was part of the agenda creating mechanism of the day. TV has become the agenda setting mechanism of today. both in terms of topics held in peoples consciousness, but also like Mcluhan in that the medium is the message. Niel Postman has written extensively on the cultural effect of the move from a literate culture (the douglas lincoln debates lasted 7 hours a time and had an active audience) to the visual cuture that TV creates. TV is watched in general alpha (meditative ) state which promotes passivity. The internet is to degree a return to a more verbal culture but with an empahsis on oneliners rather than depth of thought.

Niel Postman- amusing ourselves to death

John Lillburne

PS we are approaching the battle of the Boyne where James 2 lost in Ireland..esssentially because of conflict with the protestant church


I do agree with the general gist of this and the previous blog post ( I am a bit tempted to get nit picky about some of the data used to support the conclusions though. In particular… who joins Masonic Lodges anymore? Or the Lions’ Club? That’s like “your grandpa’s way of cooperatin'”! To some extent I think younger people have found new ways of expressing their cooperative spirit hence such measures will simply reflect the fact we’re looking at the wrong data. One thing I’ve observed – and this is anecdotal – is that a lot of socially minded young folks (i.e. during or straight out of college) choose to focus on more individualistic volunteer work (hospitals, community organizations, development work abroad, online stuff etc) rather than these established “crusty” organizations. Being a bit cynical, I’d also add that some of these new venues are nowhere near as productive as the traditional ones, but if we’re speaking to motivation it may not be as bad as it looks.

Still, the strength of the argument is in that it’s not just one but many measures which show a decline so there’s most likely something to it. Is it the supply of cooperative opportunities or the demand for such endeavors that is behind the trend? Or putting it another way, given the existence of the trend, what explains it?


I wonder if the larger number of interests and hobbies that are available to people on and off line has reduced participation in many of these traditional activities. I’m not religious, so I don’t go to Church (plus, religion in general is associated with some pretty unpleasant politics in my mind. The Rotaries allow women now, but that doesn’t interest me in the least. I’m not the type of woman who would ever have been welcome in the Junior League. Heck, I’ve only ever heard of them because I read some books by Anne Rivers Siddons. But I compete in Agility with my dogs, and our local agility club has a couple of groups we’ve started up that raise money for various canine focused charities and which help people with unexpected veterinary expenses.

I think there are other issues why people may not be as involved, even in the activities that interest them. The people who do have work often work longer hours, and with real wages stagnant, households where both parents work are much more common than they once were. So people have less time and energy for things like the PTA and Church committees and so on. Likewise, with the decline of labor unions and the increased outsourcing and automation of many jobs, job security being much lower than it once was. It used to be pretty common for someone to spend years, maybe even an entire career, working at the same plant or company. Today, more and more people go from job to job, often having to pull up roots and move to a new city or town. I even heard a commentator on npr suggesting that a decline in home ownership was good, because it meant US workers would be more flexible and willing to move to wherever the jobs were in our “rapidly changing economy.” If you know you’re going to have to move without much advance notice because you got laid off and have to pull up stakes, you’re probably not going to bother making friends with your co-workers or neighbors, joining the PTA, getting involved in a church or lodge etc.

And there’s the internet. That amazingly distracting thing that allows us to get that feeling of connection without ever stepping outside of our home.


Your Wikipedia and Linux examples actually prove your main point. There are actually very small numbers of people that produce almost all the quality content for both projects. They both have the illusion of being mass sourced projects but in reality are anything but.

  1. Home
  2. /
  3. Cliodynamica
  4. /
  5. Regular Posts
  6. /
  7. The Strange Disappearance of...

© Peter Turchin 2023 All rights reserved

Privacy Policy