Cooperation (vs?) Anarchy

Why ideals fail, why communism is similar to anarchy, and some personal digressions.

This post is a brief digression occasioned by a very recent exchange between Elle Griffin (1, 3, 5, 7) and Peter Clayborne (2, 4, 6, 8) about anarchy. Elle comes from a progress studies and utopian thinking background, with a Substack on “re-thinking the future of nation states, capitalism, and humanity.” In contrast, Peter is a “writer, queer anarchist, ex-fundie… [who is] convinced we have everything we need, in and between our bodies.” And when Elle read Peter’s blog, she commented that she doesn’t understand how anarchy is supposed to work, and a conversation started. I won’t summarize the entire debate, and think you can read this post without doing so - but it was a good discussion, and I’ll mention and quote parts of it.

One key point in the discussion was the tension between the two deeply intertwined issues of theory and practice. Griffin asked how this would work in practice, in various ways, several times. This wasn’t fully answered, and in my conclusion I’ll justify why. That said I will quote Clayborne’s last post, summarizing what it would look like.

“What that looks like in practice will vary by context, but the common denominator is the direction of power. We don’t control anyone; neither do we let anyone control us. It’s an active resistance to the accumulation of power — which requires power of a sort. Immense effort must be expended in the opposite direction than it is now: towards nurturing collaborative systems and cooperative culture, and away from social and material systems that encourage people to hoard and dominate.”

This utopian vision doesn’t seems achievable, even if it seems possible. But before talking about why anarchy is, I want to talk about a diametric opposite. I will claim that Communism cannot possibly work and does not, in order to contrast the less obviously impossible opposite.

Communism

Communism is a very different approach to utopia, one that is largely seen to have failed. And the reasons Communism failed were manifold. I’ll start from the one most sympatico with anarchist and libertarian thinking: the calculation problem. This problem was introduced by von Mises, and expounded in further detail by Hayek, but boils down to the fact that a centralized communist planning system inevitably does not have the information needed to allocate resources, and would not have the ability to solve the combinatorial explosion problem of allocating resources even if it did. As von Mises said, “[Communism¹] is not a realizable system of society's economic organization because it lacks any method of economic calculation.”

Bryan Caplan, noted economist and libertarian, was somewhat skeptical; “current events [as of the 1990s, i.e. the fall of the USSR and its sequelae,] do nothing to show that economic calculation was the insuperable difficulty of socialist economies.” He then points out that the calculation problem is a scaling problem, just like many other problems which we’ve pointed out with cooperative regimes in general. Of course, Caplan wasn’t embracing communism, he was simply pointing out that other problems, “out of hundreds on the list of arguments,” may have been critical.

He points out “the problem of work effort, or innovation, or the underground economy” as possible alternative explanations - and I will group these together as the incentive problem. That is, Communism fails because it’s not compatible with human incentives; Marx’s credo of “from each according to his ability” fails to reward effort, because it concludes “to each according to his needs.”

I am unsure who first noticed the fact that communism is an ideal system for ants, rather than humans, but T.H. White’s “Once and Future King” explicitly calls it out. Throughout his work there is a clear theme of grappling with the broader questions of fascism, power, and government². In the course of Arthur Merlin’s attempt to educate Arthur about government and rulership, he was turned into an ant. During that story, the book states that in the “fabulous past…when ants were still like men,” they had “not yet settled down to communism.” That is, society in which communism is compatible with humanity requires that humanity loses that which makes them different from ants - individuality, independence, and so on.

Kibbutzim; Communism on a Small Scale

Communism fails in practice, but in very limited ways works as an ideal for sharing and cooperation. It even works at small scales, for a limited time, as long as external pressures and exposure to the alternatives do not overwhelm it. The Kibbutz movement is a clear case study - and an excuse for a personal digression.

The early modern Jewish settlement of Israel is obviously controversial, but even before the notion of Zionism had emerged, Eastern European Jews fleeing pogroms were joined by other immigrants who circumvented the immigration bans put in place by the Ottoman empire, and those moving within the Ottoman Empire³. My great-grandfather Marco was among the latter group, and moved with his mother as a child as one of the wave of Jews who moved from elsewhere in the Ottoman Empire. This was before the establishment of the first Kibbutzim; he lived in Jerusalem until his adoption and immigration to the United States.

I currently live in Israel, along with my parents and one of my two siblings. Marco, mentioned above, had a granddaughter, my mother, who came right after finishing graduate school. Following my father’s lead, they lived on a Kibbutz in the 1970s, at the peak of the movement, living the Zionist dream, and soon got married. My father wanted to stay - but the oft-told family story of a kibbutz member who inherited a piano led my mother to refuse.

The piano was a family heirloom, but the Kibbutz, per Communist doctrine, did not believe in private property. At a Kibbutz meeting, over the objections of the otherwise owner, it was decided that it would be placed in a communal kids playroom. My mother, sympathetic to the ideals of Kibbutzim, was American enough to be horrified. So instead of staying, they ended up returning to the United States for several decades to raise their family before coming back.

And this is a microcosm of one key part of the lack of appeal of Kibbutzim⁴. That is, the appeal of collectivism is in the theory, not the practice. So even as some parts of collectivism were embraced by the west - strong social safety nets, worker protections, and taxes on income to reduce the degree of inequality - the communist program as a whole failed. And unsurprisingly, the children who grew up there generally didn’t stay.

But to return to the putative topic of the post, the reasons communism fail are not the same as the issues with anarchy - in fact, they move in exactly the opposite direction. And despite this, Anarchy also works well at small scales, but is as-yet unworkable otherwise. And we’ll need to explain this later in the post, once we’ve laid out a bit more of why each could succeed.

Anarchy is, well, anarchy

The debate which occasioned this post continually comes back to the question of how anarchy works in practice. Elle Griffon, in the penultimate post of the series, makes the point that “We were [doing the things anarchy aims for, but a] lot of the anarchist argument seems to be disgruntlement for the way those systems turned out.” That is, humanity evolved from smaller scale and more diverse approaches, and moved from there to the current world order. Therefore, as I’ve seen argued before, if anarchy leads to the modern system, why would we bother recapitulating that process?

I don’t think that is entirely fair - the evolution from tribes to kings to modern democracy didn’t ever stop at the station of anarchic self-governance. The primal state imagined by Rosseau, with some type of a noble ungoverned human existence was an illusion - though to be fair, so was Hobbes’ imagined state of nature as uncontrolled anarchy. Peter Clayborne argues, “we aim to reincorporate ideas and practices older than those of our adversaries.” But, again, those practices mostly never existed! The distant past wasn’t every-human-for-themself, it evolved from the family and tribal structures we still see in primates, with control enforced by a leader and at most semi-cooperative incentives, to other centralized and imposed power structures with the noted failure modes.

In the first post in the series, Clayborne noted that “what makes it so easy for tyrants to take over today is that our world is already structured to centralize power, with a culture already positioned to accept authority.” That is true - but as far as I am aware, the only times full anarchy has existed is when civil order breaks down - and that is inevitably accompanied by loss of trust and an inability to lead to what Griffon summarized as “communities rising up and self-organizing.” Because, as earlier posts have explored, the scaling up of cooperation requires some structure.

Of course, this does not refute the existence of successful smaller scale anarchies, any more than the calculation problem refutes the generations-long success of Kibbutzim from the 1930s to the 1970s. Clayborne continued - “ultimately what we want is an end to the imposition of ideological blueprints entirely. We want the time and space to make decisions for ourselves together.” But as we explained, we need structures for this to be possible.

That said, I don’t think what Clayborne imagines is necessarily incompatible with humanity. It’s not communism, only temperamentally suitable for insects. And I agree with Clayborne that “would-be reformers end up the best defenders of the status quo, because of their insistence on what they see as realistic or pragmatic. Anarchists try to live with the hope that our dreams are not too far-fetched.”

What could enable positive anarchy?

I do not know if the type of future that embodies the best of what Griffon and Clayborne want is possible for humans. It may turn out that it is not. But I think the previous post, where I talked about how the international system manages to cooperate without global governance, and without requiring uniformity, holds the key to what would enable anarchy at scale, if it is possible at all.

The three things I pointed out which promote the success of international relations in the anarchic realm of great-power interactions are interpersonal relationships between leaders, strong norms, and incentives.

At Scale, Relationships are Useless (for Anarchy)

The first of the above triad, unfortunately, does not work for anarchy at large scales. We cannot simply say that everyone should care about everyone else, and cooperate for that reason - if we could, it would work for communism as well!

Clayborne seems to dispute this. He suggests that “If any large region went anarchist, it would be the collective decision of almost everyone involved, and would result in deeply connected networks of support and mutual self-defense.” Unfortunately, the number of relationships grows exponentially with membership. A deeply connected network requires lots of connections - and in an anarchical system, the lack of enforced leadership means that everyone needs to cooperate directly. But relationships (at a minimum) require time to build and maintain, so humans can only sustain a finite number of relationships. (I talked about this in the past in terms of why startups turn into corporations, and why polyamory was harder than monogamy.)

The issue with relying on relationships at scale was somewhat obliquely raised in the dialog, when Griffon raised a concern that anarchy cannot scale past Dunbar’s number. But the limitation fundamentally has nothing to do with Dunbar’s number, which Clayborne (rightly) disputes - it’s a simple mathematical fact that once the group is too large, the strength of the vast majority of the possible relationships is inevitably zero. The consequence is that the deeply connected networks Clayborne wants are impossible.

Game Theory

Skipping to the last of the triad of factors which enable at-scale cooperation in geopolitics, incentives point to reasons that anarchy is possible - but it depends on the structure of the game being played. This has been a theme before, but bears repeating in this context. For example, if the game being played has a single robust equilibrium that is beneficial for all players, cooperation is easy.

The below diagram, from this paper shows that the entire bottom-left section of the map - the payoffs of which are shown below - where this is more or less strongly the case. The shaded parts of the payoff matrix show where the Nash equilibr(ium/ia) are located; in all but the top left of this subset of games, the Nash equilibria is not just good, but strongly robust.

As long as the world is playing these robustly cooperative games, and everyone knows it, there is no reason not to cooperate, and anarchy can reign, in a positive sense⁵.

But there are some games where exploitation is possible - where forcing the other person to do something that hurts them is good for you. Prisoner’s Dilemma and Chicken are like this. And in negative-sum games, exploitation can not only transfer wealth, but destroy it.

One example could be child labor, at least in developed economies. Children attending school make them better able to contribute economically later in life, and working instead of attending school not only benefits the manager at a cost to the child, but destroys future wealth far in excess of what is gained. There are arguably many other examples - leveraged buyouts of productive firms, predatory lending, or pollution and overfishing.

What stands out among the possible examples is corruption of almost any sort - leveraging a position of power to extract private benefits. Non-cooperation has costs, and these costs are far higher when there is control. If pursued fully, anarchy removes that control and makes some of the worst failures impossible - which is quite a strong recommendation. The cost of non-cooperation at the highest levels is horrendous - the sheer scale of the cost of war is enough to make a dreamer “imagine there's no countries… nothing to kill or die for.” And Lennon was right when he wrote that he’s not the only one, but changing the world requires changing norms.

Norms

After covering the other two legs of the triad of reasons that at-scale cooperation works in geopolitics, we arrive at norms. And in context of the above discussion, the reason that norms are valuable is that they can push non-cooperative games into being cooperative, and push individuals to cooperate. What does that mean?

In a game like the prisoner’s dilemma, the direct payoffs for each of the hypothetical arrestees make it beneficial to defect against the other, and cooperate with the police. But this discounts some really important questions about how people behave - and how the norms in different groups differ. Specifically, there is a saying popular in come parts of the population, “snitches get stiches” - and in some places, if you defect you might get a lighter jail sentence, but if the people you end up serving time with find out, you might not survive. In other places, cooperating with the police is normal, and there’s no honor among thieves - outside of the prison sentence you’ll receive, there’s no outside motive not to defect.

Despite the fact that the specific example is somewhat inverted - where “cooperate” means not to cooperate with police, which is a broader question of cooperation - I think the example shows that changing norms can change games to be more or less compatible with being cooperative. And Clayborne discusses part of this when he says “I would like behaviors that are currently celebrated as success (like hoarding and exploitation) shamed on a societal level.” The set of actions which are “shamed on a societal level” differ between societies, but destructive exploitation is a primary example of where we want to ensure societal costs are imposed. And if that happens, we could transform the game - impose a cost on the actions which are exploitative or controlling, and on failing to cooperate. But for this to promote anarchy, this imposed cost would not be dictated from above - it would need to be socially adopted.

To return to the question of whether norms can change the game, it seems plausible that they can, at least in the vision that Clayborne proposes. And I agree - norms can reshape the world, and more than once they have. Christian enlightenment norms largely created the western world, and as I’ve mentioned, and as Pinker has argued at length, norms promoting international peace have done so far more quickly and recently.

But if norms are so powerful, perhaps we could make a similar case for communism; if there were norms for people to work hard and appreciate societal gains, and we imposed costs on shirking duty and on wanting more than others, humans could be happy with communism. It’s no coincidence that “Imagine” was so aligned with Communism. But I think this goes back to TH White’s extended moralizing, where he argued that adopting the norms required for communism to be effective turned humans into the archetypal ant, destroying their humanity. And with communism in practice, we saw that occur to a large extent - the level of control needed to make communism successful was incompatible with human flourishing. (Not to mention the fact that it inevitably led to centralizing power in the hands of dictators⁶.)

Less cynically than White, we could say that communism is a system for angels, not for humans. And temporarily neglecting scale, if humans are virtuous, and work hard because they believe in the system, the Kibbutz flourished. This isn’t a dream - it worked for decades. This was not a fluke, it was a real outcome, showing the viability of Communism at small scales. But, to be clear, it eventually failed even at the smaller scale of communities. And the real reason for failure was not just human nature, it was exposure to the outside world, and the hard facts of what creates flourishing.

This was seen when Yeltsin went grocery shopping. The West hadn’t defeated Russia for most of the space race, even if it took a victory lap at the end - but when the Unites States tried to show off its space program, it was the incidental side trip that showed that Communism couldn’t create the material benefits which capitalism did. That is, Communism is outcompeted as an economic system, and when a choice exists, it is abandoned.

An observant person might have predicted this from the gradual failure of Kibbutzim more than a decade earlier. The children saw what the rest of the world was like. Unlike my mother, they weren’t raised to think that inheritance was normal, nor, I presume, were they shocked by things like the loss of an heirloom. Instead, even after having been shaped by the Kibbutz system, and inculcated into the norms that allowed it, they wanted the material comforts and success of the outside world more.

Despite the fact that Kibbutzimg gradually adapted to capitalism and hedged their communal nature in the 70s and 80s, the children generally left - because they had a choice. And the USSR collapsed and countries adopted democracy as soon as the same choice presented itself, and China adopted a mix of heavily regulated and centrally coordinated state capitalism, which has allowed it to (partially) recover from its slow modernization, and then its disastrous adoption of Communism⁷.

“Human Nature” and Cooperation for Anarchy

I think the norms required for anarchy are (in at least some respects,) less directly opposed to human nature than those of communism. And the problem of anarchy is related to the problem of cooperation more generally; if people can move towards more intrinsically motivated cooperation, anarchy becomes viable.

But we haven’t seen anarchy adopted anywhere at scale - and the question is why. The answer, I think, is that we’re dealing with humans, and their individual goals are not always compatible with societal good. In an earlier post, I talked about why status was always zero-sum, and opposed to enabling cooperation. But interpersonal, political, and other types of power are narrower and even better examples of anti-cooperative goals. And these goals are natural, to some extent. It seems very clear that some people derive great personal pleasure by having power over others. Of course, it’s entirely possible that most people really want to be sheep, where others have power over them - but if that’s true, humanity isn’t particularly compatible with anarchy anyways⁸.

On the other hand, lots of things we don’t do are “natural” - as any parent who pays attention should notice, as they slowly socialize their children into the norms of modern humanity. For example, property ownership isn’t natural or an objective fact, it’s a (critically useful) imposed intersubjective reality. Natural human urges include violence and theft, as well as a whole host of disease-spreading behaviors. The triumph of civilization is finding ways to curb or channel humanity into safe, productive, enjoyable lives - and that means overcoming much of what is natural.

It seems, informally and intuitively, that there exist norms which would support anarchy. The incentives to cooperate and incentives against exercising power over others are different than current norms, but not, I think, any more deeply opposed to human nature than current norms accepted in much of the world - such as property rights which forbid stealing and murder, or those which enable democracy, or those keeping religious communities viable, and so on. None of these sets of norms in absolute, but each can be viable and self-reinforcing if adopted widely enough, in the right circumstances.

So the key question for the viability of anarchy as a system isn’t, in my view, whether there can exist people and societies who could adopt it - I think that there are norms which could exist, that with mutually socially applied pressure to preserve those norms, some version of anarchy as described by Clayborne is possible, even if I cannot currently specify them exactly.

Clayborne says that ”while what we want depends largely on where, when, and with whom we live, it should not be so mysterious to figure out how to provide a baseline level of safety and security for all of us. Most political philosophies are banking on the idea that humans can and should be controlled, which necessitates homogenizing. But anarchy rejoices in everyone becoming ungovernable — a heterotopia, rather than a utopia.”

But for Clayborne’s vision to be realized, we don’t just need to say that the end state is possible, we need it to be possible to get there from here. For that to happen, it seems that any transitions involving norms need to follow incentive gradients. That is, we can’t adopt a system which requires cooperative norms in order to impose cooperative norms - and Clayborne is very clear about this, saying that the transition to anarchy would need to be effectively unanimous across large parts of the world. But this shows the much stronger requirement for a proposed system - not just a viable end state, but a path to get there.

Conclusion

The problem we’ve identified with anarchy as a heterotopian ideal is getting there, not whether the ends state is stable. And even then, I’m not sure an eventual anarchist state of the world would be stable - the pressures from small scall adoption of authoritarian structures might outcompete the anarchic system, or it may be the case that no stable anarchic system could be devised in the first place. We laid out why anarchy has scaling problems that are even more severe than the current cooperative global mechanisms. The problems of relationships and incentives are critical, and while we didn’t discuss the examples Clayborne suggests of anarchic success in native cultures, or in communal structures, these implementations are non-viable at scale, nor are they plausible competition for the modern world.

Of course, these fatal flaws are uncertain, and could plausibly be overcome - but they are irrelevant as long as there isn’t a way to get there. Marx, for all his many flaws, didn’t only outline a system for communism, he explained why there would be tremendous pressure to adopt it, and the revolutions were inevitable. And while his theory of value and his predictions were all not only wrong but damaging, he got the prediction of where pressure would come from, and how it would worsen, almost exactly right - unrestrained capitalism is self-destructive, and the working class eventually demands changes.

So I will conclude for now by saying that anarchy has other disadvantages, but the first and most critical one, in my view, is that it requires a set of norms and structures which don’t seem reachable from our present position. Of course, maybe the problem isn’t about anarchy, it’s about any future at all. So on the next post, I’ll argue that some much more utopian future is possible - and that the human and societal obstacles to overcome are surmountable.

1. von Mises and others discussed “Socialism,” and there are some really important conceptual pieces that I won’t get into, but I’m avoiding the term because the more recent usage of the term is very different.

2. The Once and Future King is very much about the tension between ideals and implementation, and the question of whether human nature is a significant barrier to ideal governance. Immediately before writing it, he published an autobiographical memoir, “England Have My Bones,” a paeon to British agricultural life, but his deep affection for the country contrasts with his ambivalence about the use of power by government in his retelling of the Arthurian legend.

3. By the 1880s, Jews were welcome to immigrate to anywhere in the Ottoman Empire - except Israel. But despite the massive effort to keep Jews out, they kept coming, fleeing even worse situations. In the following decades, both before and after the formation of the Zionist movement, the small agricultural settlements were formed. And the rise of global Communism, and the practical necessity of pooling resources to buy land, meant that a large fraction of these were communist. And by the outbreak of World War II, there were 79 different Kibbutzim - and while they were only 5% of the total population, they were seen by many as the very essence of the Zionist project.

4. The other reason for their decline was economic; small-scale farming isn’t particularly profitable, and professionals living on the Kibbutz, as well as television, showed those living on Kibbutzim that the alternative was pretty great. Privatization and other changes led to the near-complete collapse of the traditional model.

5. And the “strongly robust” piece is critical - in a previous post we talked about stag hunts, and the reason that people might defect if they are low on resources; even this failure mode doesn’t occur in the robustly positive cooperation!

6. This isn’t just about communism, but it’s historically closely aligned; communism requires centralizing power in order to centralize communal decision-making, which leads to selecting for strongmen, and that rarely goes well.

7. Cuba is the sole exception, but its relative lack of economic growth under Communism and increasing relative poverty shows the same trend.

8. This seems like a reasonable and somewhat predictive model of humans. Perhaps, however, accepting others being in charge is completely a contingent fact about humans today, and with changed norms this would no longer be true. And in that case, all such people would be happier in an anarchic society without others controlling them - and despite that, the desire for power seems pretty intrinsic to humans.