Fiscal Policy and Economic Growth in South Africa: Nonlinear Evidence for Transitory Keynesian Effects and Fiscal Risk

This study investigates whether government spending stimulates economic growth by applying the Keynesian theoretical framework across varying economic conditions. The analysis uses annual data from 1980 to 2024 to explore how fiscal dynamics change over time and across regimes. It employs the NARDL model to evaluate asymmetric effects, the STAR model to capture regime dependence, and threshold Granger causality tests to assess causal relationships across spending regimes. These approaches enable a detailed examination of asymmetry, structural breaks, and nonlinear adjustment in the spending–growth relationship. The results show that Keynesian effects remain present across economic regimes but operate only in the short run without generating sustained long-term output gains. The absence of long-run cointegration is consistent with the presence of short-run dynamic multipliers, because these multipliers reflect temporary adjustments rather than permanent effects. The findings indicate that increases and decreases in government spending have proportionate effects on output, confirming a symmetrical Keynesian response. Government debt demonstrates a consistently negative and statistically robust influence on short-run growth. Corruption, measured using an index capturing governance quality, heightens policy ineffectiveness during periods of high public expenditure. Threshold causality tests reveal that government spending Granger causes economic growth in both low and high spending regimes, confirming the short-run stimulative potential of fiscal policy. Consequently, the study supports countercyclical fiscal interventions while emphasising the importance of prudent debt management and governance reforms to reduce fiscal risks.