Abstract
This study investigates the impact of corruption on the effectiveness of government spending in promoting economic growth in South Africa over the period 1990–2020. Unlike earlier studies that treat government expenditure as a single aggregate, this study adopts a disaggregated approach, categorising spending into four groups, core governance and administration, economic development, social development, and culture and quality of life. Using the Autoregressive Distributed Lag (ARDL) model, the analysis explores both the direct effects of spending and corruption on growth, as well as their interaction.
The findings reveal that corruption significantly shapes the spending–growth relationship, with sector-specific variations. Core governance and social development spending become more growth-enhancing as corruption control improves, supporting the “sand in the wheels” hypothesis. In contrast, economic development spending and culture and quality of life display relatively stronger growth effects in higher corruption conditions, suggesting that overly strict procurement procedures and administrative inefficiencies may influence how spending translates into growth.
Overall, these results highlight the importance of considering both the type of government spending and the quality of government when designing fiscal and anti-corruption policies. The study contributes to the literature by providing the first disaggregated analysis of spending and corruption interactions in South Africa, filling a gap where most prior research has relied on aggregate data. Policy recommendations show that anti-corruption reforms should be supported by improvements in sector-specific systems to ensure that public spending contributes more effectively to economic growth.