The gender pay gap remains stubbornly large. At the same time, city workers earn more than rural workers, a phenomenon known as the urban wage premium. This column shows that up to 2008, the urban wage premium for women in the UK was more than twice that for men, but it fell drastically during the global financial crisis while the premium for men remained the same. The global recession coincided with austerity policies that reduced provisions for childcare and public transport, policies that were particularly detrimental to women in cities.
The residual gender pay gap – the part of the gender pay gap not explained by differences in qualifications between men and women – has remained stubbornly high over time (Brynin 2017). To reduce the gap, we need to identify reasons other than qualifications for why women earn less than men. How women’s personal situations affect their earnings has been considered (Dolado et al. 2010). The gender pay gap really begins once workers have children, and widens over their working life (Andrew et al. 2021). Perhaps because of gender norms, women tend to work in jobs that are geographically close to their homes rather than searching for the best-paying job (Meekes and Hassink 2023).
In a recent article (D’Costa 2024), I explore whether working in cities can boost women’s wages relative to those of men. We know that workers – even comparable workers in comparable jobs – earn higher wages in cities than in rural areas; this is the urban wage premium. However, we know little about the relative benefits of cities for women compared to men. In my paper, I measure gender differences in the urban wage premium and for the first time assess several explanations to provide insights for policies to reduce the gender pay gap.
The relative advantages of cities for women compared to men have disappeared
Until recently, most of the literature estimating and explaining the urban wage premium has used data on male workers, with only a limited number of studies conducting robustness checks with data on women or focusing on measuring the premium by gender, such as Phimister (2005), D’Costa and Overman (2014), De la Roca and Puga (2017), and Nisic (2017) (Almeida and Goncalves 2024).
I estimate the urban wage premium separately for men and women in Britain, using the Annual Survey of Hours and Earnings (ASHE). The ASHE is particularly useful because it allows me to follow 200,000 individual workers over time and across labour markets, tracking their wages, occupation, industry and, crucially, where they work (ONS 2024). With this rich set of job transitions, my estimates of the urban wage premium control for spatial sorting – i.e. the fact that more productive workers tend to sort into larger labour markets – which allows me to investigate the theoretical mechanisms behind the urban wage premium for women relative to that for men, distinguishing my analysis from previous research.
When considering the entire period of 1999 2019, urban wage premium for women looks significantly larger than that for men (2.2% versus 1.6%). But this fact hides important variation over time. Until the 2008 financial crisis, the urban wage premium for women is more than twice as large as that for men (2.8% versus 1.2%). This difference disappears during the financial crisis, as urban wage premium for women falls drastically and permanently. In contrast, the urban wage premium for men does not change significantly over time. During the crisis (2008–2013) and the recovery period (2014–2019), the urban wage premium stays between 1% and 1.5% for both men and women. Although I focus on the role of cities versus rural labour markets, the results are robust to using other size cutoffs for cities or measuring the density premium.
Figure 1 The urban wage premium in Britain by gender and period
Note: Figure plots urban wage premium estimates with 95% confidence intervals by gender and period, measured by the coefficient on a city indicator from OLS regressions of logged hourly wages on city, worker age, age squared and occupation, industry, part-time status, presence of a collective agreement, and worker and year fixed effects. This shows that the urban wage premium was significantly higher for women prior to the global financial crisis and that the gender difference is not significant during or after the crisis period, as women’s urban wage premium has dropped.
Testing the mechanisms underlying the higher urban wage premium for women
A number of theoretical mechanisms could explain why the urban wage premium was originally larger women than for men. First, I consider whether productive women tend to sort into urban jobs more or less than men. Second, I explore whether agglomeration economies are more pronounced for women than for men. If so, women (more so than men) would be more productive and therefore earn higher wages in cities. Because frequent job transitions and career interruptions negatively affect women’s wages (Adda et al. 2017, Goldin 2014), they may benefit more than men from matching in the thick labour markets of cities. In addition, contributing to the new literature on occupational matching in cities (Koster and Ozgen 2021, Papageorgiou 2022) and given the occupational segregation of women, I investigate the role of labour market differences in occupational diversity and changes in occupation that occur when a worker transitions between an urban and rural job. I also test whether women benefit more than men from faster learning or from better shared infrastructure in cities. In particular, sharing externalities from better transport links and access to childcare in cities may be more relevant for women.
Third, I consider several explanations from labour economics that may affect women’s wages differently in cities. Among these, I control for the number of hours worked, as reduced working hours by women are an important factor in the gender pay gap (Costa Dias et al. 2020). Markedly lower working hours for women in rural areas could translate into a higher urban wage premium for them. I also control for commuting distance, given the vast evidence that women’s wages are particularly sensitive to commuting costs (Black et al. 2014). Finally, I control for labour market industrial diversity due to the segregation of women into female-dominated industries, for the skilled labour share, and for local employer concentration, to account for urban-rural differences in the strength of gender discrimination (Hirsch et al. 2013).
I find that two main factors explain the pre-2008 gender difference. First, sharing externalities (such as childcare and transport infrastructure) boosted urban women’s wages compared to those working in rural areas, but there is no evidence of such benefits accruing to men. Second, women suffered a large wage growth penalty of -2.4 percentage points when they transitioned from an urban to a rural job due to poor occupational matching, while the effect of urban-to-rural job transitions on men’s wage growth was insignificant. Rather than greater occupational choice in cities, it is poorer occupational options during transitions to rural areas that have explained the greater urban wage premium for women.
Explaining the large drop in the urban wage premium for women since the global recession
The opportunities that urban employment offers to reduce the gender pay gap seem to have gone in Britain. Why did the urban wage premium for women relative to men disappear after the financial crisis? I find that women’s wages are no longer disproportionately boosted by access to shared facilities in cities. Evidence of sharing externalities disappeared during the crisis and reappeared only to a lesser extent in the recovery period (2014–2019). This is very important for policy because the global recession coincided with the start of austerity policies in the UK that reduced the provision of childcare and public transport, mostly in cities. These policies were therefore particularly detrimental to women working in cities.
The analysis of wage dynamics also reveals an important facet of gender inequality related to job transitions. In the recovery period, an urban wage premium of -1.4 points for women was still driven by a wage-growth penalty incurred while transitioning from urban to rural jobs; now, men experience increased wage growth of +1.1 points during such transitions. The wage-growth penalty for women is related to both occupational downgrading and switching to lower-paying employers when women take up jobs in rural areas. By contrast, men tend to switch to a higher-paying employer. This indicates that women’s job searches in rural areas are particularly constrained compared to those of men. Indeed, in the UK, women who switch to rural jobs typically do so after they have had children. In the absence of sufficient childcare options and transport links to jobs, gender norms compel women more than men to restrict their job searches geographically.
Figure 2 The role of urban-rural transitions in wage growth, 2014–2019
Note: Figure shows the effects of staying in an urban job, transitioning from a rural to an urban job, and transitioning from an urban to a rural job on wage growth in the same year for the period 2014–2019 with 95% confidence intervals. The yearly change in logged hourly wages was regressed on the transition indicators and first differenced worker and job characteristics, industry, occupation, and year indicators. This shows that though men and women obtain similarly higher wage growth from switching to urban jobs or remaining in urban jobs, women are penalised when they leave the urban labour market whilst men benefit.
Current outlook
To restore the role of cities in reducing the gender pay gap, policymakers should increase the provision of childcare and other forms of urban infrastructure that raise women’s productivity. In September 2024, the UK began a phased expansion in government-funded early-years childcare hours, to be continued in September 2025. The Autumn Budget on 30 October 2024 also announced free breakfast clubs in additional primary schools. However, there will be no increase in direct support to nurseries to expand supply. Nonetheless, if these interventions are significant in cities, they could help reverse the drop in women’s urban wage premium.
Reducing gender inequality when workers transition to rural jobs could be achieved by addressing gender norms and via interventions to reduce women’s geographical constraints, such as the recently announced increases in funding to local government and transport systems.
Source : VOXeu
