Matthew Yglesias writes about a study that shows that very high top marginal tax rates might grow the economy, by moving people from lucrative-but-socially-valueless jobs like bankers and lawyers to less-lucrative-but-more-valuable jobs like teachers and scientists.
I question the premise. In a very simplified economic model, in which there is a straightforward income or wealth tax with few or no exceptions, perhaps you would indeed get fewer bankers and lawyers with a higher top marginal tax rate. But this is the spherical cow of economic models. In the real world, a higher top marginal tax rate is likely to produce more bankers and lawyers (and lobbyists, by the way), because the very rich will have more incentive to find tax shelters and put more of their money into them. If you assume no tax shelters, you’re in an academic abstraction just as removed from the real world as the physicist who assumes spherical cows.