Share
The narrowing of the gender pay gap slowed around the turn of the 21st century. Meanwhile, younger workers have fared progressively worse than older workers. This column uses data for the US, Italy, Canada, and the UK to show that while pay convergence was initially driven by a narrowing of the gender gap at labour market entry, in the last two decades it has been driven solely by the retirement of older cohorts with larger earnings differentials. Even more disappointingly, the convergence of entry outcomes that persisted until the mid-1990s was not due to improved prospects for younger women, but rather to disproportionately poorer outcomes for younger men.
Between the mid-1970s and the early 2000s, the gender gap in wages and unemployment decreased in most high-income economies, leading to a significant reduction in gender inequality (Olivetti and Petrongolo 2016, Blau and Kahn 2017, Sahin and Albanesi 2018). This decrease slowed down around the turn of the 21st century, a fact that has been the focus of many recent policy debates (England et al. 2020).
During the same period younger workers have fared progressively worse than older workers, experiencing a widening age pay gap (Rosolia and Torrini 2007, Bianchi and Paradisi 2023, Bianchi and Paradisi 2024), a lower likelihood of being promoted and reaching higher-paying jobs (Bianchi et al. 2023), and declining lifetime earnings (Guvenen et al. 2022).
In a recent paper (Arellano-Bover et al. 2024), we link these two macro trends together, and document why the deterioration in the careers of younger workers is related to the narrowing of the gender pay gap and its subsequent slowdown.
Workforce ageing can contribute to shrinking the gender pay gap
The main mechanism that we highlight to explain why workforce ageing contributes to the narrowing of the gender pay gap centres around the unavailability of slots at the top of firms’ hierarchies. Many firms face constraints in adding higher-level positions to their organizations, due to slow growth and a consequent lack of available responsibilities and tasks to assign to new top jobs. Under these conditions, having more older workers in higher-paying jobs (e.g. because of population ageing or higher retirement age) crowds out younger workers as firms’ hierarchies do not expand sufficiently to accommodate the increased presence of older workers at the top (Bianchi et al. 2023, Bianchi and Paradisi 2023, 2024).
The crowding out of younger workers from top positions has implications for the gender pay gap. Indeed, the increased supply of older workers has narrowed the gender wage gap by affecting the labour market outcomes of younger women less than those of younger men, who were more likely to hold higher-paying jobs at baseline.
We develop a theoretical framework that incorporates these gender-specific cross-cohort spillovers and then test its main predictions using a combination of survey and administrative data for the US, Italy, Canada, and the UK.
The gender pay gap has narrowed across worker cohorts
The data show that the gender pay gap has been declining in all countries in our sample for at least the last four decades (red triangles in Figure 1 for the US and Italy). Crucially, we show that this narrowing has been driven by reductions across subsequent worker cohorts, rather than over the life cycle of any given cohort (shaded dots in Figure 1). Moreover, the recent slowdown in the closing of the overall gender pay gap in the economy, which began in the mid-1990s, has coincided with a slowdown in the closing of the gap between cohorts.
Figure 1 Gender pay gap between and within cohorts
Having established the importance of cross-cohort dynamics, we directly address the core prediction that workforce ageing has narrowed the gender pay gap across cohorts by worsening the outcomes of younger men more than those of younger women. Empirically, we examine where men and women at age 25 ranked in the overall pay distribution over time. The narrowing of the gender pay gap at labour market entry, which lasted until the mid-1990s, was indeed driven by younger men falling closer to younger women in the pay distribution rather than by younger women experiencing disproportionate improvements in labour market outcomes (Figure 2). In the US, the average rank of younger men at age 25 fell from the 50th percentile of the wage distribution in 1976 to the 39th percentile in 1995, while the mean position of women at age 25 remained stable around the 30th percentile.
Figure 2 Positions in pay distribution at 25 years old
Younger men have become more likely to work for lower-paying firms
We then examine changes in the distribution of younger men and younger women across different types of firms. Bianchi and Paradisi (2023) documented that older workers have become increasingly concentrated in higher-paying firms. Our paper further shows that, compared to younger women, younger men experienced larger positional losses within the pay distributions of all types of firms. These losses were largest in higher-paying firms, which younger men left at higher rates. Thus, by the end of the sample period, younger men joined younger women in being less represented among higher-paying firms (Figure 3).
Figure 3 Distribution between firms for workers age 25: Italy
The gender pay gap is not projected to close
Finally, we show that the narrowing of the gender pay gap over the last 20 years of data is almost entirely due to the retirement of older worker cohorts who, on average, had larger gender pay differentials. Therefore, in contrast to forecasts based on trends in the aggregate gender pay gap, which predict that the mean earnings of men and women will equalize in a few decades, we project that the gender pay gap will not disappear in the high-income countries in our study. At best, and in the absence of structural breaks in the labour market, the gender pay gap will converge to the level observed among recent labour market entrants, a differential that is still economically significant.
Figure 4 Projected convergence in the gender pay gap
To better understand why the gender pay gap at entry has stopped closing, we show that a substantial part of the remaining entry gap results from different choices of college majors among young men and young women (consistent with Sloane et al. 2021 and Carta et al. 2023). In the US, for example, the major-predicted entry pay gap has remained stable over the past three decades, accounting for about 63% of the total gender pay gap at entry for college graduates in more recent years. Consistent with these findings, our stylised framework suggests that a larger number of older workers cannot further reduce the gender pay gap if most of the remaining earnings differential between younger men and younger women depends on education choices made before entering the labour market.
In conclusion, our findings imply that pay convergence has evolved through the progressive entry and exit of worker cohorts. It was initially driven by a narrowing of the gender gap at labour market entry. In the last two decades, however, it has been driven solely by the retirement of older cohorts with larger earnings differentials. Even more disappointingly, the convergence of male and female entry outcomes that persisted until the mid-1990s was not due to improved prospects for younger women, but rather to disproportionately poorer outcomes for younger men.
Source: Voxeu
