Replication Package: Education Spending and Economic Growth in Sub-Saharan Africa — Governance, Colonial Institutions, and Dynamic Panel Evidence

This repository contains the replication data, code, and supplementary materials for "Education Spending and Economic Growth in Sub-Saharan Africa: Governance, Colonial Institutions, and Dynamic Panel Evidence." The paper examines the long-run growth effect of secondary education spending in Sub-Saharan Africa using the largest and longest panel assembled for this question to date: 47 SSA economies, 1970–2023, estimated with the Pooled Mean Group (PMG) estimator and cross-checked with Dynamic Fixed Effects, externally instrumented IV, and System GMM estimation. Secondary education spending alone shows no significant long-run growth effect, but its interaction with control of corruption is negative and significant, with the marginal return to education spending turning negative once governance quality crosses a threshold near the 40th percentile of the sample's corruption-control distribution — a result that runs counter to the governance-failure hypothesis (Pritchett 2001). A subsample analysis by colonial administrative origin (Anglophone vs. Francophone) shows this substitution pattern is substantially more pronounced in Anglophone economies. Contents: panel dataset (47 SSA economies, 1970–2023), PMG/DFE, IV, and System GMM estimation code, threshold and subsample robustness tables, and supplementary materials referenced in the manuscript. Author: Ernest Pele Botsyoe (ORCID: 0009-0003-5156-8667), GIMPA, Accra, Ghana License: CC BY 4.0