Monday, July 13, 2009

The Invisible Hand Meets The Five Foot Antler

Listen, I got a question about economics that's been bugging me for a really long time. Maybe someone can help me out.

Let me take as a starting point here the excellent short piece in yesterday's New York Times by the economist, Robert H. Frank on the successes and limitations of free market competition. The successes, he points out, were seen and understood by Adam Smith. As Frank puts it, "when greedy people trade for their own advantage in unfettered private markets, they will often be led, as if by an invisible hand, to produce the greatest good for all."

The invisible hand: a powerful idea. But Frank says that though most economists probably think of it as the most basic and fundamental idea of their discipline, he predicts they won't do so for long. Because alongside the invisible hand, he says, we have the powerful example of evolution putting self-interest against the interests of the group. Sometimes individual and collective interests coincide; sometimes they don't.

Consider the example of the elk, he says. Though a mutation for huge antlers tends to help any individual elk, the resulting evolutionary pressure for larger and larger antlers is a disaster for elk as a whole, who now have antlers that are a hazard in general -- getting in the way, for instance, when the elk are pursued into forests by predators.

It's like an antlers arms race. If male elk could vote to scale back their antlers, with a kind of antler-regulation, Frank says, they would have excellent reason to do so. Likewise, he says, in those cases in which unfettered competition puts individual interests against group interests, humans have excellent reason to enact regulation to, in effect, scale back our antlers.

Frank cites examples of neighborhood schools and steroid use. If everyone works extra to earn money to try to get the house with the best schools for their kids, housing prices just go up, while schools stay the same. If everyone uses steroids in sports, then you must use them to be competitive, and everyone suffers. What is good for the individual is bad for us collectively.

Frank uses these points as an argument for regulation in certain cases. Sounds good to me. In fact, it sounds so obvious to me that I'm prompted to wonder again, as I have before, about why it gets so little attention.

The idea that what is good for individuals is sometimes good for groups, but not always, seems pretty obvious just from looking around, but it was confirmed in precise terms by game theory. In games like the prisoner's dilemma, pursuing what is best for you alone doesn't always lead to the optimal outcome for all. Indeed, a classic example is the arms race: if any country disobeys a disrmament treaty while its neighbors obey it, they do best as individuals. But if all countries obey, the outcome is best overall.

But game theory, and prisoners' dilemma cases, have been known to economists for decades. These are concepts familiar to most of my undergraduates. So if the conflict between individual and collective success is underappreciated by economists in general and free market proponents in particular, as Frank suggests and as seems plausible, what is going on?

That's my question: why do economists still talk so often in terms of the rationality of doing what is in one's individual self-interest, when game theory shows clearly that optimal outcomes in certain cases are only reached by not doing so, are reached only through treaties, regulation, and antler-decreasing voting?

Is it a normative committment to the idea that that it is better to allow people maximal individual autonomy even if the outcome isn't best overall?

Or is it a factual committment to the belief that the vast range of cases we tend to see in terms of conflict between individual and collective good -- like the housing prices-schools case Frank cites -- aren't really examples of conflict after all, appearances to the contrary?

Either way, I'm skeptical.

6 comments:

Daniel said...

Another response to The Times's piece.

http://thinkmarkets.wordpress.com/2009/07/12/frank-and-stein/

Daniel said...

Hi. Me again. I just read through The Times piece again, and I'm wondering if maybe the public school example isn't such a good one. The fact that people working hard to buy houses in a community with good public schools doesn't benefit the greatest number of folks (or however it is that the best or optimal is measured) but jacks up the cost of housing in the community seems to depend on the persistence of public education as a given. The commodity in question, the education, and the ways that it gets to be good, or bad, is left unquestioned, while the competition for this clearly imperfect system / commodity is brought in as an example of the failure of the invisible hand. I guess my question is, where should we rightly locate the failure (the theory of the invisible hand, or the system of education)?

Mario Rizzo said...

First, Frank's article is extremely poor intellectual history. He contrasts a child's version of Adam Smith with a more sophisticated version of Darwin. So Darwin comes out ahead. Ok.

Second, game theory does show cases in which the narrow pursuit of self-interest does not produce the best result socially nor, sometimes, the best result for the individiual. (Some of the basic results of game theory were known before there was formal game theory, btw) But the main question is the applicability of such theory to the real world. Game theory is notorious for being institutionally empty. Agents act in very thinly defined social contexts.

Adam Smith, on the other hand, recognized the social-embeddedness of human action. His Theory of Moral Sentiments discusses many of the social norms that constrain the behavior of individuals. The primary one is "justice" but honor, beneficence, etc have roles too.

As Thomas McQuade argues in the post linked above, the natural environment in which the Darwinian struggle for existence takes place does not have the kind of social context Smith talked about.

Self-interest (supplemented by biologically evolved propensities to care about family, friends, etc) is a powerful explanatory tool. It applies not only to marklets but also to the behavior of politicians, regulatory agencies, political pressure groups, religious organizations -- I believe that there is no apriori limit here.

Game theory, however, is not a very powerful tool beyond some very basic applications. Beyond the obvious, it gets so complex that applications are few and far between. I think of basic game theory as often showing that without institutions social cooperation would be very difficult. So game theory shows us what the world would be like in that abstract case. Abstract rationality is not useful. Institutionally-embedded rationality (self-interest) is.

Mario Rizzo

Patricia Marino said...

Excellent, thank you both so much, Daniel and Mario!

Yeah, I figured Frank's original piece wasn't fair to Adam Smith as a scholar. Though I'm not an expert, I understand Smith's approach to all these things was very nuanced.

So from what I understand from these responses, the suggested answer to my question is twofold:

First, you're saying, it's not a normative commitment but a factual commitment, the factual belief in question being that although we're tempted to see many cases as ones in which the communal good conflicts with the individual, in reality very few cases are like this. Hence the limited applicability of game theoretic concepts.

Second, in cases in which it might otherwise be true that these conflict, social forces, in the form of sentiments (for justice etc) and of social institutions (as McQuade mentions) provide the necessary constraints.

So would it would be fair to characterize this by saying that in a case like that of steroid use in sports, the idea is that sentimental and social attachments tend, or should tend, to make us naturally averse to what is "unnatural" or "unfairly competitive," or something like that?"

There are various reasons for my skepticism, but they start from considerations like 1) social institutions and conventions seem often biased and discriminatory -- when we say things are "unnatural" it's often to ends I don't want to endorse, and 2) to the untrained eye, it just looks like a lot of cases concern conflict between individual good and collective good. I'm perhaps swayed here by having read so many moral contractarians (such as Gauthier) who say that the essence of moral thinking can be found in prisoners dilemmas type reasoning, in which we act "morally" in order to do best, not individually, but overall.

Anyway, thanks!

Mario Rizzo said...

I think that there is a limit to our ability to deal with these issues on an abstract or theoretical level. This is because I am not saying that conflicts between the individual pursuit of interest and a broader notion of social welfare do not exist; they do. The questions are under which conditions and what the alternatives are.

First, I think we need to qualify or clarify the idea of self-interest. As Smith shows, we are not always or even usually narrowly selfish. It is hard to know what this would mean. Even Madoff shared his wealth with family and some friends. He bought things in voluntary transactions that benefited other people. So I prefer the term Bentham used "interest" -- the interests of the individual are sometimes more confined, sometimes more encompassing.

Second, we need to clarify the notion of "social welfare." This doesn't have a clear meaning in most discussions. And even when the intent is clear there is often no practicable way to measure it (as in the idea of aggregate happiness). Economists often mean social wealth -- a measure of the value of things in money. F.A. Hayek criticizes many notions of "welfare to society" as incoherent -- he coined the phrase "the morage of social justice [read income distributional fairness]"

So there is plenty of room for confusion at the abstract level -- what is "self" interest? What is "social welfare"? And then when we settle on these there is plenty of room for discussion of specific issues, including the causes of a apparent conflicts between the indiviudual and the rest of society.

For example, in the current financial crisis, some economists say: Selfishness is always there; what made this situation so different from all the previous cases? They say in this case the Federal Reserve policy of low interest rates made it likely that self-interest would involve serious risk taking and ultimately bad consequences.

Blogger said...

Get daily suggestions and instructions for generating THOUSANDS OF DOLLARS per day FROM HOME totally FREE.
GET FREE ACCESS TODAY