What is the Distribution of Income in the Country and Why does it Matter

Alan Reynolds argues that the standard measures of income inequality distort the actual realities in an article in the Wall Street Journal this morning. It forced me to go back to a body of work that I have looked at in the past. There are two sets of issues here. One is methodological, how should we determine income shares? The second is political, why does it matter that the income share arguments are wrong.

In the methodological arguments, the questions are complex but fairly clear. Beginning in 1993 (in the Quarterly Journal of Economics - Vol CXVIII, Issue 1) Thomas Piketty of EHSS in Paris and Emmanuel Saez of UC Berkeley began a series of studies looking into income inequality. The two scholars have periodically updated their work.

The numbers from Piketty-Saez, if correct, should be of concern. They argue that the top 1% of taxpayers hold 16.1% of the total income and that the share has increased substantially from 2004 (when it was 14.7%) and 1980 (when it was 9.3%). And there are a number of visuals that can confirm that trend. (Bonuses at Goldman, which average more than $600,000 were a major story on ABC news last night.) But visual impressions should not guide policy.

Reynolds points out a number of problems with the model used by Piketty and Saez. First, their model does not include transfers (such as Social Security) which are not counted in tax data and are skewed to the lower end of the sample. Second, alternative data on incomes suggests that the total figure in the income totals is about a third higher than the one used for the data in the Piketty/Saez paper. Third, the data understates middle and lower income incomes because its figures fail to account for such middle class savings vehicles as §529 plans and §401K plans. Fourth, it fails to account for the significant increase (estimate by one source at a growth rate of 9% per year since the 1986 tax act) of movement from corporate to Subchapter S corporate returns. Those returns are business income that seems to be reported as personal. The net effect of all of these differences is that middle income income is underreported and higher income income is over-reported AND that the trend for both seems to be accelerating.

But then there is the political side of it. In a recent Rolling Stone article headlined The Great Wealth Transfer liberal darling Paul Krugman argues that not only is the data on wealth correct but that it is even worse than suggested by the two real academics. "The reason most Americans think the economy is fair to poor is simple; for most Americans it really is fair to poor.", Krugman clucks. One wonders whether Krugman would say anything different even if the data were different. He then goes on to compare the 1970s and today - 1) the common employee in the 1970s worked for GM and had generous defined benefit pension plans and other benefits and now works for Walmart with no such things. That is a nonsensical comparison. The employee in the 1970s did not have a §401 plan and much less control over his own fate than today. 2) Krugman then goes on to argue that more unions and a higher minimum wage would save the day. That is something that most legitimate economists would disagree with. 3) He then goes on to suggest that the corporate scandals were somehow covered up by the invasion of Iraq. That is absurd on its face but hasn't he heard of Sarbanes Oxley and doesn't he own a calendar (we went to Iraq before he dates the corporate scandals)? Lefty places like the Center on Budget and Policy Priorities trumpet the data as they did in October when they suggested that there was an "Extraordinary Jump" in inequality in the country.

Here is what we know. First, workers in the middle class have much more control over their lives (in things like §401 savings opportunities) and less apparent security than their counterparts in the 1970s. The flexibility of employment works both ways. The benefit or cost of that depends on some value judgments. I happen to think that individuals benefit from more freedom, but benefit is in the eye of the beholder. Second, by most reasonable comparative opportunities today's middle class is better off - larger square foot houses and better technology - which may or may not show up in the income statistics. I am amazed every time I go back to my aunt's house in North Carolina to understand just how small their very prosperous family home is by modern standards. Third, we understand that the Top 1% in income (or 5% for that matter) are not necessarily the same people in two time periods. Part of this requires some value judgements but part also require a clear headed analysis of data.

I am interested in this set of issues, because I believe that the ability of people to make their own way and to benefit from their energy is critical to maintaining democratic institutions. Krugman is not interested in the facts here but merely a way to push his own centralized and looney alternatives. Here the wrong data can lead to wrong conclusions.