Income Inequality

Alan Reynolds wrote a fabulous piece in the Wall Street Journal last Thursday concerning America's alleged income inequality. It is a masterpiece in demonstrating that attention to details, and knowledge of data sources, can make all the difference between a valid conclusion, and useless speculation.

Specifically, Reynolds, who is a senior fellow at the Cato Institute, and co-founder of Polyconomics, identifies the source of Senatorial candidate Jim Webb's statement that,

"the top 1% now takes in an astounding 16% of national income, up from 8% in 1980."

The researchers responsible for these numbers, Thomas Piketty, of Ecole Normale Superieure in Paris, and Emmanuel Saez, of the University of California at Berkeley, used tax returns for their denominator, rather than total income. Major income sources which they omitted, according to Reynolds, were Social Security and other transfer payments. Of course, we would expect these types of payments to go to lower income earners, thus further skewing the findings of the two researchers.

Other sources of error include the non-filing of some income earners, municipal bond and other tax-exempt income sources, and the inclusion of two, joint filers on single tax returns at the higher end of the tax-reported income spectrum. Reynolds estimates that total personal incomes, the denominator for the inequality statistics, was roughly $3.3B in 2004, or slightly more than 1/3 larger than Piketty and Saez estimated.

Another source of error which Reynolds discusses is the tax-policy-driven shift of small, formerly conventionally-filing corporations, to Subchapter S corporations. These are often higher income sources taking advantage of a different reporting structure, thus improperly appearing to inflate upper incomes, when, in reality, they were simply measured as business income sources in prior years.

Reynolds' fine, detailed and sensible work demonstrates how easily such inflammatory statistics as the ones Senator-elect Webb (D-Va) used can become commonly held "wisdom," or "fact."

In reality, it appears that the work on which those statistics are based was very flawed, and has generated suspect results. As time-consuming, dry and tedious as it may be, doing the fundamental work of investigating definitions and research methodologies can often shed valuable, and, occasionally, disconfirming light on apparently important and troubling results. Such is the case with the now-well-publicized 'increasing income inequality' in America, as demonstrated by Alan Reynold's good work and articulate explanations.