Experimental work on market-based or non-market-based incentives

July 9, 2012

(This article was originally published at Statistical Modeling, Causal Inference, and Social Science, and syndicated at StatsBlogs.)

Mark Patterson writes:

I found a discussion at the Boston Review that I thought you’d be interested in, given your posts on the potentially dubious foundations of many neoclassical economics models.

Michael Sandel cites a few examples of markets crowding out moral behavior. His longest discussion regards Frey and Oberholzer-Gee’s work demonstrating Swiss citizens’ willingness to admit a nuclear waste facility to town decreasing when offered monetary incentives. It seems like this is a situation that really demands a discussion of the available empirical evidence (Uri Gneezy and Aldo Rustichini have two papers, “Pay Enough or Don’t Pay At All” and “A Fine is a Price” that seem especially relevant.)

While the essay has sparked the usual sort of libertarian response, I’m struck by the fact that most people aren’t talking about the experimental work that’s actually available—it seems like this is the best way forward.

My reply: I don’t have much to add here, but this sort of discussion is fun. For example, Sandel writes, “Consider friendship. Suppose you want more friends than you have. Would you try to buy some? Not likely. A moment’s reflection would lead you to realize that it wouldn’t work.” That’s not completely true, though. I hire postdocs and assistants and they are somewhat friend-like. And sometimes I’m friends with students, who are (indirectly) hiring me. I don’t know if Michael Sandel is friends with Anne T. and Robert M. Bass, but it’s possible.

I agree with Patterson that it would be good to bring the discussion to the level of empirical research. But this can be difficult. I was involved a couple years ago in a project involving the gift of low-cost stoves to poor people in Africa. There’s been a debate for years about whether it’s better to give these sorts of things out for free, to sell them at or below cost, or to involve local middlemen who can make money off the deal. This sort of decision arises in all sorts of development aid, from factories to mosquito nets, and there are persuasive arguments on all sides. My impression is that the empirical work in this area leads to no clear conclusions, and I am suspicious of various clever-clever arguments in this area.

I’m also reminded of Dale Miller’s classic article, The Norm of Self-Interest. As I wrote in comments:

Tautologically, all actions are self-interested in the sense that people are doing whatever they happen to be doing. For example, if person A goes hungry so as to feed a complete stranger (person B) who is even hungrier, this could be viewed as self-interested in the sense that person A must have received some psychological benefit from it.

When we speak of “self-interest,” we’re going beyond this tautology to speak of actions that benefit oneself instrumentally. Webster’s describes self-interest as “a concern for one’s own advantage and well-being.” Yes, you can define well-being completely tautologically to mean “whatever you happen do be doing” but that’s not very helpful. I think it’s more reasonable to describe some–maybe most–of our actions as self-interested, but with other possible motivations also.

Please comment on the article here: Statistical Modeling, Causal Inference, and Social Science

Tags: ,