From Disclosure to Accountability: A Working Model

The UK reporting case has established that a public pay statistic can be followed by a measurable change in firm pay dynamics. The measurement discussion has also shown why a gap is not a diagnosis, while the wage decomposition has shown that a smaller gap can come from slower male pay growth rather than higher women’s pay. This checkpoint connects those ideas into one working model, so the reader can judge where information becomes pressure, where the evidence is strong, and where the chain can break.

The value of the model is practical. It prevents two errors at once: treating disclosure as a self-executing solution, and dismissing a real policy effect because it does not solve every source of gender inequality. The question is not whether transparency is simply good or bad. It is which link in the chain is doing the work.

A number becomes an intervention only through comparison

A reported gap begins as a description of a distribution. It says that women and men in a defined population occupy different positions in pay. That description becomes an intervention only when someone can make a meaningful comparison and use the result to question a decision.

The first link is therefore measurement. A mean or median firm-level gap can show that an organisation’s pay distribution is uneven, but it combines occupation, hierarchy, hours, tenure, care-related career interruptions, promotion and reward patterns. It can signal a problem without identifying its cause. If the report’s categories are too broad, it may not show whether people doing comparable work are paid differently. If the relevant work is divided between male- and female-dominated occupations, it may miss a difference in the value assigned to those occupations altogether.

The second link is comparison. A meaningful comparison may be between workers in the same job, between groups performing work of equal value, or between a firm and comparable employers. Each comparison answers a different question. The UK public indicators made firms visible to one another and to outside audiences, but they did not by themselves provide a full evaluation of every job or pay decision. The near-threshold research comparison likewise served a different purpose: it estimated how the reporting obligation changed worker outcomes relative to similar firms that were not required to publish.

That distinction explains why the UK result is both important and limited. In firms near the 250-employee threshold, the study estimates a relative increase of about 3 percentage points in women’s hourly pay compared with men’s, corresponding to a 19% reduction in the pre-policy gap in that sample. This is evidence that the policy changed the measured pay relationship under the study’s comparison design. It is not evidence that the economy-wide gap fell by 19%, that every firm responded in the same way, or that the reported number identified discrimination directly.

A useful way to read the result is to ask what kind of comparison generated it. The study compares treated and control firms across time and compares women with men within those firms. Individual-by-firm fixed effects help isolate changes within observed employment relationships.