The Price of flexibility: Revealing salaries in job postings. Guest post by Amen Jalal

This is the 2nd in this year’s series of posts by PhD students on the job market.

Around the world, women sort into lower-paying firms than men, which explains up to half of the gender pay gap. My job market paper asks: do women prefer these low-paying firms, or do they face barriers to accessing high-paying ones?

The conventional explanation emphasizes preferences. The argument is that women's employers are not only lower-paying but also more family-friendly, so women may be "buying" amenities with foregone wages. However, to fully account for the observed wage gaps, women's willingness to pay for these amenities would have to be incredibly large. In practice, lab-based estimates of such preferences are orders of magnitude smaller. So while the preference explanation is important, it is also incomplete.

My paper brings a new fact to this conversation: information frictions. Specifically, most workers apply to jobs without knowing the salaries. Across many high- and low-income countries – including the U.S., Germany, China and Ethiopia – over 80% of job postings omit salary information. This means that when women trade wages for amenities, they may not know the price they're paying.

To test the importance of this information friction, I partner with Pakistan's largest online job search platform. In a field experiment with 20,088 jobs from 8,906 firms, I randomize whether pay disclosure in job ads is mandatory (treatment) or optional (control). I find that mandating pay transparency nearly doubles women's applications to jobs at large, high-paying firms (from 20.6 to 40.2) and reverses gender gaps in directed search, even though the wage and amenity landscape of these firms does not change. This shows that gendered sorting arises not because women prefer family-friendly amenities to wages, but because they do not observe the price of those amenities.

The Equilibrium Job-Seekers Face

To separate preferences from information frictions, we need to observe not only whether men and women end up in different jobs, but also whether they apply to different jobs for given choice sets. Through my partnership with the platform, I track application decisions as well as consideration sets. I acquire the platform’s data spanning five years, where 3 million workers send 29 million applications to over 300,000 jobs and 62,000 firms. A distinctive feature of these data is that I observe salaries even when job-seekers do not because firms are required to report them to the platform. I supplement these data with firm and worker surveys. The latter include a discrete-choice experiment to measure preferences over amenities and wages, and incentivized belief-elicitation to capture perceptions about hidden wages.

Three key facts emerge from these data. First, while large firms pay more, they are less likely to disclose salaries in job ads. Consequently, salaries are more likely to be hidden when higher for otherwise similar jobs. Second, only 11% of surveyed job-seekers – men and women alike – believe hidden salaries to be higher than posted ones, revealing a potential window for policy action through pay transparency mandates. Third, large firms offer fewer family-friendly amenities like flexibility, which women – as revealed by the discrete-choice experiment – value moderately more than men. Together, these facts imply that pay is most often hidden precisely where it may be high enough to compensate women for inflexibility.

The following figure shows that men and women respond similarly to pay information but differently to its absence. Holding job characteristics fixed, when pay is posted, male applications increase 30% and female applications 28% for each log-point increase in salary. But when salaries are hidden, this pattern reverses. Male applications become unresponsive—reflecting the lack of information—while female applications decline by 16%.

Figure 1

My conceptual framework proposes a mechanism to rationalize these patterns. Informed by data, the framework assumes that large firms pay more, offer less flexibility, and hide wages; small firms do the reverse. Men and women value pay equally but place different weights on flexibility. When pay is hidden, uncertainty reduces the appeal of large firms that conceal wages. Transparency therefore raises applications to these firms for everyone. But the effects are gendered even though preferences over wages and wage uncertainty are not. This is because wage uncertainty magnifies the relative appeal of flexibility, which men and women value differently. If women value flexibility more, more of them may sit on the margin where revealed wages just offset the amenity loss. I test this explanation with the field experiment.

The Field Experiment

Mandating pay transparency increases applications by 49%. Effects are driven by large firms: their applications surge by 66% versus 23% at small firms. This suggests that workers learn new wage information about large firms and find it appealing. Indeed, large firms in the control group hide salaries 30% more often, and offer 32% higher minimum and 21% higher maximum salaries, with a narrower range. In treatment, large firms continue to offer higher and more compressed salary ranges.

Consistent with the model, the impact is disproportionate for women. Treatment increases male applications to large firms by 59% and female applications by 95%. This reverses the gender gap in directed search: in the control group, large firms – relative to small firms – get 20 percentage points fewer female than male applications. In treatment, this gap flips in favor of women, to 43 percentage points.

The Mechanism: Substituting Flexibility for Pay

As women shift toward large firms, they also substitute away from flexibility. To capture this trade-off, I develop a natural language processing algorithm that classifies mentions of remote work, flexible hours, transport support, safety, and inclusive language in job ads. I then test whether the appeal of these ‘amenities’ depends on salary disclosure. Remote work stands out: large firms are 50% less likely to offer it, while women value it highly. Figure 2 shows that remote jobs without posted salaries receive 169% more female applications but only 71% more male applications – all else constant. Mandating transparency leaves the amenity supply unchanged but sharply reduces its appeal to women. In treatment, women’s applications to remote jobs fall by 79 percentage points, while to large firms, they rise by 47 percentage points. Male behavior remains stable in contrast. I also replicate these patterns in a discrete-choice experiment that cross-randomizes salary and amenity information.

Figure 2

What Happens on the Demand Side?

In baseline surveys, most non-disclosing firms expect transparency to raise applications but reduce quality – and they are partly right: at large firms, the share of applicants meeting the job’s skill criteria falls by 6%. Yet, firms appear to overestimate the potential screening costs that come with more applicants and reduced quality: after the experiment, while untreated large firms remain about 80% likely to hide pay, those fully exposed to the experiment are 30% less likely to do so.

A potential reason is that transparency improves the top of the applicant pool, especially among women. Among firms in the control group, 90% of jobs already attract at least one high-ability man but only 57% attract a high-ability woman. With transparency, that probability rises by 11% for women and just 1% for men.

Treated jobs become 7% more likely to download a female CV after reviewing applications on the portal, and report similar female hiring rates despite increased competition. Back-of-the-envelope calculations suggest that a representative woman becomes 18% more likely to be hired by a large firm in the treatment group than in control because the increase in women's applicant share offsets the increase in overall competition.

Key Takeaways

This paper shows that pay non-disclosure hides the true price of flexibility, steering women toward low-paying firms. It also tests a policy solution: pay transparency. Women’s applications to high-paying firms who post the salary in this experiment increase 95% (compared to men’s increase of 59%), which reverses the gender gap in directed job search toward these firms. What’s in it for the firms? Despite small declines in the average quality of applicants, transparency improves the quality of top applicants, especially women, inducing high-paying firms to voluntarily disclose salaries.

Amen Jalal is a PhD student at the London School of Economics.