U.S. Employment Survey Discrepancy Increases

Markets reacted positively (initially) on today's U.S. employment report. And the payroll survey was indeed strong-at least by the standards of the last few months. Payroll employment increased 117,000 and even more impressive was a 0.4% increase in average hourly earnings. This suggests a nominal increase of 0.5% (6% at an annualized rate) in aggregate labor income.

By contrast, the household survey showed a loss of 38,000 jobs that together with the increase in population reduced the employment rate to 58.1%, even lower than the previous low of 58.2% reached in December 2009 (and November 2010 and June 2011). The drop in the unemployment rate was thus entirely the result of a continued decline in the labor force participation rate as more and more unemployed view the prospect of getting a job so low that it's not even worth trying to apply for jobs (only people who actively apply for jobs are officially counted as unemployed), even though they would still prefer to have a job.

The markets, as usual, only focus on the payroll survey and more specifically on the number of jobs reported there. That might seem sensible since the payroll survey is probably more reliable when it comes to monthly fluctuations than the often in the short term erratic household survey. However, over longer periods of time the household survey is more reliable and the fact that during the latest year, it has reported nearly a million less new jobs (household survey employment is up just 305,000 during the latest 12 months, compared to the 1.258 million increase in the payroll survey) suggests that the payroll survey likely overestimates the strength of the U.S. labor market.