« [SSJ: 20] Response from Reed on Rational Choice | Main | [SSJ: 22] Cite search, Japanese interest groups »
June 4, 1995
[SSJ: 21] Response to Reed on Rational Choice
From: David C Kang
Posted Date: 1995-06-04
The real question regarding RC or anti-RC should be what it tells us about an empirical question. When is RC useful and under what circumstances?
When I think about RC or anti-RC, I always come back to an empirical puzzle. Does using RC help us understand the world better, or help us resolve debates? Lets take the infamous Johnson-Rosenbluth disagreement over MITI and politicians. Johnson says MITI dominates, Rosenbluth says that Politicians have some power. The debate should be not over methodology, but how we resolve this puzzle. For those more inclined to the Johnson opinion, what evidence would be convincing that politicians have the power? The same question should be applied to those more supportive of Rosenbluth: what evidence would support the Johnson approach and contradict your own? If the answer is "nothing will ever convince me that I'm wrong" well, then you do have an ideology, and not a theory. If the response is "if you disagree with me you must be an idiot" we're also not getting anywhere.
Let's quit quibbling about whether people are rational or they have preferences. This is a tired debate that really doesn't go anywhere. Everybody (including rat.choice modelers), as Frances Rosenbluth pointed out, realizes that people aren't perfectly rational with perfect information. The real test for me is whether a rational-choice approach provides some new way of looking at the world.
Approved by ssjmod at 04:08 PM