The Labor Market for Men and Women
Over the course of this recession, men have experienced significantly higher unemployment rates than women. The unemployment rate for men rose by 6.7 percentage points from its 2007 average level, peaking at 11.4 percent , while the unemployment rate for women increased by 4.3 percentage points, peaking at 8.8 percent. This pattern of more cyclically sensitive unemployment rates for men has been apparent over the last four recessions.
The higher unemployment rates for men compared to women during this cycle are seen in all age groups. For both men and women, unemployment rates generally decline with age. While the elevated unemployment rates of younger men relative to younger women account for some of the overall difference in rates between the two groups, the difference explained is small. This is because the share of younger workers in the labor force is relatively small and the share of women in the young age categories is actually slightly larger than men.
The key to explaining the difference in unemployment rates between men and women appears to be the fact that men and women are employed in different proportions in different industry sectors. Some sectors dominated by men have experienced above-average job losses and higher unemployment, while the reverse is true in industries with a relatively large share of female employees. This is especially true in those industries at the extremes of the male-female distribution. Construction, the most male-intensive industry, has experienced the highest increases in unemployment rates and the weakest employment growth, whereas the education and health service sector, which has by far the highest share of women, has actually added employment over the recession.
However, if one looks within industries, employment growth for men and women is quite similar. On average, changes in employment growth rates within individual industries did not differ systematically between men and women, as indicated by the fact that the growth rates for men and women within industries fall relatively symmetrically along the 45 degree line of the chart below. If for example, women’s employment growth was higher, on average, across industries, one would expect the data points in the scatter plot to lie generally above the 45 degree line. The key point is that within industries, men and women employees experienced very similar employment loss.
The pattern for changes in the unemployment is not quite the same. The change in men’s industry-level unemployment rate is, on average, somewhat greater than women’s, so the points tend to lie somewhat below the 45 degree line (i.e. less symmetry).
What might explain this difference between the patterns of employment growth and changes in unemployment rates for men and women? One possibility is differences in the duration of unemployment for men and women working in the same industry. While employment growth measures the net job change in a sector, the unemployment rate is a more complicated metric, as it incorporates both the incidence and duration of job loss. Hence changes in unemployment rates can be due to both forces and thus industry patterns of job loss and changes in unemployment rates need not correspond precisely. To be sure, at the aggregate level, men stay unemployed longer than women. In the second quarter of 2010, for example, women remained unemployed 4.2 fewer weeks than men.