ByHoward J. Wall
Between the fourth quarter of 2007, when the current recession began, and the first quarter of 2009, men bore 78 percent of the job losses. Over the same period, the unemployment rate for men rose from 4.9 percent to 8.9 percent, while the rate for women rose by only half as much, from 4.7 percent to 7.2 percent. As reported by economist Mark Perry of the University of Michigan-Flint in his blog Carpe Diem, this gap in unemployment rates has no precedent during the post-war period. In light of the disproportionate employment effects of the recession on men, some commentators in the press and elsewhere have labeled the current recession a "man-cession" or even the "Great Man-Cession."
The dominant explanation for this phenomenon is that it follows from the severity of the recession across industries. According to Christina Hoff Sommers of the American Enterprise Institute, "Men are bearing the brunt of the current economic crisis because they predominate in manufacturing and construction, the hardest-hit sectors." Women, on the other hand, "are a majority in recession-resistant fields such as education and health care." Harvard economist Greg Mankiw echoes this in his blog, conjecturing "that a large part of the explanation is the sectoral mix of this particular downturn in economic activity, including a significant slump in residential construction."
Despite the sudden interest in the phenomenon, the relative effects of the recession on men and women are not the least bit unusual. At least since the 1969 recession, men have borne the brunt of job losses during recessions, and, compared with previous recessions, men have actually borne a smaller proportion of job losses in the current recession. Between 1969 and 1991, male employment fell by an average of 3.1 percent during the five recessions experienced during the period. Female employment, on the other hand, actually tended to rise by an average of 0.3 percent during recessions.1 Women have a much larger presence in the work force now than between 1969 and 1991; so, a more-relevant comparison is to the 2001 recession.
For that recession, employment peaked in the first quarter of 2001 and bottomed out in the third quarter of 2003, with a total loss of a little more than 2.6 million jobs. Men accounted for 78 percent of those job losses, just as they have during the current recession. So, in terms of job losses, the current recession has hit men in roughly the same proportion as did the previous recession, but by a much smaller proportion than during earlier recessions.
Still, according to unemployment rates, the gap between men and women is higher than it has ever been. It is a bit of a mystery as to why the gap in unemployment rates shows much more of a man-cession than is indicated by jobs numbers, but unemployment rates indicate much more than simply changes in employment status.
The rates reflect not only the net number of people who lose their jobs, but also the net number of people who are in the labor force either already employed or looking for a job. During this recession, the male labor force has been shrinking as the number of unemployed men has been rising. The female labor force, in contrast, is actually larger than it was when the recession began, accounting for much of the increase in the gap between the male and female unemployment rates.
In sum, the proper perspective on the current recession is that its effect on the employment of men relative to women is very similar to the effects of the 2001 recession and much milder compared with earlier downturns. Although this perspective debunks the notion of this recession being an especially bad one for men relative to women, the fact remains that recessions hit male employment much harder than female employment. Total employment has fallen by 3.1 percent between the fourth quarter of 2007 and the first quarter of 2009, while male and female employment fell by 4.8 percent and 1.4 percent, respectively. Put another way, men lost jobs at 3.4 times the rate at which women did. Despite what has been presumed, however, for the current recession, this is not necessarily due to the different mixes of industries in which men and women tend to be employed.
It's easy to see the reasons for supposing that the disproportionate job losses for men are due to the disparate impacts of the recession on the goods-producing sector, in which 77 percent of employees in the fourth quarter of 2007 were men. The two hardest-hit industries have been construction and manufacturing, which lost 12.7 percent and 9 percent of their jobs, respectively, between the fourth quarter of 2007 and the first quarter of 2009. These two industries also happened to have had two of the three highest shares of male employment. At the other end of the spectrum, two of the three industries that saw positive job growth over the period—the government sector as well as the education and health services sector—are among the three with the lowest shares of male employees. As illustrated by Figure 1, there is a strong negative relationship between the share of male employment and the rate of job growth. A notable exception to this tendency is the relatively small natural resources and mining sector, which has seen strong job growth in the wake of high energy prices.
The problem with explaining the man-cession only in terms of industry mix is that it's not really possible to separate the industry-mix effects from other effects. The evidence that something else is going on is that men have been hit disproportionately in almost every industry; that is, within an industry, men have tended to lose jobs at a higher rate than have women. In the service sector, in which men accounted initially for only 46 percent of employment, men lost jobs at 4.2 times the rate that women did (3.1 percent versus 0.7 percent), resulting in the same 78/22 split for the economy as a whole. If the 78/22 split of total job losses were due to male-majority industries being hit hardest, we wouldn't see the same split in the goods-producing and service-producing sectors. Looking deeper into the industry-level numbers, we can see more evidence that the man-cession is more than an industry-mix story.
The man-cession in the service sector is laid out in more detail in the table. In the trade, transportation and utilities industry, men began the period holding 59 percent of the jobs. In percentage terms, their job losses were 1.4 times that of women, meaning that they accounted for 67 percent of the industry's total losses. Similarly, in professional and business services, leisure and hospitality, and "other" services, men lost jobs at 1.3, 1.2, and 5.8 times the rate that women did and, consequently, accounted for a disproportionate share of job losses.
In the two industries that gained jobs, women began the period accounting for large majorities of employment and gained disproportionate numbers of new jobs. The education and health industry, which began the period with 77 percent women employees, experienced job growth of 3.3 percent, 80 percent of which went to women. Women accounted for 57 percent of employees in government, which saw a 1 percent increase in employment, all of which was for women.
There were two industries that bucked the trend and saw job losses that fell disproportionately on women. Whereas men comprised 58 percent of initial employment in the information service industry, they accounted for only 49 percent of the job losses. This industry is relatively small, however, making up only about 2 percent of total employment. In the financial services industry, the job losses fell almost proportionally, with women seeing 61 percent of the job losses while starting the recession with 59 percent of the jobs.
Because men tended to have been affected disproportionately across all industries, whether goods-producing or service-producing, the story behind the man-cession cannot be about industry mix alone. Clearly, then, the man-cession phenomenon is not a story about the goods-producing industries but reflects something much broader about the economy and how firms respond to downturns by deciding which workers they will let go and which they will hire. As we have seen, employment losses are not felt the same by men and women within the same industry, and, in fact, recessions have widely varying effects across demographic groups. Perhaps the male/female differences within these categories can shed some light on the man-cession phenomenon.
Up to this point, all of the data have come from payroll employment series produced by the Bureau of Labor Statistics (BLS) and which are derived from a monthly survey of 150,000 or so employers around the country. These data, however, are not broken down by demographic categories other than sex; so, a different data source is needed. Fortunately, the bureau also surveys households on a monthly basis and categorizes the responses by demographic categories. The employment measures from the payroll and household surveys are not the same in that they cover different types of employment. For example, payroll employment does not include farm employment or self-employment. Although the two employment measures do not coincide perfectly, they do capture the same broad patterns in male/female employment. In fact, by fortunate coincidence, the household survey indicates the same 78/22 split in the male/female employment losses that arise from the payroll employment data and each of its two major components, the goods-producing and service-producing sectors.
Figure 2 illustrates the differences across demographic groups and between men and women within each group. For every demographic group except for those aged 55 and above, fewer were employed in the first quarter of 2009 than in the fourth quarter of 2007, and men fared worse than women within every group. There were, however, significant differences in the impact of the recession across the groups and on men relative to women. Note that the demographic groups overlap a great deal; so, the explanations for the differences across them also often overlap. Further, across groups, employment changes over the period reflect not only the effects of the recession but also ongoing trends in the tendency to participate in the labor market.2
Married men and women saw smaller job losses than did their single counterparts. Moreover, the effect of the recession on the employment of married men was almost nine times that on married women, whereas the effect for single men was 2.4 times that for single women. In part, the fact that married women are the least likely subgroup to see employment losses can be explained by what has been called the "added-worker effect."3
According to this effect, some married women enter the labor force during recessions following their husbands' job losses. The added-worker effect can account for some of the increase in the female labor force during the recession.
Another explanation for the difference between married and single people is that married people are more likely to have children and are, therefore, more likely to take a new job at lower pay after they lose their old job. Also, much of the differences according to marital status are reflections of other demographic differences that make them more likely to be affected by a recession: Compared with married people, single people tend to be younger and, therefore, have less work experience and lower education levels.
The differences across racial categories are intertwined with differences in other categories. Black men, who have less education on average than black women or whites, saw the largest decrease in employment. Black women, on the other hand, have seen the smallest reduction in employment of any of the six sex-race categories. Underlying these differences is the long-term trend of women, especially black women, becoming more likely to be employed.
Figure 2 also illustrates the changes in employment across age groups, for which there are significant differences across groups and between sexes within each group. Teenagers, for example, have seen the biggest decrease in employment during the recession, but there was little difference between the percentage decreases for male and female teenagers. In contrast, for the next lowest age group, those aged 20 to 24 years, men saw about a 9 percent decrease in employment, which was 5.5 times the decrease for women. Very large differences between men and women were also seen for ages 25-34 and 45-54. Partly reflecting the ongoing trend of increasing employment, the number of employed people aged 55 and above rose by more than 3 percent during the period. This increase might also be due to the effects of delayed retirements in the wake of dramatic decreases in savings and investments for retirement.
The final demographic category is educational attainment, for which there were dramatic differences in male and female employment changes during the recession. For every category, men fared worse than women (Figure 3). Much of these differences reflect the industry-mix effects: Men without a high school diploma, for example, would make up a significant proportion of men in the construction and manufacturing industries, whereas women with associate and bachelor's degrees would make up a large portion of the education and health industries. Nevertheless, given the differences in education levels between the sexes within other demographic categories, such as race and age, education level is probably an important part of the man-cession story.
The first thing to take away from this blizzard of data is that the so-called Great Man-cession of 2008-09 is nothing unusual when compared with the previous recession. Even so, a greater than three-to-one employment impact on men relative to women is still large relative to the nearly equal representation of the sexes in the work force. This certainly has something to do with the differences in the industries for which men and women are in the majority, as there is a strong tendency for industries with large shares of men to have been hit hardest by the downturn. These differences, however, are only part of the story, which must be completed by examining the sometimes large differences in the educational and demographic characteristics of men and women. The differences in employment changes between men and women within these groups are usually larger than those across industries.
|Share of Industry Employment Q4.2007||Share of Industry Change||% Change Q4.2007 to Q1.2009||Men Relative to Women|
|Trade, Trans. and Utilities||–4.1|
|Professional and Business||–5.6|
|Education and Health||3.3|
|Leisure and Hospitality||–2.2|
DeRiviere, Linda. "Have We Come a Long Way? Using the Survey of Labour and Income Dynamics to Revisit the 'Pin Money' Theory." Journal of Socio-Economics, Vol. 37, No. 6, December 2008, pp. 2340-67.
DiCecio, Riccardo; Engemann, Kristie M.; Owyang, Michael T.; and Wheeler, Christopher H. "Changing Trends in the Labor Force: A Survey." Federal Reserve Bank of St. Louis Review, Vol. 90, No. 1, January/February 2008, pp. 47-62.
Goodman, William; Antczak, Stephen; and Freeman, Laura. "Women and Jobs in Recessions: 1969-92." Monthly Labor Review, Vol. 116, No. 7, July 1993, pp. 26-35.
Hoff Sommers, Christina. "No Country for Burly Men." The Weekly Standard, Vol. 14, Issue 39, June 29-July 6, 2009, pp. 22-24.
Stephens Jr., Melvin. "Worker Displacement and the Added Worker Effect." Journal of Labor Economics, Vol. 20, No. 3, July 2002, pp. 504-37.