Replication package: Redistribution without reallocation — Exchange-rate shocks and the resource-sector wage premium in Azerbaijan

Replication package for "Redistribution without reallocation: Exchange-rate shocks and the resource-sector wage premium in Azerbaijan." The paper asks when the distribution of resource rents through the Azerbaijani labour market changed, and through which channel. Rather than fixing February 2022 as a treatment date, it dates the structural break first and examines the mechanism afterwards. Version 2 adds the sector-year panel: average monthly nominal wages by kind of economic activity, 19 sectors annually from 1999 to 2024, or 494 sector-year observations, together with the difference-in-differences, placebo-date and event-study code that uses it. Version 1 covered the aggregated two-sector series and the synthetic control exercise only. Contents: code/panel.py — sector-year panel construction, difference-in-differences with sector and year fixed effects, placebo treatment dates, event study, cross-sector dispersioncode/fig.py — event-study figurecode/wages.tsv — nominal wages by economic activity, 19 sectors by 26 yearscode/scm.py, main.py, placebo.py, loo.py, spec.py — synthetic control estimator, baseline, in-space placebos, leave-one-out and specification checksdata/ — source files from the State Statistical Committee of the Republic of Azerbaijan and the World Bank World Development Indicators, unmodified, plus the compiled datasetsoutput/ — every estimate reported in the paper, as JSONfigures/ — the three figures in the paper All data are public. Azerbaijani series: State Statistical Committee of the Republic of Azerbaijan (https://www.stat.gov.az). Cross-country series: World Bank World Development Indicators (https://databank.worldbank.org). Downloaded September 2026. Code is released under the MIT licence; data carry the terms of the original statistical agencies.