This paper estimates the causal effects of state-level Pay Transparency Acts (PTAs) on wage levels and within-occupation wage dispersion in the United States using Bureau of Labor Statistics Occupational Employment and Wage Statistics data (2015–2024). Leveraging staggered policy adoption across six states, I distinguish between two regimes: mandatory disclosure (salary ranges posted; Colorado, California, Washington) and upon-request disclosure (information provided only when solicited; Connecticut, Nevada, Rhode Island). I develop a Nash bargaining model with occupation-specific observability, informed by the Autor–Levy–Murnane framework, predicting that mandatory disclosure raises wages—especially in non-routine cognitive occupations—while upon-request policies may yield null or negative effects due to costly worker self-selection. Empirical results partially support these predictions: Mandatory PTAs increase wages by 1.5–2.6% at the 75th and 90th percentiles, with strongest effects in nonroutine cognitive fields (e.g., computer science, engineering). In contrast, upon-request PTAs reduce median wages by 2–5% in occupations such as sales, community services, and transportation, and substantially widen dispersion. The central conclusion is that compliance design is the primary determinant of policy impact; upon-request regimes may exacerbate inequality by shifting information costs onto less-advantaged workers.
