Abstract
Abstract This study examines the relationship between taxation structure and macroeconomic performance within a comparative panel comprising Nigeria, South Africa, the United Kingdom, and Canada over the period 2000–2024, while undertaking a forward-looking assessment of Nigeria’s 2026 progressive tax reform using a simulation-based counterfactual framework. The analysis combines two-way fixed effects and threshold regression models to examine historical associations between taxation and economic performance and employs a Synthetic Control Method (SCM) to compare projected reform outcomes with a counterfactual trajectory constructed from comparator economies under explicitly defined simulation assumptions. A machine-learning T-learner is further implemented exclusively as a predictive validation tool to assess the consistency of projected outcomes rather than to establish causal effects. Unlike conventional ex-post policy evaluations, the analytical framework is specifically designed to support transparent ex-ante assessment in the absence of observed post-reform outcomes, with all projected estimates interpreted as conditional on the stated modelling assumptions. The historical panel estimates indicate that higher tax-to-GDP ratios and greater reliance on direct taxation are associated with stronger real GDP per capita growth, with these associations appearing more pronounced in economies operating above estimated fiscal thresholds. Institutional quality and gross fixed capital formation are likewise associated with more favourable growth outcomes. Under the simulation assumptions adopted, the counterfactual analysis suggests that Nigeria’s 2026 progressive tax reform may be associated with modest short-run adjustment costs, potentially reflecting transitional constraints related to institutional capacity and tax administration efficiency before longer-term fiscal benefits may emerge. The projected short-run deviation should be interpreted as a scenario-dependent counterfactual estimate rather than a realised policy effect or economic forecast. These findings emphasise the importance of distinguishing between historical associations and forward-looking policy simulations, while highlighting the role of institutional capacity in shaping projected reform outcomes. The study contributes to the literature by integrating structural panel econometric analysis, simulation-based counterfactual evaluation, and machine-learning validation within a transparent ex-ante policy assessment framework for fiscal reform in developing economies, while explicitly recognising the uncertainty inherent in forward-looking simulation-based analyses.