November 6, 2013
-
The Logic of Collective Action
JDN 2456594 PDT 15:55.
The blinders neoclassicists wear are large indeed. This book was written in 1971, and the neoclassical understanding of collective action has, as far as I can tell, not been substantially advanced since then—in over 40 years, the essential conclusions haven’t changed.
This would not be a problem, of course, if those conclusions were correct; but they aren’t. Indeed they are wrong at a fundamental level; they are almost literally reversed from the truth.Here is the basic argument:
1. The interests of a group are not the same as the interests of its individual members. [This is true; moreover, I don't think it was recognized by most people before Olson. So that could be marked as the significant achievement.]
2. Individuals act in their own self-interest. [Fundamental neoclassical assumption; and herein lies the problem.]
3. Therefore, groups do not act in their own self-interest.The logic—and the title did say “logic” after all—is absolutely valid. The problem is that the argument should actually run the other way:
~3. Groups act in their own self-interest.
1. The interests of a group are not the same as the interests of its individual members.
~2. Therefore, individuals do not act in their own self-interest.That answer has been staring them in the face all along… and yet even now in 2013 most economists still fail to recognize it. Humans are neither selfish nor altruistic; we are tribal. We identify ourselves with a group, defining those inside as good and those outside as bad. The definition of the group can vary a great deal; it might be Americans, or White people, or Christians, or Red Sox fans, or economists, or University of Michigan alumni; and indeed most people will identify in varying degrees with many different groups at once—and much of the conflict in our lives comes from being torn between such identifications. (The rest probably comes from conflict between the groups we identify with and that one special group that contains only ourselves.)
Actually, even neoclassicists have been forced to accept the most extreme examples: They speak of “households” and “firms” as though they were indivisible entities, even though a household contains several individuals and a firm may literally contain millions. Why does this work? Because they are indivisible, at least in terms of their tendency to act in their own self-interest. A corporation is far more likely to act in its own self-interest than any individual human being would be. Households are also not strictly selfish, but they are at least close to it sometimes; whereas, anyone who would stop feeding their own child because they ceased to be amusing we would all immediately and rightfully recognize as a horrible and defective human being. (Yet this is how you would behave, if you actually acted in your own self-interest independent of that of your “household”, that is, your family.) True, they might be punished for it; but then again they might not (most child neglect goes unreported), and in any case, this does not factor into a normal human being’s decision process. We do not ask ourselves, “Would the police punish me if I stopped feeding my children?”; indeed, we do not even ask, “Should I feed my children?”; we simply feed our children, whenever it is possible, without hesitation. Indeed, we will go to great lengths to feed our children, even if it harms ourselves substantially.
That makes perfect sense in terms of evolutionary psychology—indeed, it would be completely baffling if it were not true, and might force us to radically redefine our understanding of evolution. But it is completely ignored when we assume that human beings are rational self-interested economic agents.
Olson even slips into arguing for group cooperation himself sometimes, apparently without realizing it: “[…] a more detailed analysis of the kind outlined above could help to explain the apparent tendency for large countries to bear disproportionate shares of the burdens of multinational organizations, like the United Nations and NATO,[...]” Once you start saying that The United States acts in its interests, you have already committed to an enormous scale of group cooperation. (Moreover, I also question whether he is even correct that this would be something to explain; it is not obvious to me that the disproportionately large amount of money the US gives to the UN offsets its disproportionately small amount of soldiers, for example. It could just be comparative advantage, or even the US bearing too little a share.)
Throughout the whole book, Olson seems completely unable to fathom why anyone would every do anything for anyone other than themselves.
He says things like “If the groups, or at least the economic groups, are often interested primarily in their own welfare, it could only be so because the individuals in these groups were primarily interested in their own welfare.” Why? In fact, that is one thing we know doesn’t work—as Olson himself goes on to say: “But if the individuals in any large group are interested in their own welfare, they will not voluntarily make any sacrifices to help their group attain its political (public or collective) objectives).” Olson goes to great lengths to prove that cooperation requires altruism, but simply dismisses the idea that altruism could actually exist.
He likewise dismisses the notion that people have an innate instinct for cooperation as “meaningless”, when in fact it is the consensus conclusion of modern evolutionary theory. It’s even formalizable; you can actually quantify the human instinct for altruism, and it’s quite notable that the figure considerably exceeds the calculation we would get from simple kin selection, C < rB. If you're anything like me and most people, you've paid more than rB to give someone else B on many occasions in your life, and will probably continue to do so many times in the future. You probably haven't paid more than B, which would usually be bad—though some people do so, and it may not always be bad. You are probably even unlikely to pay exactly B, though at times you may. Instead, you pay pB,where p is a discount factor, functioning akin to r—but no longer equal to your genetic relatedness. Instead, it might be called your "cognitive relatedness", or to use a simpler and more evocative term, your solidarity. Usually p far exceeds r, though in some cases it might be less—e.g. if you have had a terrible falling-out with your sibling and now avoid all contact with them. I also suspect that for identical twins, even though r = 1, p < 1. Identical twins usually care about each other a great deal, but it seems unlikely that they value each other exactly equal to themselves. p = 0.99 seems like a reasonable guess.
The reason we call it an "instinct" is of course that it is, and this has a rather precise definition in evolutionary biology—a genetically encoded, evolutionarily selected behavior pattern that is triggered in response to certain environmental conditions. We have a moral instinct in exactly—exactly—the same sense in which we have a hunger instinct and a sexual instinct. One of the more important unsolved questions in modern evolutionary psychology (and moral science generally) is how much of human moral behavior is genetic and how much can be changed by environment; but there is no serious dispute about the fact that both are important—and that means that by definition we have a moral instinct.
Yet Olson doesn't seem to understand this instinct in the slightest, which makes him seem like he thinks all humans are psychopaths. And indeed, The Logic of Collective Action is a book about how people would behave if they were neoclassical rational agents, which is to say, omniscient psychopaths.
Don't get me wrong, there are some useful insights in this book. It's worth studying the ways that individual and collective decisions can be at cross-purposes. There are good reasons to have enforcement mechanisms in your system of policy, just in case people are tempted to act against the interests of your group in favor of themselves (or in favor of another group!). It's worth considering the fact that many forms of cooperation, from international treaties on down to class projects, are far below their optimal levels, and trying to find ways we might fix that.
His analysis of how the most stable groups provide both collective and non-collective goods simultaneously is also useful; it goes far to explain why churches hold bake sales and lobby politicians. His explanation of why the lobbying groups for oil companies and utility companies is so powerful is sound. His concept of "selective incentives" can be adapted, I think, to include our concepts of in-group and out-group and social belonging.
But if we were really atomistic individuals, acting always in our own self-interest, there would be virtually no cooperation at all; even enforcement mechanisms are costly, and if we were all the psychopaths we are imagined to be, we wouldn't even be willing to punish people who do bad things, because that takes effort and time (and in a modern economy, money). In reality, of course, we are very willing to punish people, even at rather substantial cost: Heinrich et. al. (2006) "Costly punishment across human societies." Science 312(5781): 1767-1770.
The few neoclassical economists who acknowledge this fact at all make ridiculous excuses for it, like "People enjoy punishing others for wrongdoing; they gain utility from a sense of righteousness." Okay, that may be true to some extent… I doubt it's the main motivation, but even if it is: Why is that true? Because we have a moral instinct. Yes, sometimes it feels good to do the right thing; but why does it feel good? Moral instinct. That is literally the only plausible explanation. Psychopathy is at a selective disadvantage.
(Given that this is the case, you may wonder how there are psychopaths at all. It's probably due to frequency-dependent selection: In a world of moral individuals, it can be adaptive to be a psychopath. But a population of psychopaths would rapidly self-destruct. Under frequency-dependent selection, the two traits converge to an equilibrium where they are equally adaptive.)
Only at the very end of the book does he acknowledge any motives other than self-interested calculation; and as neoclassicists are wont to do, he immediately declares them "irrational". I do appreciate that he at least mentions the possibility, because I've seen others who refuse to even do that.
There are of course limits to human altruism; and that is well worth studying. Many people are apathetic in large groups, and it's worthwhile to try to figure out why that is and what can be done about it. People are not always altruistic to everyone, and they do not always do their fair share. The tension between self-interest and group interest—and between different types and scales of group—is fundamental to economics, and indeed to human nature.
But you're never going to understand how that works if you pretend one side doesn't even exist.
