Public-Sector Compensation and the Private-Public Earnings-Dispersion Gap: Stata Replication Package

This replication package contains the analysis-ready country-year dataset, Stata code, variable dictionary and figure files supporting the study of public-sector compensation and the private-public earnings-dispersion gap. The harmonised panel contains 4,646 observations covering 202 economies from 2000 to 2022. The primary estimation sample comprises 826 observations from 105 countries. The master Stata script reproduces the main and supplementary analyses, including country and year fixed-effects models, strict one-year lag and adjacent-year change specifications, Mundlak estimates, alternative covariance estimators, sample and measurement sensitivities, stability diagnostics, OECD validation and exploratory workforce-composition analyses. It also generates the manuscript and supplementary figures. The code uses official Stata commands and was verified in StataNow/SE 19.5. The analytical data draw on the World Bank Worldwide Bureaucracy Indicators and World Development Indicators. Independent validation uses the OECD Data Explorer series Decile ratios of gross earnings, full-time dependent employees (OECD.ELS.SAE:DSD_EARNINGS@DEC_I(1.0)). The archive reproduces the reported estimates and figures from the deposited analytical panel. It does not include raw provider downloads or panel-assembly code. All findings represent observational within-country associations and do not identify causal effects or public-to-private wage leadership. File-specific licensing, source attribution and SHA-256 checksums are provided in the archive.