To get an overall picture of the relative contributions of revenue and expenditure forecast errors to fiscal slippage in South Africa, today’s post by Lisa Martin plots annual forecast errors at each forecast horizon
for both categories expressed as a percentage of each year’s expenditure and revenue. The post demonstrates that, on average since 2012/13, positive (under-estimated) expenditure forecast errors and negative (over-estimated) revenue forecast errors, contributed to larger-than-expected budget deficits. Before the COVID-19 pandemic, budget overruns were dominated by weaker-than-expected tax revenues. Since the pandemic, large unplanned expenditures were the greatest contributors to larger-than-expected borrowing requirements. As we show in our recent paper, free higher education and social relief of distress grants,
and unexpected borrowing to support state-owned enterprises, were the greatest contributors to larger-than-expected borrowing requirements since the pandemic.
Codera makes it possible to automate sophisticated analysis of economic and financial data. Our EconData platform, for example, makes it easy to evaluate historic IMF, SARB and National Treasury forecast errors and data revisions.
Footnote
In the chart above, the T-1 label shows forecast errors for in-year ‘revised estimates’. The labels T, T+1, T+2 compare outcomes from the latest Budget Review (the 2026 edition) to the three Medium Term Expenditure Forecast years published in earlier Budget Reviews. For example, the
outcome published for 2024/25 in the 2026 BR is compared with
previous estimates/forecasts to create the ‘T+2’ (grey) bar at 2022/23, ‘T+1’ (red) bar at 2023/24, ‘T’ (light blue) bar at 2024/25 and the ‘T-1’ (dark blue) bar at 2025/26.
