Men and women fare differently in the labor market. There is, for instance, a large literature documenting earning differences between men and women who work the same job and have comparable education and experience. Similarly, it is well-known that progression and promotion in the workplace often seem more difficult for women than for men.
In this article, we discuss another facet of the difference between men and women in the labor force: their exposure to unemployment. We complement our analysis with a discussion of the black-white exposure to unemployment and show that it behaves noticeably differently than male-female exposure.
Figure 1 shows the difference between the unemployment rates of women and men since the late 1940s. We call this the “gender unemployment gap.” The shaded areas represent recessions dated by the National Bureau of Economic Research (NBER).
Figure 1
A few observations are worth noting. First, the gap tended to be positive before the 1980s; it was arguably large during the 1960s and 1970s, when the gap was between 1 and 2 percentage points. This means that for a long period of time, the unemployment rate of women was above that of men, i.e., women faced higher unemployment risk than men.
Second, throughout the 1980s and until the last recession, the gap was no longer as large as it had been. Instead, the gap seemed to hover just above or below zero, suggesting that women and men faced a similar risk of unemployment during this period.
Finally, the unemployment gap exhibits a tendency to decrease during recessions. This is particularly clear in the last recession. The unemployment rates of men and women were very close in the months leading up to the recession. In June 2007, for instance, the unemployment rate was 4.7 percent for men and 4.4 percent for women. But the unemployment rate of men rose to 11 percent in January 2010 versus 8.4 percent for women, causing a gap of almost –3 percentage points between them at the end of the recession.
The decline of the gender gap in the unemployment rate indicates that women appear to be less exposed to increased unemployment during recessions than men.
Changes in the Gender Gap during Recessions
Changes in unemployment, which are large during recessions, can have important welfare consequences. But, how do recessions affect the unemployment gap? We address this question in Figure 2. To build this figure, we considered the last seven recessions identified by the NBER.
Figure 2
First, we collected the data on gender gap in unemployment that start with the beginning of each recession and end 24 months later. Then, to allow for an easy comparison between each of the seven series of numbers, we “normalized” the gap to zero at the beginning of each recession. This explains why the lines in Figure 2 all start at zero.
An example can help. Take the case of the 1980 recession. When the recession started in January 1980, the unemployment gap between men and women was 1.1 percentage points, i.e., the women’s unemployment rate was 1.1 percentage points higher than that of men. In May 1980, which was the fourth month after the start of the recession, the gap was 0.1 percentage point. Thus, the gap decreased by 1 percentage point. Hence, the –1 can be seen in Figure 2 at the fourth month after the start of the 1980 recession.
There are three groups of recessions that stand out in Figure 2. Consider first the 1960 and 1969 recessions. The unemployment gap did not decrease significantly; it remained positive or near zero for one to two years after the start of the recession. This confirms the first observation made about Figure 1, that is, women faced higher unemployment risk than men.
The second group comprises the recessions from 1973 to 2001. These recessions show an approximately 0.5 to 1 percentage point decrease in the unemployment gap two years after the start of the recession.
Finally, the Great Recession—that is, the 2007 recession—stands out. Like recessions from 1973 to 2001, the Great Recession was followed by a reduction in the unemployment risk of women relative to men. But the magnitude of the reduction is dramatically different: Two years after the start of the recession, the gender unemployment gap was about 2 percentage points lower. In summary, these post-1970 recessions imply a lasting reduction in the unemployment risk of women relative to men, but the last recession stands out in the magnitude of this reduction.
The Racial Gap
A similar analysis can be conducted across race instead of across gender. Figure 3 is analogous to Figure 2, but the gap analyzed there is the difference between black and white. A positive gap means that the black unemployment rate is higher than the white unemployment rate.
Figure 3
The lesson from Figure 3 is remarkably different from that of Figure 2. First, the 2007 recession does not particularly stand out. Second, all the plotted recessions exhibit an increasing gap in the two years following the start of the recession. Black workers become relatively more exposed to unemployment than white workers after a recession.
Conclusion
We do not have a theory of the different patterns exhibited across recessions in Figure 2. Similarly, we do not have a theory of the difference between Figure 2 and Figure 3. We have documented the patterns of these gaps, but an explanation of these patterns is beyond the scope of our article. Yet the patterns raise important questions. Why are women relatively less exposed to the unemployment risk after recessions? Why are black workers relatively more exposed? Why does the Great Recession appear so different for the gender gap but not the race gap? Further research aimed at explaining these patterns would be of great interest.
- We do not consider the recession that started in July 1981 since it is subsumed in the two-year period after the start of the recession that began in January 1980.
- Data for the black unemployment rate are not available for the 1960s.
Women textile workers perform quality control tests in İzmir, Turkey. / © International Labour Organization
The COVID-19 pandemic has had a negative impact on both women's and men's employment – but at different stages of the crisis due to the gender segregation of economic activities in many countries.
UNCTAD analysis shows that early measures to curb the spread of the virus first hit jobs held predominantly by women, such as personal services. At the outset of the pandemic, a higher prevalence of the virus correlated with a higher rate of female unemployment (Figure 1).
But as the crisis worsened and disrupted cross-border value chains, the impact on men’s employment increased because they tend to work in sectors and jobs that are more dependent on international trade.
Even more worrisome, though, than how the pandemic has affected unemployment rates is its impact on women's participation in the labour market.
The available data reveals that even in countries where men’s unemployment rate outpaced that of women, more women left the labour market entirely in 2020. The significant drop in the number of women actively searching for a job threatens to reverse decades of progress on women’s empowerment.
Men’s unemployment increased as the crisis worsened
Overall, the volume of international trade in goods dropped by a dramatic 16% in the second quarter of 2020 compared with the same period of 2019.
In manufacturing, the global trade slump affected sectors that primarily employ men. In the United States, for example, exports of capital goods, industrial supplies and automotive industries were hit hard. Men comprise more than 70% of people employed in manufacturing, according to the US Bureau of Labour Statistics.
By August 2020, data started to show increases in men’s unemployment. Meanwhile, in countries where the spread of the virus had begun to slow, governments started easing confinement measures, allowing people to return to their jobs, especially women in the services sector.
But women’s labour force participation still fell more
Yet even as men’s unemployment overtook that of women in some countries, women’s labour force participation continued to decline more rapidly, adding an important nuance to the COVID-19 unemployment landscape.
Discouraged by the pandemic, many people who lost their jobs stopped actively looking for work. Since such people are excluded from unemployment statistics, it’s important to also analyse labour force participation.
Figure 3 shows the strong correlation that exists between COVID-19 prevalence and a decline in women’s labour force participation. The International Labour Organization also found that women’s working hours dropped dramatically in 2020.
Part of the explanation lies in women’s high share of tourism jobs. Globally, tourism suffered an estimated loss of $1.3 trillion in export revenues – more than 11 times higher than during the 2009 global economic crisis, according to the World Tourism Organization, which warned that up to 120 million tourism jobs were directly at risk.
A closer look at specific countries
In the US, 4.4% of both women and men were unemployed in March 2020. But in just one month, unemployment jumped to 16.1% for women and 13.6% for men. The gender difference disappeared gradually and both rates fell to 6.7% in December 2020. But during the year, women’s labour force participation dropped by 3.4% compared with 2.8% for men.
Chile provides another good example of how COVID-19 has affected men’s and women’s unemployment and labour participation differently.
In the South American nation, men’s unemployment increased faster than women’s in spring and summer 2020 – a 90% drop for men compared to less than 60% for women. Yet in August 2020, women’s decline in labour force participation was worse, falling 16.6% year-on-year compared with 10% for men.
Statistics from Italy show that women’s employment was hit harder at the start and throughout the pandemic. According to the national statistics institute (ISTAT), 326,000 women had lost their jobs in 2020, compared with 141,000 men. Overall, women’s labour force participation in the European nation was notably lower than that of men before the crisis (56.2% compared with 74.6%) and decreased even more during it (by 3.7% for women and 1.2% for men).
From the viewpoint of women’s empowerment, it’s particularly alarming that in these and other economies more women than men are leaving the workforce.
More gender-disaggregated data needed
COVID-19’s impacts on unemployment are not gender neutral, and recovery efforts should include targeted action to bring more women back into the workforce.
Effective response measures and policies require comparable and granular statistics to tell the full story of the pandemic’s socioeconomic impacts on women and men. Governments must therefore provide national statistical offices with the resources they need to produce more gender-disaggregated data that can be used to assess the economic aspects of gender equality.
UNCTAD is working closely with them and international organizations to develop better statistics and policies to enable the design of recovery responses that improve gender equality in the future economy and job markets.
Simonetta Zarrilli is the chief of UNCTAD’s trade, gender and development programme, and Henri Luomaranta is an UNCTAD statistician.
Sources for figures 1-4: The University of Oxford, Blavatnik School of Government (2020), United Nations Department of Economic and Social Affairs (2019), and ILO (2021).