Nonfarm payroll jobs rose sharply last month, by 288,000, following an upward revised 337,000 leap in March. It was the eighth monthly gain in a row and was widespread among industries. Notably, manufacturing jobs improved for the third successive month, reversing a downward trend that began in August 2001.
So far this year payroll jobs have risen by an average of 217,000 a month. If this pace continues, as many economists expect, by year’s end the job total will have exceeded its pre-recession peak in March 2001.
Though the unemployment rate ticked down to 5.6 percent in April and was well below its cyclical high of 6.3 percent last June, it was the same as in January despite the strong showing in payroll jobs. But it’s not payroll jobs, collected by business survey, that swing the unemployment rate. Rather it’s civilian employment, collected from the same household survey as unemployment. Household measured employment has been stagnant since the beginning of the year, contributing to the lack of further improvement in the jobless rate. Most economists in and out of government agree that the larger payroll survey produces the more accurate employment data. In the past, employment differences between the two surveys have converged over time, and it may well be that household measured employment will catch up with the rise in payroll jobs in the months ahead.
The outlook going into the presidential election is for further gains in output and employment and a resumption of the downward trend in unemployment. Regarding unemployment, however, an alternative possible outcome must be considered, namely that the jobless rate may rise for a time despite continued robust growth in payroll jobs, and even if civilian employment turns upward.
Most of the time during economic expansions, when employment goes up, unemployment declines. However, over the past half-century, there have been brief periods when both employment and unemployment have risen together. This happens when people outside the labor force, who are desirous of a job, see that the labor market is improving and are encouraged to enter the work force in the expectation of finding employment.
When the labor force influx is greater than the increase in jobs, unemployment rises. This phenomenon of many jobseekers chasing the available jobs is usually temporary. As employment continues to improve, more of the labor force entrants find jobs, the ranks of the labor reserve thin, and unemployment resumes its more normal inverse relationship to changes in employment.
Looking at the historical record, in 1954 the unemployment rate rose from 5.6 percent to 6.1 percent between June and September, while in the same period civilian employment rose by nearly a half million. In 1956 the jobless rate rose from 3.9 percent in February to 4.4 percent in July, during which time civilian employment rose by nearly 300,000. In 1979 the unemployment rate rose from 5.6 percent in May to 6.0 percent in August, as civilian employment expanded by nearly 450,000 and payroll jobs rose by over a half million. The unemployment rate rose from 7.3 percent in December 1991 to 7.8 percent in June 1992 as civilian employment grew by nearly 400,000 and payroll jobs by more than twice that.
The duration of these episodes was from three to six months. In general, the greater the labor reserve build-up outside the work force that has to be absorbed into employment, the longer the episode will last. Conversely, the stronger the rise in jobs, the fewer the months before unemployment reverses direction and begins to decline in response to further increases in employment.
Because of an extended period of a weak job market in the current economic recovery, until recently, there has been a fall-off in labor force participation and an unusually large build-up of people waiting on the sidelines for signs of a sustained improvement in job opportunities. The labor force participation rate declined from 67.1 percent at the cyclical peak in March 2001 to 65.9 percent last month.
If, as economists expect, the job growth of recent months continues in the months and quarters ahead, it is likely that both the existing jobless and labor market entrants will be absorbed into employment, and unemployment will eventually decline. In the shorter run, however, there could well be an influx of hopeful would-be workers into the job market, in sufficient numbers to drive up the unemployment rate. The size of the labor reserve in such that a surge of labor force entrants could push up the unemployment rate for several months, as it did in the first half of 1992, even with a healthy expansion in employment, however measured.
The last unemployment report before the presidential election will be in early October, showing data for September. Rising unemployment going into the election would be a political nightmare for Republicans. On the other hand, with continued increases in employment, the jobless would have reason to be encouraged, and the positive political impact of more people getting jobs could more than offset the negative perception of a rise in unemployment. A temporary increase in the measured unemployment rate that is a transition to a high-employment economy is not bad news. But there is little doubt that Republicans would prefer the alternative scenario of rising employment accompanied by a declining unemployment rate.
Alfred Tella is former Georgetown University research professor of economics.