The Geography of Stagnation: Spatial Divergence and the Low-Wage Formalization Trap in South Africa, 2014–2024

This paper investigates whether South Africa’s nine provinces are experiencing economic convergence or divergence over the period 2014 to 2024, drawing on novel administrative data from the South African Revenue Service (SARS) merged with Statistics South Africa’s Quarterly Labour Force Survey (QLFS). Despite three decades of democratic governance and redistributive fiscal policy, the spatial structure of the South African economy remains profoundly unequal. Gauteng and the Western Cape concentrate the majority of formal economic activity, while provinces such as the Eastern Cape, Limpopo, and the North West languish in persistent structural poverty. The central question this paper addresses is whether market forces and existing policy frameworks are generating a natural catch-up process among provinces, or whether the spatial divide is permanently widening. Employing a multi-method convergence framework, the analysis proceeds through sigma-convergence, absolute and conditional beta-convergence, and panel fixed-effects estimation across provincial, district, and municipal levels. The sigma-convergence analysis reveals a 25 per cent increase in income dispersion across provinces (coefficient of variation rising from 0.136 to 0.170), decisively rejecting the hypothesis of narrowing spatial inequality. Cross-sectional beta-convergence tests yield a positive coefficient for GDP (beta = +0.012), indicating divergence, though the result is not statistically significant due to the small sample of nine provinces. Panel fixed-effects models show unconditional convergence (β = 0.342 for GDP, β = 0.767 for income), and conditional models with structural controls yield qualitatively similar convergence signals (β = 0.391 for GDP, β = 0.712 for income). However, robustness checks reveal this apparent convergence to be mechanical mean reversion that vanishes at a five-year lag horizon (β = 0.975, p = 0.905 for GDP). The integration of QLFS data demonstrates that a one percentage point increase in provincial unemployment decreases real GDP by 4.7 per cent (p = 0.036), confirming unemployment as an econometrically verified structural barrier to convergence. The district-level analysis reveals a devastating new phenomenon: apparent convergence is a statistical mirage driven by the collapse of formerly higher-income districts, not the rise of poor ones. In the Eastern Cape, metropolitan areas have lower median incomes than surrounding rural districts, destroying the assumption that provincial metros serve as engines of convergence. At the municipal level, the paper identifies a “Low-Wage Formalization Trap” where formal employment (FTE) grows while median income collapses and inequality rises, fundamentally undermining the national “job creation” KPI. The trap is most vividly illustrated by Mnquma municipality, where an 18 per cent increase in FTE employment was accompanied by a 52.2 per cent collapse in median income over the same period. The policy implications are significant and demand urgent attention. Spatial trickle-down economics has failed, and the Government of South Africa requires forced economic decentralisation, hyper-SEZs with zero corporate taxation, geofenced employment tax incentives, and sub-provincial infrastructure decoupling to arrest the deepening spatial crisis. The paper further recommends reforming the national job creation KPI to incorporate employment quality metrics, addressing the Eastern Cape metropolitan anomaly through targeted industrial policy, and investigating the promising growth trajectories observed in the Overberg and iLembe districts for potential replication. These findings contribute to the academic literature on convergence in developing countries and provide actionable evidence for the Government of the Republic of South Africa.