Cooking Unemployment Data to Make the U.S. Look Better
Laurent Guerby made a post on the prior topic about European-U.S. unemployment comparisons, I was just at a conference sponsored by the OECD where exactly this issue came up. The basic point is that proponents of the U.S. model want to add people in employment training programs and disability roles in Europe to their official unemployment rates for purposes of international comparisons.
This seems bogus on several grounds. First, the employment training programs are obviously heavily subsidized by the government (often 100 percent), but there are many situations in the U.S. where jobs enjoys substantial government subsidies. The EITC peaks at more than 35 percent of wages, throwing in work related child care benefits can easily push the subsidy to more than half of the wage. At what point do we say that the job is simply concealing unemployment, a 60 percent subsidy?, an 80 percent subsidy?, or does it have to be 100 percent? Furthermore, what if the government paid the full wage, and the person did absolutely nothing, but we didn't call it a training program? Then is the person unemployed? Suppose we can keep a person from looking for work by paying for their school and a small stipend (or a low interest loan). Should this person be counted as unemployed?
I don't think that there is a consistent definition under which people in European training programs can be counted as unemployed which would not require people in other subsidized employment (both in Europe and the U.S.) from being counted as at least partially unemployed. (Subsidies also don't have to go through the government. Prohibitions on age discrimination mean that younger workers subsidize the purchase of employer provided health insurance for older workers.) In short, this seems like a bogus measure of unemployment.
Similarly, if we grant that the people on Europe's less strict disability programs are at least somewhat less able to work on average than the non-disabled, then the question becomes the determination of disability. There is no basis for saying that the U.S. definition is the correct one, and it is an entirely reasonable decision on a society's part that they would be willing to pay more in taxes so that people who have some impediment that makes it difficult to work, do not have to work. It is reasonable to have a more strict standard as well, but no one made the U.S. Congress god, so that its rules should be the universal standard of who is disabled.
There is a reasonable question to be asked about the overall sustainability of a given situation, but if the comparison is with the Nordic countries, the U.S. loses badly. They have small budget deficits or surpluses and current account surpluses. The U.S. has a large budget deficit and clearly unsustainable current account deficit.
On the other side, the U.S. prison population (2 million) is approximately 1 percent of our working age population (15-65 for international purposes). Also, our official data probably overstate the employment rates by about 1.5 percentage points because people who are likely to be unemployed do not respond to the survey. (The coverage rate for young black men is about 70 percent.) My colleague John Schmitt recently did a “paper” on this.
In short, I think it's great to try to look at the unemployment/employment data more closely when making international comparisons, but I'm not convinced that a closer look improves the relative standing of the U.S.