The new data on retail sales should give us grounds for concern. The issue here is the Bureau of Labor Statistics' imputation of jobs for new firms that are not included in the survey. This imputation was was very large in April, 317,000 to be exact. Since the establishment data showed a net gain of 88,000 jobs, this means that survey actually showed a loss of 229,000 jobs. In fairness, most of this imputation is due to seasonal factors. Many new businesses open or expand in the spring. But, the imputation was 46,000 larger than in April of 2006 when the economy seemed to be growing considerably more rapidly by most measures. One area that seems especially suspicious is restaurant employment. There was an imputation of 95,000 new jobs for the larger leisure and hospitality sector, which includes restaurants. The restaurant sector itself showed a gain of 25,000 jobs in April. Since the retail sales data from the Commerce Department show a nominal decline in sales of 0.1 percent from March to April, it seems implausible that the sector added 25,000 workers. Over the last year, restaurant sales are up by 4.9 percent in nominal terms, which translates into 1.6 percent real growth. (Inflation for food away from home is reported at 3.3 percent.) This seems hard to reconcile with the 336,400 jobs reportedly added to the sector, a gain of 3.6 percent. The moral of the story is that it seems likely that the economy has hit a turning point, which causes the new jobs imputation to overstate job growth. Actual job growth is likely somewhat slower than reported growth. Since reported growth is already quite weak, actual job growth may now be close to zero, at least in the private sector. Of course, the good news is that productivity growth is somewhat higher than is being reported.
--Dean Baker
