There is a lively debate as to why the gender wage gap has closed in industrialized countries in recent decades.1 When investigating possible explanations, most research has focused on factors such as education and experience, for which changes have been more favorable for women than for men; increased labor force participation and the changing characteristics of working women rank high on the list of likely causes.2 All of these explanations come from the supply side, implicitly assuming that there were no changes in the skill requirements of women’s work relative to that of men. In our recent research, we apply a task-based framework in order to evaluate the importance of technological change--one of the major sources of changes in labor demand--in explaining the closing of the gender wage gap.3
Evolution in the skill-mix in the labour market
Since the late 1970s, the diffusion of computers had a large impact on the skill requirements of jobs. Some tasks, such as bookkeeping, calculating or operating machines (so-called cognitive and manual routine tasks) have been taken over by computers. At the same time, computers have increased the productivity of employees who perform tasks such as researching, analyzing, selling or buying things (so-called analytical and interactive non-routine tasks). As a consequence, the structure of tasks performed by employees has changed.
Using a unique, survey-based data set from West Germany that covers the time period 1979 to 1999, we are able to measure skill requirements directly by using the task composition of occupations. Figure 1 illustrates the percentage point changes in tasks relative to 1979 for West Germany. From this figure, one can see that the task content of work has shifted away from cognitive and manual routine tasks towards analytical and interactive non-routine tasks.4 These task changes are economically important, as they have an impact on the type of labour demanded in the market. High skilled employees have a comparative advantage in performing the tasks that are now in greater need at work, a conjecture that is consistent with the empirical evidence that suggests that the task changes explain a considerable fraction of the education upgrading we observe.
Figure 1: Trends in Aggregate Skill Inputs
Task changes and the closing of the gender wage gap
Our recent research documents that these task changes are also important in explaining the closing of the gender wage gap. It turns out that the aggregate changes we observe in Figure 1 actually mask important differences in the evolution of skill requirements at work by gender.
In the late 1970s, men and women were differentially distributed across tasks. Specifically, women were more like to be performing tasks for which computers are substitutes. As a result, the changes in workplace tasks have been larger for women than for men. Figure 2 illustrates the evolution of task inputs of women relative to men between 1979 and 1999 by showing the proportional difference in task changes relative to 1979; that is, growth in female task inputs minus growth in male task inputs. It is striking that all the changes in task inputs have been larger for women than for men. An inspection of the underlying data shows that, in the earliest period, men’s analytical task inputs were more than twice as high as those for women, while women had higher routine cognitive and routine manual task inputs. However, by 1999, women appear to be catching up to men in terms of analytic skills and, even more, in terms of interactive skills. For routine cognitive and routine manual skills, where women had dominated 20 years earlier, men have taken over.
Figure 2: Female Relative Task Inputs
We relate the relative task changes for women to the closing of the gender wage gap. To do so, we decompose the rate of wage convergence between the genders into changes in the tasks performed by men and women and changes in relative task prices. The results show that relative task changes explain a substantial fraction of the closing of the gender wage gap.
This research shows that the one-sided focus on supply side factors of the literature so far only conveys part of the picture. In particular, it conveys too positive a picture of the counterfactual; that is, how the gender wage gap would have evolved had women not increased their educational attainment and experience. Given the changes in task inputs observed in the labour market in recent decades, it is very likely that the labour market position of women would have deteriorated had they not upgraded their skills.
Policies to promote employment of women need to take these changing skill requirements into account. Affirmative action policies intended to increase female employment, for example, must go hand in hand with educational programs that ensure that women can meet the skills now demanded at workplaces.
Autor, D.H., F. Levy, and R.J. Murnane (2003), “The Skill Content of Recent Technological Change: An Empirical Exploration,” Quarterly Journal of Economics, Vol. 118(4), 1279-1333.
Blau, F.D., and L.M. Kahn (1997), “Swimming Upstream: Trends in the Gender Wage Differential in the 1980s,” Journal of Labor Economics, Vol. 15(1), 1-42.
Blau, F.D., and L.M. Kahn (2003), “Understanding International Differences in the Gender Pay Gap,” Journal of Labor Economics, 21, No. 1, 106-144.
Blau, F.D., and L.M. Kahn (2004), “The U.S. Gender Pay Gap in the 1990s: Slowing Convergence,” NBER Working Paper No. 10853.
Spitz-Oener, A. (2007), “Technical Change, Job Tasks, and Rising Educational Demands: Looking Outside the Wage Structure”, Journal of Labour Economics, 24(2), 235-270.
1 See various Vox columns on the issue (Manuel F. Bagues and Berta Esteve-Volart, Stefania Albanesi and Claudia Olivetti, Doris Weichselbaumer, Rudolf Winter-Ebmer and Martina Zweimüller, M Daniele Paserman, or Juan Dolado)
2 For a comprehensive review, see work by Blau and Kahn (1997, 2003, 2004).
3 “Explaining Women’s Success: Technological Change and the Skill Content of Women’s Work” by the same authors available as NBER Discussion Paper No. 13116 at www.nber.org/papers/w13116.
4 The source of these figures is Spitz-Oener (2007). For a similar figure for the United States see Autor, Levy and Murnane (2003).