If Greg Mankiw did not know that his latest column on the incentive effects (on Greg) of higher marginal tax rats would be ill-received, then he is somewhere along the autism spectrum. Tyler Cowen tries to return the discussion to issues of status. Let me make a few points.
1. Sometimes, pay conveys status. To the extent that this is true, the marginal tax rate could be greater than 100 percent without affecting behavior. For example, if you want to be the highest-paid person in your field, then you can enjoy that status benefit no matter how high the marginal tax rate.
2. People often care more about status than pay. Most college-educated parents would rather see their offspring work for $40,000 a year at a "professional" job than earn $60,000 a year as an air conditioning repairperson.
3. Perhaps one reason for the steep decline in the employment/population ratio this century is that people are reluctant to search for lower-status jobs. It could be that, at the margin, the value of income is not as high as it once was, but the importance of status remains high.
4. It is possible to increase public school teacher pay relative to that of other occupations and yet not raise the status benefits of teaching. I suspect this has happened over the past thirty years.
5. As the time horizon gets long, the effects of monetary incentives and status motives may be hard to disentangle. In the short run, you raise marginal tax rates, and few people reduce work effort, because the status of being "employed" is much higher than the status of being a homebody. However, once a few people decide to become homebodies, the status of being a homebody goes up enough that many people choose not to work. So the long run effect of the higher marginal tax rate is much higher than anything you might have predicted, because it has affected cultural norms. I worry about this much more near the median of the income distribution than at the very high end.