It is transfers, not taxes, that really generate such progressivity as we have in the United States. As Lane Kenworthy shows, the overall U.S. tax system — federal, state, and local — is not all that progressive in its effects, despite a very progressive graduated federal income tax. What low-income workers don’t pay in federal taxes, they make up for in state and local taxes, particularly sales taxes, which are basically a flat income tax for the poor. Kenworthy finds that each quintile pays about 30 percent of its income in taxes. But the system becomes much more progressive when transfers are accounted for.

Tax hikes on the so-called rich may decrease the private sector’s share of income, but they probably will not do much to decrease the real income of high-wage workers and may in reality increase government revenue at the expense of low-wage workers in the long term, though it is very difficult to disaggregate the complex relationships between taxes, wages, and prices. But those who say that they are most interested in economic inequality would do well to follow Kenworthy’s example and look at transfers rather than taxes. Means-testing Social Security and Medicare would do more to make the total package of taxes and transfers more progressive than any tax hike likely to pass Congress in the foreseeable future. It is also a reform that many conservatives and deficit hawks could support. This should be persuasive to those on the Left whose interest in tax hikes on the high-income is not strictly punitive, but I am afraid they are a very small minority.