ByHoward J. Wall
Since the U.S. economy entered into its current recession in December 2007, steep job losses have been seen for most demographic groups and industries. By any measure, news from the labor market has been dire: Between the fourth quarter of 2007 and the third quarter of 2009, nonfarm employment fell by about 6.8 million jobs while the unemployment rate rose from 4.8 percent to 9.6 percent. The overall picture has been bleak, but the bad news has not been distributed evenly across demographic groups.
A recent report produced by the Federal Reserve Bank of St. Louis analyzed the effects of the recession on the employment of a variety of demographic categories—sex, marital status, race, age and educational attainment—relative to the previous five recessions. To provide a more complete picture of what happens to employment during recessions, the report accounted for two important considerations that are usually overlooked. First, it considered employment losses that occurred outside of official recessions because changes in employment do not always coincide with official recession periods (i.e., jobless recoveries). Second, the report also used estimates of the growth in employment that would have occurred if the recession had not happened—foregone employment—to more accurately measure the total effect of recessions on the level of employment.
This article follows the methodology of the report and applies it to the 2008-2009 recession, focusing on the different effects of the recession by sex and race.
When the word recession is used to describe specific periods of economic weakness, it refers most often to the official recession dates determined by the business-cycle-dating committee of the National Bureau of Economic Research (NBER). It used to be that NBER recession periods coincided with periods of falling employment, but this link collapsed starting with the 1990-91 recession. For the current recession, employment growth first dipped below zero in early 2007, months before the start of the official recession, and has continued to fall even after the second quarter of 2009, the date that most analysts presume will be considered the last quarter of the recession. Therefore, the effects of the recession should be measured as starting with the second quarter of 2007 through the most recent data available.
Typically, the effects of a recession on employment are seen as simply the difference between the levels of employment at the start and end of a recessionary period. This assumes, though, that there would have been zero employment growth if the recession had not occurred. However, the recession results not only in a drop in employment from its pre-recession level, it also prevents employment growth that would have occurred. This “foregone” employment needs to be accounted for in an analysis of the recession’s total effects on employment. This consideration is especially crucial for present purposes because average employment growth varies a great deal across demographic categories.
The figure illustrates what has happened to employment because of the recession. For all categories combined, the level of employment has fallen by 4.6 percent. As already described, however, this does not tell the whole story. What is needed is a notion of what the level of employment would have been if there had not been a recession. If the estimate of the missing employment growth is taken to be simply the median growth between 1985 and 2009, the foregone employment is 3.3 percent, making the total effect of the recession on employment a 7.9 percent decrease.
The difference across demographic groups that has received the most attention is that between men and women. In fact, some have labeled the recession the “Great Man-Cession” because employment losses have fallen disproportionately on men. As shown in the figure, employment for men fell by 6.4 percent whereas that of women fell by only 2.6 percent. Put another way, the fall in men’s employment is about 2.5 times that of women’s.
It is not at all unusual for men to bear a greater burden of direct employment losses during a recession. In fact, this recession is one of the milder ones in this regard. Further, because women’s employment has tended to grow faster than men’s employment, the difference between men and women is reduced dramatically once foregone employment is taken into account. Specifically, because foregone employment for women was about 60 percent higher than for men (4.1 percent versus 2.5 percent), the total effect of the recession on employment was about 33 percent larger for men than for women (-8.9 percent versus -6.7 percent). So, although the burden of the recession has fallen disproportionately on men, the discrepancy between men and women is not nearly as large as it appears from looking at the simple employment changes alone.
There are interesting differences in the effects of the recession if the employment data are broken down by race. In fact, the differences by race are intertwined with the differences by sex. As reported in the figure, there is a substantial gap in the changes in employment between whites and blacks: White employment is 4.4 percent below its pre-recession level, whereas black employment is 7.5 percent below its pre-recession level. Because black employment has tended to grow faster than white employment, the gap widens after foregone employment is included: The total effect of the recession on black employment is about 50 percent higher than the total effect on white employment (-11.3 percent versus -7.4 percent).
The final section of the figure shows how the breakdowns by race and sex are intertwined. First, whereas the effect of the recession on black men’s employment is about 50 percent of its effect on white men’s employment, the effect on black women is twice the effect on white women. Further, for men, it is the change in employment that is dominant, although there are significant differences by race: For white men, the change in employment is about 2.5 times foregone employment, while for black men the employment change is about 4 times foregone employment. For women, on the other hand, it is foregone employment rather than the employment change that is the larger of the two effects, although both effects are much larger for black women than for white women. Finally, note that the total effect of the recession on black women’s employment is double that for white women, and the difference between the total effects on black women and black men is relatively small.
So, what accounts for the differences in the effects of the recession across these categories? The most obvious explanation for the differences between men and women is that they are a reflection of the recession’s effect on the industries in which they tend to be employed: Construction and manufacturing, which are male-dominated, have seen the biggest declines in employment, whereas education and health services, where women are a majority, have actually seen job growth during the recession. No doubt these industry differences play a role, but they cannot go very far in explaining the differences outlined above. To do that, one should begin with education because there is a strong link between the demographic categories that have been hit hardest and those with low average educational attainment. For example, women are more likely to have finished high school and whites are much more likely to have a bachelor’s degree than are blacks.