Author: Kathleen Beegle ;Lead Economist with the World Bank Gender Group
A number of recent posts on Development Impact have focused on gaps in labor market outcomes for men and women. This includes some posts in the job market series in late 2019 and, in 2020, a look at the ‘dismal statistic’ on female-to-male labor force participation rates in India and Pakistan, changing the way women think about themselves and what they can do as a means to closing gaps, and a summary of the World Bank’s Research Group recruitment. The latter was underpinned by goal setting in the shortlisting process with respect to between men and women applicants. Fun fact: such quota setting with respect to the composition of candidates to interview is known in the U.S. as the Rooney Rule — established by the U.S. National Football League in 2003 and named after the former owner of the Pittsburgh Steelers (and intended not in regards to women hires but with regards to ethnic-minority candidates in coaching and senior league jobs, though it is now used more generally in the private sector).
Solutions to address gender inequality in the labor market are often described in terms of supply and demand side interventions. Here I to take a look at a domain that doesn’t squarely fit in the demand or supply space: legislating approaches to closing gaps. In this post I look at pay transparency (and to come in the next blog that I post: parental leave).
Women are paid less than men in most of the world’s countries, sometimes by a wide margin. Is this “just” a problem in developed countries? No. Gender pay gaps are as larger or larger in low and low-middle income countries (see the ILO 2018/2019 global wage report).
Pay secrecy has been proposed as a key contributor to gender pay gaps in firms. A number of countries have fairly recent legal initiatives to require employers to put sex-disaggregated salary statistics in the public domain, including Australia, Denmark, Germany, Iceland, and the United Kingdom. Reporting gender wage gap data is not the only way to legislate pay transparency. Some states/cities in the U.S. have laws that prohibit employers from enforcing pay secrecy or laws that ban employers from asking potential hires about past earnings. Additional approaches (including by the European Commission) include insuring an employee’s right to request information on gender pay levels for the same work or work of equal value, requiring employers to conduct an audit on pay and pay differentials on grounds of gender, and measures to ensure that the issue of equal pay, including pay audits, is discussed in collective bargaining.
Greater pay transparency is theorized to close pay gaps because it can support individual worker action/reaction (especially relevant for women who may have weaker networks by which to glean information on wages) and/or spark public pressure. On the other hand, there are arguments against pay transparency, including the cost burden to firms to provide this information, potential productivity effects (via disgruntled workers), and higher worker turnover.
Does greater pay transparency shrink the pay gap? Bennedsen, Simintzi, Tsoutsoura, and Wolfenzon (2019) (earlier ungated version) tackle this question in Denmark, where a 2006 legislation change required firms to provide sex disaggregated wage statistics. Prior to the law, in the firms they study, male employees had a 18.9% wage premium over females; they find that this gap was not driven by differences in demographics, work experience, macro trends, or selection into specific occupations. Bennedsen et al. use a difference-in-differences approach, comparing firms with 35-50 employees on average from 2003 to 2005 (those covered by the law) to those uncovered with 20-34 workers. They find that the pay gap declined by 7 percent with mandatory reporting, and firm response was largest among firms with greater pre-law gender pay inequality. But did women’s wages go up? Not much. Rather, the gap shrank largely because male wage growth fell. (They do a number of robustness checks to support their causal interpretation which I won’t describe here.) From the firm perspective, despite concerns of the costs of such laws for firms, Bennedsen et al. find that the firm wage bill fell. However, this is offset by negative effects on productivity so as to yield no change in profitability. Can the law close other gaps? They find that the law resulted in an increase in female hires and promotions of women. They argue that the former is consistent with the argument that closing the gender pay gap cam result in greater female labor supply (that fairer compensation draws women into work).
To date, there are few other studies to add to the inventory. The impact found on gender pay gaps in Denmark is smaller than the findings on pay transparency effects on pay gaps in public university faculty salaries in Canada (Baker et al 2019). In this setting, the starting conditional pay gap is much smaller, around 6-7 percent. This study found that the pay gap declined by 30%, but also found that the effect is virtually zero for those with a large initial gap. However, the nature of the pay transparency approach in Canada is quite different than in Denmark, with the former being non-anonymous and publicly accessible pay data, as well as pertaining to public sector jobs.
Moving to another distinct setting, Cullen and Pakzad-Hurson (2019) model the implications of pay transparency, study data from online markets for low-skill, temporary jobs (back-end data from TaskRabbit), and conduct an online experiment (using Amazon Mechanical Turk). This paper covers a lot of ground! As with the two papers above, though not specifically in reference to gender pay gaps, they find that wages become more equal with transparency, but are lower, on average. The intuition behind this finding is that greater pay transparency results in employers negotiating more aggressively. This is theorized to be brought on because transparency changes the implications of one highly paid worker: other workers can more easily learn of the higher wage and leverage this information in their own negotiations (information spillovers). Looking at the nuances of pay transparency, they note that partial transparency can disadvantage women (that men have a ‘communication advantage’) and potentially increase the gender pay gap. Their experimental data show that full transparency is more beneficial for women’s wages than for men.
So while pay transparency efforts seem to be gaining momentum by governments as a tool to tackle pay gaps, these studies raise notes of caution. They do find that such efforts can reduce the pay gap. But (there is always a but) these efforts may not make a huge dent in the pay gap, nor result in higher wages for women.