Mixed Signals on the Health of the Job Market

The BLS employment report for January offers a puzzling and mixed picture of the health of the current job market. The household survey shows a continued and large drop in unemployment, caused almost entirely by an increase in the percentage of Americans who report holding a job. Respondents to the BLS household survey reported a jump in employment of 589,000 in January following an increase of almost 300,000 in December. With employment gains this large it is hardly surprising the unemployment rate has fallen, dropping 0.8 percentage points over the past two months. The jobless rate has now declined more than a percentage point since reaching a recession peak in October 2009.

The Labor Department’s employer survey offers a much bleaker picture of the current job market. On a seasonally adjusted basis, January job gains were estimated to be just 36,000, a disappointing slowdown compared with rates of employment growth in recent months. Some of the slowdown was due to bad winter weather, but over the past three months job gains have averaged only about 83,000 a month, a rate of increase that is too slow to reduce the unemployment rate.

As was the case last year, government employment was weak in January. Total public employment has shrunk 253,000 over the past 12 months, with virtually all of the loss concentrated in local government payrolls. Given the weak fiscal outlook of many state and local governments, these numbers are not likely to improve over the next year. Therefore, the nation’s hope for an employment rebound depends on the appetite of private employers for new workers. The latest job report and the Labor Department’s revisions in the payroll numbers for last year suggest that private-sector demand for additional workers remains weak.

When there is a large discrepancy between the household and employer surveys, it is natural to ask which is more likely to be accurate. Statistical analysis of the historical record suggests there is little reason to put greater weight on one survey than on the other. In measuring month-to-month changes in the employment situation, there is some justification for slightly favoring the trends uncovered in the payroll survey. Over a two- or three-month period, however, there is not much evidence that the employer survey is preferable to the household survey.

Both Labor Department surveys agree in showing that the 2008-2009 recession was the worst downturn of the post-war era. The two are also consistent in showing that the job market rebound since late 2009 was unusually weak. The two surveys do not offer a consistent story about the health of the job market in the past couple of months. The employer survey shows very meager job gains; the household survey shows robust employment growth and a sharp drop in the unemployment rate. It is wrong to think the truth must lie exactly halfway between these two possibilities. However, a betting man would not put all his money on just one of the surveys either. Both surveys show continued gains in the job market. It may be several months before we know whether the pace of recovery is fast enough give hope to the nation’s unemployed.