Last week I noted that, in an hour of searching, I couldn’t find any defenders of the economic-theory-supported tall tax, even among economists. Today I report it was easy to find supporters (1, 2, 3, 4, 5, 6, 7, 8) of Alesina, Ichino, and Karabarbounis’s proposed man tax, also well supported by economic theory:
Gender Based Taxation … changes spouses’ implicit bargaining power and induces a more balanced allocation of house work and working opportunities between males and females. Because of decreasing returns to specialization in home and market work, social welfare improves by taxing conditional on gender. When income sharing within the family is substantial, both spouses may gain from [it].
So riddle me this: if careful economic analysis had instead favored taxing men less than women, how many supporters do you think that proposal would have found, even among economists? And what does that say about how much economic theory influences economists’ policy conclusions?
P.S. The best "man tax" is not a fixed tax amount for being a man, nor is the best "height" tax a fixed amount per inch of height. Instead, each gender or height has a different income tax schedule. So there is no special reason to fear some "couldn’t pay the tax."
Added 20Dec: Here’s a paper on a more general tax on genetic features.
"And what does that say about how much economic theory influences economists' policy conclusions?"It says that this kind of economic reasoning is very plastic and can be twisted to support a wide range of conclusions, and generally will be used to support whatever is most expedient for the author. When you read these papers, you need to be wary of every step.One trick which is often used is for the author to mention a few assumptions explicitly. This is done in order to distract you from all the other assumptions they are implicitly making but not mentioning.For example, let's suppose there is an emasculation factor e which causes disutility to the man and increases as his wife's wage rises. That could change the conclusion of the paper, but I don't see this factor in any of the equations.Or, assume there is a p(a|acw) (propability of adultery occuring given presence of attractive co-worker) that increases as a function of market participation rate. I don't see that in the equations either.
It seems the problem is the statistical definition of "endowment." Endowment is a causal concept, it refers to things that 'cause good fortune.' Presumably we ought to spend the research effort into figuring out direct causes of good fortune, rather than wasting time taxing wealth correlates which may or may not have a causal basis.