Tuesday, April 30, 2013
One Thought about the Seeming Change in Income Distribution
There has been a lot of discussion in the last few years about the seeming growth in income inequality. That first came from a pair of unlikely names called Thomas Piketty (who is an economist at the University of California) and Emmanuel Saez (so called Pikkety Saez paper) who have argued in a number of papers that income inequality is getting larger and that the middle class is disappearing. The PS data sets are both from IRS and Census Current Population Survey (CPS) data - supposedly unimpeachable. Unfortunately, according to Richard Burkhauser, who is a Cornell economist, their use of data is flawed. In essence, if you make minor changes in the way data is counted and depending on which data set you choose to use, the data varies widely from that reported by PS. Burkhauser wrote a paper for the NBER (which is gated but pretty inexpensive) which disputes the PS conclusions. Most of the left has taken the PS data a gospel, and indeed, will vilify people who question the PS conclusions.
From my read of Burkhauser's paper, his conclusions are sound. A lot of interpretation depends on how you count income. For example do you use household or taxpaying unit data? Do you include AGI or gross income? Each of those variables can vary the data considerably. And if you use the Burkhauser methodology you come out with very different results than PS.
But as I have watched us come out of this recession, I have been more concerned with labor market participation, which has (even before Obama) been declining. It turns out that at least a partial explanation of that change in behavior is related to the explosive growth in Americans claiming Social Security Disability payments. SSDI - now equals 5% of the people between 25 and 64. That is double what it was in the 1990s. What is more the recipients differ from earlier times in three significant aspects. First, the rapid growth in SSDI recipients must be part of the explanation on why labor market participation is declining. Second, current recipients are generally younger than in previous generations. Thus, they will be on SSDI for a longer period of time than they might have been before the program started to expand. Third, a significant portion of the new SSDI recipients are claiming SSDI with soft injuries (mental and tissue related disabilities - which are significantly harder to observe and congruently easier to fake). Here is the kicker - all those younger people who are on SSDI have consigned themselves to an income structure which is not as lucrative as many other employment possibilities but they are willing to accept the lower incomes in part because SSDI also comes with a medical program that is similar to Medicare.
There are two possible explanations of what this might do to income distributions. First, that people would choose to live on a modest annuity at an earlier age than in the past might show how truly dreadful employment prospects for the undereducated in society are. Second, increased utilization of SSDI may in fact have changed the income distribution not based on available wages but on the perverse incentives established by SSDI. I am not sure which of these explanations is correct; indeed both may be. But it does suggest that even if the PS data is accurate, as far as it goes, the change in distribution of income has come about as a result of the negative effects of governmental policy.