The Post had a good piece noting the large number of people dropping out of the workforce, presumably because they can't find jobs in the weak economy. However the problem is likely worse than the piece indicates.

There are a large number of people who do not respond to the Bureau of Labor Statistics' Current Population Survey (CPS), the standard survey used to measure labor force participation. In recent years the non-response rate overall has been close to 12 percent, as opposed to just 5 percent three decades ago. The non-response rate varies hugely by demographic group. For older white men and women it is 1-2 percent. By contrast, for young African American men it is close to one-third.

The Bureau of Labor Statistics effectively assumes that the people who don't get picked up in the CPS are just like the people who do. This assumption may not be plausible. The people who don't respond may be more transient or may have legal issues that make them less willing to speak to a government survey taker. For these reasons they may be less likely to be employed than the people who do respond to the survey.

My colleague, John Schmitt, examined this issue by looking at the 2000 Census (which has a 99 percent response rate) and comparing the employment rates overall and for different demographic groups in the CPS and the Census for the months when the Census was conducted. He found that the overall employment rate was 1.0 percentage point higher in the CPS. For groups with high non-response rates the gap was larger, with a gap of 8 percentage points for young African American men.

These results imply that the problem of people dropping out of the labor force is likely much worse than is generally recognized. The problem is that many of those who are dropping out are not responding to the surveys we use to measure the problem.



I forget to mention, the reason that so many people are just dropping out of the workforce now is the shortening of the period of extended unemployment benefits. As long as people are receiving unemployment insurance they have to be looking for work. When their period of eligibility ends, most people just drop out of the labor force. The period of extended benefits was shortened in most states at the end of 2012. As a result, many people went from being classified as unemployed (no job, but looking for work) to being out of the labor force (no job and not looking for work). This was reflected in a sharp drop in January in both the average and median duration of unemployment spells, since many of the people who had been out of work longest were no longer counted as unemployed.