Gender pay gap disclosure works, up to a point

9 June 2019

Evidence from Denmark shows that gender pay disclosure has caused companies to cut the pay gap, through restraining men’s pay rather than boosting women’s. 

The growing focus on gender discrimination

Gender pay disparities are becoming topical the world over. Blatant pay discrimination – different pay for the same job – is illegal in most advanced economies. But economy-wide pay differentials persist. These are driven by two factors: greater prevalence of women in lower paying occupations, and greater prevalence of men in senior higher paying positions within occupations.

In response to the persistent gender pay gap, governments are resorting to regulation to accelerate the glacial pace of change. A popular tool is requiring disclosure of the ‘gender pay gap’ which measures the difference between average pay of women and men. Critics say that this is a meaningless measure – the important point is whether men and women doing the same job are getting paid the same. Supporters say this is beside the point – the gender pay gap sheds much needed light on the structural injustice of women’s under-representation in the best-paying jobs. 

Assessing the impact of gender pay gap disclosure in Denmark

Does gender pay gap disclosure have an impact? Anecdotally it seems clear that organisations are focussing on explaining, and seeking to do something about, the unflattering metric. But hard evidence is more difficult to come by. Fortunately, Denmark collects extremely detailed individual-level pay data that enables gender pay gaps to be tracked over time at the firm level. Even better, a law change in 2006 created the conditions to assess the causal impact of gender pay gap disclosure requirements. 

This introduced a requirement for all firms in Denmark with 35 or more employees to disclose the gender pay gap (there is a complicated wrinkle in the legislation that makes this an oversimplification, but it is sufficiently accurate for the purposes of this article). The existence of the size threshold creates sets of firms just on one side of the threshold or the other (e.g. 20 to 34 employees versus 35 to 50) that are very similar in every respect other than one group has to disclose the gender pay gap, and the other doesn’t. This enables a test of whether gender pay gap disclosure causes firms to act differently.

A group of academics, Morten Bennedsen, Elena Simintzi, Margarita Tsoutsoura, and Daniel Wolfenzon, have undertaken the painstaking work of assembling the data to investigate this question. The results have triggered much press interest all round the world. Bennedsen discussed the results at the 3rd BI Conference on Corporate Governance in Oslo last month. There were three main conclusions:

• Gender pay gap disclosure works: the pay gap in firms required to disclose reduced by 7% over three years compared with no change for those that did not disclose.

• The gap is closed by holding back the pay of men rather than boosting the pay of women: women’s pay progression was indistinguishable in the two groups of companies, but men’s progression was lower in the companies that had to disclose.

• Company productivity falls: revenue per head reduces in firms that are required to disclose, but this is offset by the lower pay per head in these firms and so overall profitability is unaffected.

In addition to these main findings, there were also some interesting insight to the dynamics of pay gap closure at affected firms. First, firms required to disclose hired and promoted more women, particularly promotions from lower to middle levels in the hierarchy. Second, pay gap closure was more rapid in firms that had a higher pay gap prior to disclosure. Third, pay gap closure was more rapid in firms whose leaders had more daughters. This is a commonly used proxy used to indicate a higher likelihood of female-friendly management practices. They do not report on whether the productivity difference persisted in these firms.

The fact that pay gap disclosure has an impact should not be surprising. The benefit of the disclosure is that it creates a strong moral incentive to show progress in reducing the gap. However, the study also gives sustenance to critics of gender pay regulation. Can the policy really be said to have worked if it results in a levelling down of men rather than a levelling up of women and if productivity takes a hit? 

The productivity puzzle

At the BI conference seminar, Bennedsen pointed out that while the pay conclusions were strong, the productivity result was relatively weak by comparison, but too strong to be ignored. He cautioned that lazy assumptions about a win-win on gender pay gap and productivity needed to be challenged. Life may not be that accommodating. But what could explain the fall off in productivity? 

The paper suggests that the impact of the pay gap may have both positive and negative implications for motivation and hence productivity:

• On the positive side, a fairer pay system may lead to better employee engagement and motivation, particularly for women. 

• On the negative side, men may be demotivated at getting lower pay rises, and even women could be demotivated because of the knowledge, furnished by the disclosure, of the extent to which pay is unequal.

The implied conclusion is that the two negative factors outweighs the positive one. Perhaps not a surprise given that 70% of employees in the studied firms were men.

However, there’s another plausible explanation. Claudia Goldin has undertaken extremely detailed research into the gender pay gap using job classification data and pay data for the US. She found that the gender pay gap in the US is explained more by pay differentials within certain particularly highly-paid occupations than by different prevalence of women across occupations. Occupations such as senior corporate leadership, law, and finance exhibit significant pay gaps, which are in turn driven by the tendency for non-linear relationships between hours worked and the wage rate: in other words, in these occupations the pay per hour for individuals working 70 hours per week is more than twice the pay per hour for those working 35 hours per week. 

There is a significant wage premium in certain occupations for submitting oneself to the constraints of being available to work long hours and to be available at specific times for the convenience of and at the choosing of others. Senior corporate leadership, professional services, finance are all examples. Goldin notes that women are generally less inclined (or in light of social norms around child care less able) to sacrifice ‘temporal flexibility’, which incorporates all of: the number of hours to be worked and also the particular hours worked; being ‘on call’; providing ‘face time’; being around for clients, group meetings and the like. Alternatively, they may take the roles but achieve less in them. Goldin contends that the gender pay gap can only be closed if jobs are restructured and also remunerated in a way to enhance temporal flexibility, for example by improving the ability of individuals to substitute for each other. In some occupations this has already happened with the result that the pay gap is close to zero, but much less so in high paying ‘winner takes all’ occupations such as corporate leadership, the law, finance and so on.

Therefore, a possible explanation for the Danish findings is that addressing the pay gap by putting more women into roles that are designed for men, and where they are as a consequence unlikely to flourish, may lower the pay gap but not be good for productivity. Only when jobs are restructured to suit women can the dual benefit of greater equality and greater talent utilisation (and hence productivity) be realised. 

Flexibility for all

This is all speculation. But if it is the explanation it gives some cause for hope. The Danish study focusses on small companies with around 35 employees with 70% male employees on average. Such companies may find it more challenging to change job types and working patterns to create a more women-friendly workplace. Anecdotal experience of larger firms is that they are very aware of the cluster of HR practices that need to be changed in order to create true gender equality. They also have the resources to achieve it, although as acknowledged in PwC’s Time to talk publication from last year, much remains to be done.  Nonetheless, an optimistic view would be that it’s possible in larger companies the pay gap is being closed at the same time as maintaining and increasing productivity. 

Regulation almost always seems to have a surprising unintended consequence. While it shows that regulation can close the pay gap, the Danish study could also be used as an argument against regulation in this area. That would be a pity. While some differences in gender preferences over career choice may persist, it is lazy to use this as a justification for the gender pay gap when so little has been done to challenge accepted male-oriented norms about job structures and working patterns. The voluntary pace of change has just been too slow. Regulation can provide a useful kick start. But the change will only be productive if firms put in the effort to adapt how they operate to reflect the new reality. This will benefit men as well as women by enabling more flexible working patterns to emerge without today’s severe wage penalty.

---