South Africa’s Investment Slump Is Exposing a Deeper Structural Growth Problem

South Africa's economy has not simply endured a sequence of bad years. Its underlying growth engine has been losing strength for decades, while the investment needed to expand productive capacity has become weaker, more volatile and less effective. A new study, "Investment Decline and South Africa's Growth Slowdown: Evidence from Bayesian Model Averaging," published in the journal Economies by Kazeem Abimbola Sanusi and Zandri Dickason-Koekemoer of North-West University, argues that the country's investment problem is inseparable from a deeper erosion of economic momentum.

Using quarterly data from 1960 to 2025, the researchers find that investment growth is the strongest contemporaneous macroeconomic correlate of GDP growth among the variables they examine. However, their results also complicate the tempting conclusion that more investment alone would restore expansion. Once the persistence of GDP growth itself is taken into account, investment's independent predictive power weakens sharply, shifting attention toward the structural conditions that determine whether capital can produce durable economic gains.

A Slowdown That Became Embedded After the Global Financial Crisis

South Africa's long-run growth trajectory has deteriorated markedly. The study's Bayesian Structural Time Series estimates show that underlying trend growth exceeded 5% near the beginning of the historical sample but fell substantially over subsequent decades, reaching around 1.1% by the third quarter of 2025. The weakening became especially pronounced after the global financial crisis, suggesting that recent stagnation cannot be understood as a temporary cyclical setback.

Comparisons across economic regimes make the shift more visible. Average real GDP growth stood at 3.31% in the pre-global-financial-crisis period, dropped to 1.43% between 2008 and 2019, and fell further to 0.63% during the COVID and post-COVID period. Investment growth followed an even steeper decline, moving from an average 4.67% before the crisis to -0.06% afterward and -1.79% in the most recent period examined.

The researchers estimate that underlying trend growth in the post-crisis, pre-COVID period was about 1.54 percentage points below its pre-crisis level, with a posterior probability of decline of 0.996. Post-COVID trend growth was around 2.19 percentage points below the pre-crisis average, although the evidence for a distinct additional break after the pandemic is less definitive. The implication is important: much of South Africa's growth deterioration appears to pre-date COVID-19 rather than originate with it.

Such persistence changes the diagnosis. A weak recovery after a temporary shock can often be addressed through demand support, easier financial conditions or short-term fiscal intervention, but a multi-decade loss of trend growth points toward deeper constraints on productivity and productive capacity. South Africa's electricity shortages, logistics bottlenecks, declining capital accumulation, institutional weaknesses and policy uncertainty therefore sit closer to the centre of the problem than any single downturn.

Investment Is the Strongest Signal, but Not a Smoking Gun

Investment initially appears to dominate the statistical evidence. In the study's baseline Bayesian Model Averaging analysis, growth in gross fixed capital formation receives a posterior inclusion probability of 1.000, meaning it is retained across the highest-probability specifications considered. A one-percentage-point increase in investment growth is associated with roughly a 0.214-percentage-point increase in real GDP growth, although the authors explicitly caution that this is an association rather than a causal estimate.

Exchange-rate movements also display substantial explanatory relevance, while interest rates and government expenditure receive weaker or more specification-dependent support. Investment's prominence remains stable when the researchers alter statistical priors and when the treatment of interest rates is changed. Such robustness strengthens the case for viewing weak capital formation as a defining feature of the country's low-growth environment.

A more demanding test, however, produces the study's most revealing result. When lagged GDP growth and lagged investment growth are allowed to compete within the same model, previous GDP growth receives a posterior inclusion probability of 1.000, while the probability attached to lagged investment collapses to 0.238. Economic activity therefore appears to possess substantial inertia of its own, meaning weak investment partly moves with an already weak economy rather than independently predicting its future trajectory.

Such a result does not make investment unimportant. Instead, it prevents an overly simple policy reading in which raising capital expenditure is assumed to generate a proportional increase in growth. Investment and output are strongly pro-cyclical, gross fixed capital formation forms part of aggregate expenditure, and both respond to wider economic conditions, making causal claims difficult without a stronger identification strategy.

Why More Capital Spending May Still Produce Weak Growth

Gross fixed capital formation as a share of GDP averaged about 17.7% during the post-crisis, pre-COVID period, higher than the roughly 14.5% recorded before the global financial crisis. GDP growth nevertheless slowed dramatically, while investment growth itself stagnated, showing why investment ratios can give an incomplete picture of productive dynamism.

Capital can be substantial in accounting terms without delivering equivalent gains in productive capacity. Infrastructure delays, unreliable electricity, inefficient freight systems, regulatory uncertainty and weak public-sector implementation can lower expected returns, raise business costs and reduce the productivity of each unit invested. Under such conditions, increasing spending without improving the environment in which investment operates may simply reproduce low returns at a larger scale.

South Africa's case shifts the growth debate from capital quantity toward capital effectiveness. The study argues that electricity reliability, transport and logistics performance, infrastructure delivery, institutional capacity and policy certainty determine whether investment can translate into higher productivity and potential output. Weak investment is consequently both a symptom of structural constraints and a channel through which those constraints become embedded in slower growth.

The finding carries wider relevance for developing economies facing infrastructure deficits and tight public finances. Mobilising additional capital can attract political attention because investment targets are measurable, but headline spending provides limited information about project quality, execution capacity or economy-wide spillovers. Countries seeking faster growth may therefore need to evaluate not only how much capital they mobilise, but how effectively institutions, infrastructure networks and regulatory systems convert it into productive assets.

The Policy Test Is to Make Investment Productive Again

For South Africa, the authors argue against treating indiscriminate fiscal expansion as a growth strategy. Limited public resources should instead be directed toward infrastructure with strong economic spillovers, while existing capital budgets need to be executed more efficiently. Greater private participation, well-structured public-private partnerships and credible regulatory frameworks could help mobilise finance without placing the entire burden on the public balance sheet.

Electricity, freight rail, ports and municipal infrastructure emerge as particularly important because failures in these systems raise operating costs across the economy rather than within a single industry. Reforms involving major network institutions have significance beyond their own balance sheets. Their economic value ultimately depends on whether they reduce system-wide costs, restore reliability and encourage productive private investment.

Policy caution is equally important. The study uses a deliberately limited set of macroeconomic variables and does not directly model human capital, institutional quality, trade openness or every external channel affecting South Africa. Its long historical dataset also spans major structural and institutional changes, while Bayesian Model Averaging addresses uncertainty over model specifications without resolving endogeneity or establishing causal effects.

Forthcoming research will need sector-level investment data, explicit structural-break analysis and stronger causal methods to establish which investments produce the greatest growth returns and under what conditions. Cross-country analysis could also show whether South Africa represents an exceptional case or a broader pattern among emerging economies constrained by weak infrastructure and institutional capacity.