Housing wealth accounts for up to 65% of the average South African family's total wealth, yet its potential to reduce inequality is diminished by the persistent “stickiness” of apartheid geography. Using data from 2008–2017, the authors found that more than 90% of individuals from marginalized groups live in areas historically designated for their race, and even those who move to formal urban centers face returns that are deeply stratified by race. As the absolute gap in housing wealth between white and African urban residents continues to grow, research highlights the emergence of “spatial traps” – where state-provided housing provides physical shelter but fails as a wealth-building asset.
Housing accounts for between 40 and 65 percent of the average South African household's total wealth. But whether rising property prices help reduce or re-exacerbate inequality depends on the dynamics between housing markets in urban, rural, formal and informal sectors.
In our recent research, we examined how post-apartheid spatial mobility has shaped housing wealth. We asked whether South Africans are able to move out of historically marginalized slums, and whether such moves have led to meaningful gains in housing wealth. We found that spatial mobility remains limited, and when it occurs, housing wealth advantage is increasingly stratified by race.
housing wealth mobility
South Africa's extreme wealth inequality is well documented. Recent estimates indicate that the top 10 percent of South Africans held approximately 85 percent of total household wealth over the period 1994–2018, with the top 1 percent's share reaching 55 percent of the total in 2017 – one of the highest levels of wealth concentration globally.
The spatial distribution of wealth inequality in South Africa today is a legacy of its colonial and apartheid past. Laws such as the Native Land Act of 1913 and the Group Areas Act of 1950 systematically excluded black South Africans from owning land and property, particularly in economically productive urban centres.
Anecdotal evidence from two homes with nearly identical physical characteristics presented a starting point for our analysis: the change in home value in a historically white neighborhood over a ten-year period was at least twice as large as the change in home value in a historically Indian neighborhood over a fifty-year period.
A large literature has examined post-apartheid changes in income, employment and service access. Little is known about how spatial change – or the lack thereof – has affected housing wealth accumulation. Our research examined how moving from one historically defined settlement type to another affects changes in housing wealth over time, and how these processes vary by race.
In other words, who has been able to realize the wealth-accumulating potential of housing assets in post-apartheid South Africa?
Mobility and wealth tracking
Using panel data from the National Income Mobility Study (NIDS), we tracked individuals from 2008 to 2017 and classified residential areas based on their historical position within apartheid group areas. These settlement types are:
Tribal Authority Areas (TAAs), corresponding to former 'homelands' or Bantustans. They are largely African, rural and characterized by communal land ownership,1 and account for 27–32 percent of the NIDS sample.
Urban informal areas (UIAs), or informal settlements, typically on the urban periphery with insecure tenure. They are inhabited almost exclusively by African and colored communities and 8-9 percent of the sample live here.
the rural formal sector (RFA), which includes commercial farms; These are historically white-owned areas and account for 6-7 percent of the sample.
Urban formal areas (UFAs), the most economically desirable category, include cities, suburbs and formal townships. This is the largest and most racially mixed (though internally segregated) category, with 45–47 percent of the sample.
Within our framework, 'moving out' from a different location is defined as a person moving from a tribal authority area or urban informal sector to a rural or urban formal sector. It is important to note that we do not track activities within settlement types. Although there may be significant mobility within these categories – such as moving to a better located home within the urban formal sector – this specific framework does not capture such movement.
We used imputed rents as a proxy for housing wealth, which measures the value of housing services consumed by a homeowner – what they would pay to rent their home on the open market.
Finding 1: Spatial mobility is rare
We found strong evidence of spatial inertia. The majority of people from marginalized race groups – more than 90 percent – continue to live in areas that were designated for their race groups under apartheid.
This “stickiness” of apartheid geography is clearly demonstrated in Figure 1, which tracks mobility pathways over a decade. The densest flows show individuals remaining in their original settlement type. Nearly 90 percent of the residents of the urban formal area remained in the urban formal area, and importantly, an average of 94 percent of the residents of the tribal authority area remained in the tribal authority area. The 'exit' rate for those living in urban informal areas was also modest. This suggests that for most South Africans, particularly those living in historically isolated areas, spatial mobility remains the exception, not the rule. The 'stickiness' of apartheid-era geography acts as a powerful structural barrier, limiting access to opportunities and perpetuating poverty. The minorities who 'move out' from isolated places are highly selective groups. On average, they are younger, more likely to be male, and more economically active than those who remain. This implies that for those who are able to move, the positive outcomes are not only associated with the move, but also reflect the pre-existing human capital and low dependency ratio that enabled the move in the first place.
Figure 1: Flow of persons across settlement types
Each colored stream traces the settlement trajectories of individuals across survey waves. Most individuals remain in the same settlement category over time, while narrow branches indicate transitions from one category to another. Source: NIDS Waves 1-5; writers analysis
Finding 2: 'Exiting isolation' does not mean 'entering integration'
For individuals who move to urban formal areas, the average rent (and hence housing wealth) increases significantly. However, this process of wealth creation is neither neutral nor equitable. Economic returns to spatial mobility are deeply and persistently stratified by race.
The first step in understanding this is to differentiate the urban formal sector. The data shows that UFA is not a single, unified market but rather a collection of racially and economically distinct sub-markets. Figure 2 shows that the 'UFA' unit means something different depending on the ethnicity of the residents.
Figure 2: Average imputed rent by population groups and settlement types
Source: NIDS Waves 1-5; writers analysis
In Wave 1 (2008), the average estimated rent for white-occupied dwellings in the urban formal sector (R2 740) was almost six times higher than for African-occupied dwellings in the same settlement type (R461). By Wave 5 (2017), while both groups saw increases, the absolute gap had widened dramatically: the average white imputed rent reached R9 424, while the average African imputed rent was R1 627. This suggests that the urban formal sector is, in all likelihood, an analytical simplification, possibly concealing a deeply segregated internal geography where property values are determined by race. While both groups saw growth over the past decade, the fact that the absolute gap between them has increased dramatically indicates that different racial groups continue to operate in largely separate housing submarkets. On the other hand, the relative growth of African-occupied dwellings was greater than that of white-occupied dwellings, suggesting a slight convergence in growth rates despite the widening of the absolute wealth divide.
This stratification is not merely stable; It also defines the benefits to be gained from moving forward. Analysis of mobility pathways (Figure 3) shows that the very act of 'moving out' of segregation yields racially differentiated returns.
Figure 3: Changes in fares charged on routes by population groups and settlement types
Source: NIDS Waves 1-5; writers analysis
UIA