Developing an Original Econometric Methodology for Estimating Causal Economic Effects in Non-Experimental Nigerian Data

Causal inference from observational data in Nigerian economics is constrained by limited availability of natural experiments, randomisation opportunities, and instrumental variables, and developing an original econometric methodology for credible causal effect estimation in the Nigerian data environment constitutes a significant methodological contribution. This study developed and validated an original causal inference methodology for non-experimental Nigerian economic data, addressing Nigeria-specific challenges including endogeneity from patron-client network effects, measurement error from informal economy activities, and selection bias in programme participation. A methodology development approach combining systematic review of 56 causal inference methodology publications from 2018 to 2024, original theoretical development of a Nigeria-adapted synthetic control method, and empirical validation across four policy evaluation case studies was employed. The Nigeria-adapted Synthetic Control Method introduces a weighting scheme that accounts for the spatial autocorrelation of economic outcomes across Nigerian states, the non-stationarity of key economic series, and the structural breaks induced by major policy events such as devaluations and subsidy removal. Empirical validation against known policy effects confirmed that the adapted method outperformed standard synthetic control in coverage probability and root mean squared error. Application to four policy evaluations (RMAFC transfer formula change, state investment incentive programmes, agricultural finance reforms, and security intervention effects) demonstrated the method's practical utility. Expert review by 20 econometricians confirmed the methodology's original technical contribution. Recommendations include the method being incorporated in graduate econometrics curricula at Nigerian universities.

Keywords: causal inference, synthetic control method, Nigerian economics, non-experimental data, econometric methodology