And so Reynolds is once again on the Wall Street Journal editorial page declaring that income inequality is a statistical illusion brought about by technical changes in the tax law that alter what income gets reported to the Internal Revenue Service and what income does not. "[W]hat the Congressional Budget Office presents as increased inequality from 2003 to 2007 was actually evidence that the top 1 percent of earners report more taxable income when tax rates are reduced on dividends, capital gains and businesses filing under the individual tax code." Um, Alan? The CBO report documented increased inequality from 1979 to 2007.
As the foregoing demonstrates, Reynolds's technique is to bury you under a mountain of hard-to-follow, often irrelevant, and sometimes entirely erroneous statistical quibbles to the point where you cry "Uncle!" and agree to believe that the existence or nonexistence of inequality is a matter of personal taste, like preferring Cherry Garcia to Chunky Monkey. Brad DeLong calls it "intellectual three-card monte."
This latest piece ends with the triumphal assertion that the CBO report concluded that income taxes were more progressive in 2007 than in 1979. But Reynolds doesn't tell you that the CBO report also concluded that the totality of federal tax and benefit programs were about one-quarter less redistributive in 2007 than in 1979. He also doesn't tell you that the income tax got more progressive because of changes at the bottom (mainly expansion of the Earned Income Tax Credit), not because of changes at the top. At the top, effective tax rates went down. That was detailed in an earlier CBO document from 2008. It showed dramatic declines for top incomes between 1979 and 2005 in both the effective income tax rate and in the effective rate for all federal taxes. For the top 0.01 percentile (i.e., incomes above $10.6 million in 2005 if you include capital gains and above $6.7 million if you don't) the effective income tax rate dropped from 21 percent to 17 percent, and the effective rate for all federal taxes dropped from 42.9 percent to 31.5 percent. Changes in income-tax rates, surprisingly enough, haven't benefited the superrrich as much as other changes in federal tax policy.
A central problem with Reynolds' argument that a lay person may grasp without much difficulty is that even though the technical changes he describes have, during the past 32 years, gone in different directions at different times--for instance, the capital gains tax went up in 1986 and down in 1997--income share for the top 1 percent has pretty relentlessly gone up, falling only during economic downturns, and even then not very far. That's true if you include taxes or not, if you include capital gains or not, etc., etc. (You can slice and dice the data ad infinitum at my favorite Web page, the World Top Incomes Database.) To believe Reynolds's argument you have to believe that the rich are reporting steadily more and more previously-hidden income to the IRS. We should be so lucky!