Economists, until recentily, have described the world using the assumption that all agents in the economy have access to all of the information that is necessary to make their decisions in the way that makes them best off, so-called "perfect information" assumption. This has all kinds of incredible implications. Gary Becker (U Chicago) pointed out, for example, that this implies that drug addicts are rational addicts -- they start taking drugs with full knowledge that they will become addicted and are happy to become so. In his paper he writes this without a hint of irony, though it is a ridiculous conclusion. Assuming everyone has perfect information also tends to imply that all unemployment is voluntary -- unemployed workers have no job not because they can't get one, but because they are simply unwilling to work at the going wage in a job in which they could get hired; this is a rational choice. Sorry, I don't believe this one either.
By the way, I'm an economist.
In any case, though, these examples amount to reducto ad absurdum for me -- the perfect information assumption is simply untenable in many contexts. Economists who support it generally defend the use of this strong assumption with a very weak argument: that geesh, if consumers and workers aren't rational decisionmakers with perfect information, we just don't know what they're gonna do! (So we might as well stick with a simple model with useless predictions!)
However, I have a different view: I think trying to understand how the world really works without perfect information is much more interesting. For example, why are so many consumer products branded? One reason might be that is that it is costly for consumers to find out detailed information on the quality of different products, so brands substitute as a proxy for product quality. Why do people become addicted to drugs? One reason might be that they systematically misevaluate what effect taking drugs now will be on their future happiness. Etc. Determining whether these things are true is, I find, a much more interesting task than continuing to push to prove things that are hard to believe because they fit into a simple framework.
In addition, it matters for policy. If drug addicts are rational, there is no need to prohibit drugs in any way, or even to prevent young children from getting started using them. If drug addicts are not rational, then there is a need to intervene. If brands are proxies for product quality, it changes how we think of competition -- the ideal situation for consumers may not be many small firms (as is suggested by perfect information) but rather allowing firms which can build up a brand name to grow large in size.
To non-economists, it may sound obvious that we should try to understand how the world works and form policy based upon that, and fortunately the field of economics appears to be headed towards taking this approach. So I'm ready for the fun ahead -- I'm rationally addicted to economics.Posted by ethan at March 08, 2003 10:40 PM