Economics remains a male-dominated field. In the US, women account for 28.8% of PhD graduates but a mere 13.9% of full professors in economics (CSWEP 2017). This underrepresentation of women is perhaps nowhere as visible as in central banks (OMFIF 2019). For instance, as of the date of writing, there is not a single woman on the 30-member General Council of the ECB. A recent survey conducted by the American Economic Association on the professional climate in the economics profession paints a disheartening picture of an overly competitive and hostile environment for women (AEA 2019). What holds women back from pursuing a career in economics?
What is behind the underrepresentation of women in economics?
Several explanations may account for the lack of women in high-level positions in the economics profession. One possibility is that the pool of potential applicants is male dominated. Despite recent efforts to turn the tide, women remain less likely to study economics, and macroeconomics in particular (Ginther and Kahn 2004). An alternative explanation is that women are less likely to apply for promotions because of gender differences in the preference for competitive environments (Niederle and Versterlund 2007) or in bargaining abilities in the labour market (Blackaby et al. 2005). The presence of children and trade-offs between family and career may also hold back women from pursuing promotions (Bertrand 2013). Finally, there may be gender-based discrimination in promotion decisions (Goldin and Rouse 2000).
Which of these explanations is more relevant? And can corporate diversity policies mitigate these biases? Despite a large body of literature on gender differences, there is no agreement on the importance of diversity policies and their impact on labour market outcomes.
Addressing gender balance at the ECB
In a new paper (Hospido et al. 2019), we analyse the career progression of men and women at the ECB, using confidential anonymised personnel data from professional staff during the period 2003-2017. Our analysis focuses on expert staff across four different salary bands representing different levels of seniority (expert, senior expert, principal expert, and advisor) in the policy areas, the research department, and the statistics department. These are business areas across which we observe considerable flows of staff over time. With this selected group, we focus on a broadly homogeneous pool of staff in terms of human capital and experience, ensuring comparability across individuals.
We find that a wage gap emerges between men and women within a few years of hiring, despite broadly similar entry conditions in terms of salary levels and other observables. One important driver of this wage differential is the presence of children. We also find that women are less likely to be promoted to a higher salary band up until 2010, when the ECB issued a public statement supporting diversity and took several measures to support gender balance. Following this change, the promotion gap disappears.
Have you read?
Figure 1 shows in more detail that this change in diversity policies had material effects on gender differences in promotion outcomes. The figure focuses on promotions from salary band F/G, which is the entry-level salary band for professional economists at the ECB. The gender gap in promotions is defined as the difference in the promotion rates of men and women. The promotion gap narrowed from 2011 onwards, following the policy change. While prior to 2011, the gender promotion gap stood at over 36% after ten years since entry, this gap decreased to about 8% on average after 2011, or a decline of about 80%.
Figure 1 Gender gap in the probability of promotion from salary band F/G (before 2011 and from 2011 onwards)
Using 2012-2017 data on promotion applications and decisions, we explore the promotion process in depth, and confirm that during this most recent period, women are as likely to be promoted as men. This results from a lower probability of women applying for promotion, combined with a higher probability of women being selected conditional on having applied. We coin this reluctance to apply for promotions the ‘gender applications gap’. Following promotion, women perform better in terms of salary progression, suggesting that the higher probability of being selected is based on merit, not positive discrimination. We do not find evidence that the composition of the selection committee, including the fraction of women on the panel, alters these results. Taken together, these results point to the effectiveness of corporate diversity policies in reducing gender bias in promotions and lend support to supply-side explanations for the existence of remaining gender differences in promotion outcomes.
We make three contributions to the literature.
- First, we are the first to exploit the complete personnel records of a large organisation to analyse gender bias in career progression and promotion decisions. This allows a more comprehensive analysis of career progression across various job levels within an organisation, in contrast to much of the literature that focuses on gender differences at corporate board or leadership levels.
- Second, in contrast to much of the literature on promotion decisions, we simultaneously consider the role of promotion applications and decisions when identifying the drivers of the promotion gap. Analysing promotion decisions without accounting for gender gaps in applications would bias the results. We are able to do so because we have information on both promotion applications and decisions, while existing literature has focused on only one of these dimensions.
- Third, we exploit a change in the ECB's gender policy to assess the impact of corporate diversity policies on promotion outcomes. While the economics literature has assessed the impact of gender quotas for corporate board seats on corporate decisions, to our knowledge we are the first to consider the impact of broad-based corporate diversity policies on female labour market outcomes.
Our results suggest that institutional efforts to reduce the gender promotion gap may have to include measures aimed at lowering the barriers for women to seek and apply for promotion opportunities. Such measures could range from offering assertiveness and interview trainings to enhanced child support benefits and services. More generally, efforts to curb the overly competitive and hostile environment in the profession seem desirable. Understanding the main drivers of the observed gender promotion gap is critically important to improve our understanding of how we can close the gender gap and ensure that women are adequately represented.