Nonlinear Inflation–Growth Nexus in Asia, Latin America, and Sub-Saharan Africa: A Cross-Regional Institutional Analysis

This repository contains the replication dataset and estimation code for a study testing whether the inflation–growth threshold is conditioned on institutional structure, using panel threshold regression across 115 countries in Asia, Latin America, and Sub-Saharan Africa (1970–2023). Data coverage: Annual panel data for 115 countries across three regions, with Sub-Saharan Africa further disaggregated into Francophone (CFA franc zone) and Anglophone monetary-regime subsamples. Three-year and five-year non-overlapping temporal averages are also included alongside the annual series. Variables: GDP per capita growth (dependent variable); CPI-based inflation, semi-log transformed following Ndoricimpa (2017) (threshold variable); and control variables covering government expenditure, gross fixed capital formation, terms-of-trade level and volatility, population growth, political stability, institutional quality (rule of law and regulatory quality), trade openness, and initial income (lagged log GDP per capita). Sources: World Bank World Development Indicators (WDI), World Bank Worldwide Governance Indicators (WGI), IMF International Financial Statistics (IFS), and UNCTAD. Data construction: Missing values are addressed via within-country linear interpolation where feasible, and via peer-group median imputation (colonial-legacy or sub-regional grouping) at series endpoints. Threshold-relevant variables are winsorized at the 1st and 99th percentiles. All imputation and winsorization decisions are logged in the package. Contents: Cleaned analysis datasets (annual, 3-year, and 5-year averaged panels), Stata do-files for Hansen (1999) fixed-effects panel threshold regression, linear IV (xtivreg2), threshold-consistent IV-GMM, and system GMM (xtabond2) estimation, plus output logs. License: CC BY 4.0