PALI LEHOHLA | The intolerable cultural economic geography of apartheid

The current configuration of South Africa’s spatial economy is intolerable.

It is a Russian roulette with a 97% certainty that the poor are born to lose and, like the sun rises in the east and sets in the west, not one day will a miracle happen.

Andrés Rodríguez-Pose, an institutional economic geographer whose research frequently addresses themes of culture, trust and shared norms to explain why certain regions flourish while others are stuck in a “development trap”, provides an apt practice and theoretical frame that explains why flares of political and social discontent are perennial in South Africa.

While this spatial dysfunction displays remarkable inertia and persistence — a testament to its structural robustness — it has become a source of profound economic, social and political dysfunction. The poor are therefore like flies trapped in an opened Consol bottle where they are destined to stay, lose and die.

The geography of apartheid, originally designed for racial separation and economic extraction, has transitioned seamlessly into a modern spatial economy characterised by extreme poverty density and economic exclusion. The central question facing the nation, especially ahead of any local government election cycle, is whether these statistical and spatial patterns are immutable, or if the localised governance structures possess the numeric and financial instruments required to disrupt them.

This is where Morena Mohlomi’s adage and wisdom on new instruments of power is apt. But have we these new instruments of power?

The new tools of power are important for us to understand why this persistence is intolerable. We must look past superficial symptoms such as service delivery protests and employ a rigorous geostatistical analysis.

The Lehohla Ledger has successfully created these new instruments of power and they are based on and drive integrated reporting as an essential property of Mohlomi’s Intergenerational Value Creation.

This is where Sekodi in Daveyton remains true to its name — one US President Donald Trump appropriately calls a “sh*t hole”. For it is so by design.

The new tools of power are important for us to understand why this persistence is intolerable. We must look past superficial symptoms such as service delivery protests and employ a rigorous geostatistical analysis. This necessitates utilising instruments such as Anselin’s Local Indicators of Spatial Association (LISA) and the Global Moran’s I statistic.

A LISA analysis, when applied to a municipality like Ekurhuleni, reveals a landscape of intense spatial autocorrelation. This is defined by a statistically significant clustering of socioeconomic attributes.

For a 14-year-old student like Thabo, growing up in grade 8 in Sekoti Phola, Daveyton, the LISA map does not show a neighbourhood; it defines a mathematical certainty. Sekoti Phola is located within a “High-High” cluster of multi-dimensional poverty and youth unemployment. This means Thabo is surrounded by thousands of others in the exact same economic position.

Given this geostatistical reality, can the local government elections fix it? The conventional rhetoric suggests yes; better service delivery and governance should lead to local economic growth. However, a rigorous analysis of the numerical constraints suggests otherwise.

This positive spatial autocorrelation creates an intolerable functional constraint. It acts as a mechanical determinant of his future development trajectory. In a High-High cluster, the social capital necessary for standard job referrals is mathematically non-existent.

Furthermore, the distance from these clusters to primary economic hubs — the periphery-to-core gap — imposes a severe, often impassable, “tax” on work-seeking behaviour and labour market participation. The density of unemployment in these clusters (often reaching 6,000 to 8,000 individuals per square kilometre) overwhelms any localised market density, trapping emerging human capital in a cycle of marginalisation.

Given this geostatistical reality, can the local government elections fix it? The conventional rhetoric suggests yes; better service delivery and governance should lead to local economic growth. However, a rigorous analysis of the numerical constraints suggests otherwise.

The power to break these High-High clusters does not reside at the municipal level, principally because of how municipal infrastructure is funded.

The current municipal infrastructure grant framework and localised property tax systems create a functional paradox. The highest need for spatial restructuring exists within these dense High-High poverty clusters. Yet, the property value within these areas is, by definition, low, providing minimal rate-paying income. Municipalities, therefore, lack the internally generated revenue required to fund massive spatial restructuring.

The state’s funding of infrastructure often follows, rather than leads, economic activity. Investments flow into corridors with existing economic density to protect current GDP contributions, further reinforcing the established spatial regime.

Local elections change the administrators of the poverty, but they rarely change the numeric capability of the ward to fund its own escape from the cluster. A better-run council can fix the water pipes, but it cannot move the township 5km closer to the economic core or reverse the historical spatial planning that guarantees peripheral isolation.

The worst-case scenario: the decentralisation of precarity

If the established spatial trajectory continues unchecked, the worst-case scenario is not a sudden collapse but a creeping, formalised stagnation.

The East Rand census mesh, when analysed over longitudinal decades (2001 to 2022), shows that these High-High poverty clusters are expanding at the edges while simultaneously densifying in the core.

In this scenario, local government elections fail completely, not due to lack of trying, but due to numerical irrelevance.

Municipal infrastructure funding remains centralised at the National Treasury, and grants are allocated using formulas that prioritise current population counts rather than future-focused spatial restructuring.

High-value economic nodes in formerly affluent suburbs continue to attract investment, increasing land surface temperatures (LST) and building density (NDVI deficits) that further degrade the health of labourers who commute to them.

The spatial lag model, which defines economic potential based on the outcome of neighbours, solidifies. In a pure spatial trap, the coefficient of neighbouring influence approaches 1.0, meaning individual characteristics like Thabo’s education level are completely dominated by the surrounding High-High context.

This scenario creates an unsustainable pressure valve where entire peripheral wards become geographically distinct reservoirs of untapped human capital, perpetually excluded from the core, with no democratic avenue to alter the math of their existence. The spatial persistence becomes, quite literally, the statistical formalisation of hopelessness.

The South African spatial economy is intolerable because it has turned geographical coordinates into a powerful, multi-generational predictor of individual life chances.

Using instruments like LISA mapping, we can visualise that Thabo’s fate in Daveyton is tied to the outcomes of his neighbours in a way that should be unacceptable in any equitable society. Local government elections, as currently constructed and funded, do not possess the necessary numeric instruments to dissolve these High-High clusters.

Breaking this geostatistical trap requires more than just better localised governance; it demands a radical, centralised restructuring of the nation’s entire census mesh and infrastructure funding model. Without it, the spatial autocorrelation of poverty will remain the defining, intolerable arithmetic of our economic landscape.

Dr Pali Lehohla is a Professor of Practice at the University of Johannesburg, a Research Associate at Oxford University, and a distinguished Alumni of the University of Ghana. He is the former Statistician-General of South Africa.