Ethnic Attainment Gaps and Labour Market Outcomes: Evidence from England: Replication Files
Description
This replication package contains Stata code for "Ethnic Attainment Gaps and Labour Market Outcomes: Evidence from England" (Sean Brophy, Manchester Metropolitan University). The paper examines the labour market cost of the ethnicity degree awarding gap across employment and earnings, using HESA student records linked to Graduate Outcomes survey data for five cohorts of first-degree qualifiers at English higher education providers (2018/19–2022/23). A sequential decomposition grounded in the Cunha-Heckman skill formation framework decomposes ethnic employment and earnings gaps into prior attainment, degree classification, geographic sorting, and residual labour market channels. The employment analysis uses 870,045 survey respondents; the earnings analysis uses 376,740 graduates in full-time paid employment at 15 months post-graduation. The package contains three do-files implementing the full analysis pipeline (data import, sample construction, and analysis), a README, a codebook, and a data availability statement. The underlying data are not publicly available and are not included; see the data availability statement for access information.
Files
Steps to reproduce
he analysis requires Stata 17 or later and four community-contributed packages: reghdfe, ftools, estout, and leebounds. Install via: ssc install reghdfe; ssc install ftools; ssc install estout; ssc install leebounds. Run the three do-files in order from a working directory containing the analytic dataset (go_analysis.dta): 00_import.do — imports the raw Graduate Outcomes CSV extract and saves grad_outcomes_clean.dta 01_data_prep.do — constructs the analytic sample and saves go_analysis.dta 02_analysis.do — implements the full analysis, producing all tables, figures, and saved estimates The working directory path in 02_analysis.do must be updated to match the local directory before running. The do-file creates tables/, figures/, and results/ subdirectories automatically on first run. The analytic dataset (go_analysis.dta) is not provided. It is constructed from a bespoke HESA/Graduate Outcomes extract obtained via Jisc under a Data Sharing Agreement. Researchers wishing to replicate the analysis should apply to Jisc/HESA for equivalent data access; see the data availability statement for details. Key outputs: Tables 1–7 and Figures 1–2 in the main text; Appendix Tables A1–A9 and Appendix Figure A1.