Yesterday Jon Bakija of Williams College presented his co-authored paper, "Jobs and Income Growth of Top Earners and the Causes of Changing Income Inequality: Evidence from U.S. Tax Return Data."
The paper looks at an excellent set of 1979-2005 IRS tax return data that includes the self-reported professions and industry sectors of people with high taxable income, in order to shed light on where the rise in income inequality has been coming from. Unsurprisingly, it's mainly salary income, such as of CEOs and people in the financial industry.
Obviously, to the likes of us at the colloquium, the issues of particular interest include the tax policy implications, as well as what light the paper sheds on alternative explanations for rising U.S. income inequality. Globalization and technological change are weakened as explanations - at least, as exclusive ones - by divergence as between countries that one might think were relevantly similar. The U.S. has of course led the way in rising income inequality, with the U.K. being closer to us than anyone else. So other stories must presumably be playing a role, including the reduced influence of social norms in restraining high compensation and an increase in "tournaments" that have a big winner and lots of losers. Reduced marginal tax rates may have played a role as well - though partly just with respect to labeling, such as by encouraging the owners of closely held companies to use S corporations or at least pay out more salary to themselves - but appear to fall well short of serving as a plausible primary explanation.
For my money, one of the big tax policy implications is as follows. Both executive compensation and the rise of the financial industry - again, two huge contributors to the overall story - involve flawed markets in which we have reason to believe that high earnings often do not denote high social value or productivity. While it may be true that the pre-mega stock option corporate executives of the 1950s through the 1980s had at times too weak an incentive to seek greater profitability, it's clear (as Lucien Bebchuk and others have shown) that we have now a fundamentally flawed market, where the incentives are very strong, but misdirected given all the short-termist and other games that executives can and do play. Likewise, financial sector profitability to a large extent reflects bad incentives, whether to find the correct price one second before everyone else (leading to personal gain far in excess of the social gain) or to do the sorts of things that the Greg Smith op-ed talked about.
In principle, one would want to respond to these problems by addressing executive compensation and financial sector issues directly, rather than sweeping in all high-income people whether their income comes from this stuff or not. (Although there would still be further reasons for greater progressivity in light of the changing income distribution - for example, the equity gains from redistribution are greater when the top has pulled so far away from everyone else, and it's more efficient to tax income at high levels when there are lots of people SO wealthy that these taxes are inframarginal for them).
But if one can't respond directly to the governance and financial sector issues, then by raising marginal rates one is not entirely misdirected - they cover a lot of the waterfront here, even though not all of it. So I consider these problems, and the sort of information that the Bakija paper provides, as strengthening the case for higher marginal rates at or near the top.