A reflection on property wealth and taxes in Cape Town

Property Wealth in Cape Town

Who or what makes a house worth more?

In a local context deeply entwined with a history of forced displacement, even the slightest inkling that long-standing communities can be priced out of their homes - whether by urban renewal, private development, or rising costs - can raise tensions almost instantly. Cape Town's recent experimentation with its municipal rates has proven to be yet another such instance.

The narrative across numerous neighbourhoods is that residents who have lived in their homes for generations suddenly find themselves holding (real estate) assets worth millions, thus falling into the City's higher property rates; but at the same time, their incomes have not kept pace, and the inability to pay their bills is forcing them out of these neighbourhoods that they call home (Davis, 2026; Naidoo, 2026). Civil society wants valuations capped, the City wants to protect its revenue, and the courts have struck down the current version of the policy (Kretzmann, 2026) - the debate now extends a conflict already years in the making.

But this fight rests on a deeper question: when a home doubles in value while its owner has changed nothing about it, who or what created that value? The answer matters for what a city can fairly tax, and from whom. We take this question to public data on Cape Town's property market.

Two sources of property value

A property's value has two components: the structure, i.e., the dwelling and whatever improvements the owner has invested in it, and the location. The value of a location is produced almost entirely by others: by public investment in roads, transport, and amenities, and by the presence and activity of other residents. The owner produces the value of the structure, while the surrounding city produces the value of the location. The argument goes back to Henry George (Brown, 1997; Wenzer, 2016): because location value is socially created, the public has a strong claim to tax it, and because land is immobile, taxing it introduces little distortion.

This argument has renewed relevance in the current (global) discourse on wealth taxation. Capital-based wealth taxes struggle with enforcement, because financial assets are mobile and can be restructured, relocated, or concealed (Alstadsæter et al., 2018; Saez, 2017; Zucman, 2019). Real estate cannot be hidden: a house has a cadastral entry, and even if it does not map to an individual, it is still “real” and taxable. Furthermore, for most households, across the income and wealth distribution, real estate is often their largest asset (Causa et al., 2019). South Africa is no exception: household wealth is among the most concentrated in the world, and property dominates the household asset portfolio across most of the wealth distribution (Chatterjee et al., 2022; Orthofer et al., 2019).

A tax on assessed property value is therefore arguably among the more enforceable ways to tax wealth. But its fairness relies on the split above: the public claim on location value is strong, since the public produced it, while the claim on the structure is weaker, since it reflects private investment. We therefore ask two questions of the data. How much of Cape Town's property appreciation is location rather than structure? And who is holding the gains?

What can the data tell us?

We draw on open and public data to explore these lines of inquiry. We begin with the City's valuation rolls, which report the median residential property value for each of Cape Town's 700-odd planning suburbs over four time points (2012, 2015, 2018, and 2022)--we compare between 2012 and 2022. In parallel, we use satellite-derived layers of building footprint and population - the Global Human Settlement Layer (Pesaresi, 2023) - over 2010 to 2020 to track how the built fabric of the city changes. All monetary values are reported in 2015 rands. The growth rates below are therefore real, i.e., adjusted for inflation, and are much smaller than the nominal increases visible on a rates bill.^[1]

Most of the appreciation in Cape Town is location value

How do we separate the appreciation due to structure from the appreciation due to location? We use two approaches. The first compares raw value with value per square metre, netting out the size of the dwelling to isolate the price of the location. The second investigates where private investment is directly visible, in suburbs where the dwellings themselves have grown.

The first approach divides each suburb’s median value by its median dwelling size. Value per square metre nets out dwelling size: a larger house raises a suburb’s median value but not its value per square metre, which, in turn, is driven mainly by the price of the location.

Raw value and value per square metre tell different stories. On raw value, inequality between suburbs narrowed over the decade: the Gini coefficient fell from 0.466 to 0.457. On value per square metre it moved the other way, rising from 0.231 to 0.242: location values are mildly diverging even as raw values converge. The priciest fifth of locations widened its lead over the cheapest from 3.1 to 3.2 times the value per square metre. Dwelling size reconciles the two: in the two cheapest fifths of suburbs in 2012, median dwelling size grew at about 1.6 percent a year, twice the pace of the rest, so much of the value growth at the bottom of the distribution came through larger dwellings rather than more expensive locations.

The ranking itself shows the same split (Table 1 and Figure 1). On raw value, the ladder does not change: 84.8 percent of suburbs sat in the same fifth of the distribution in 2022 as in 2012, and almost none moved far from their origin quintiles. On value per square metre, only 55.3 percent stayed put: location prices are re-sorting beneath a stable hierarchy of raw values. And the fastest repricing is at the bottom of the ladder, in cheaper suburbs. These patterns can co-exist: a minority of cheaper suburbs appears to leap upward in location value and re-sort the ranking, even as the gap between the top and bottom fifths, taken as groups, widens. This repricing at the bottom is a first identifier of the under-pressure list we turn to next.

Table 1 Movement and inequality across the value distribution, 2012 to 2022.

Figure 1 Quintile transition matrices, 2012 to 2022 (suburb counts, row percentages in brackets), for value (a) and value per m² (b). Diagonal cells count suburbs that stayed in their 2012 fifth.

The second approach asks where private investment is visible. Among suburbs that appreciated faster than the metro average, those where median dwelling size also grew faster than the metro's - the most direct sign of private investment in the structure - are a minority, and among the wealthiest suburbs rarer still, led by Baronetcy Estate, Boskloof, and Fresnaye. Prima facie, in both readings, most of Cape Town's appreciation appears to be growth in location value rather than a return on private investment.

Where the wealth is, and where the pressure is

If most of the appreciation is location value, who is holding it? We first ask where the accumulated wealth sits. 

Taking the top fifth of suburbs on both 2022 value and 2022 value per square metre gives 54 premium suburbs, which we call the existing-wealth list: the established high-value core of the Atlantic Seaboard and the Southern Suburbs, from Clifton and Bantry Bay to Camps Bay, Bishopscourt, and Fresnaye, with a mean value of R4.9 million.

The second (and more pertinent) geography is where rising values are accruing to residents whose capacity to pay may not have risen with them. Using the satellite data, we compare each suburb's growth in building footprint and in population with the metro trend to get a better understanding of how the built fabric has been changing (Broitman & Koomen, 2015). Our estimates using the GHSL data show that Cape Town has added population at 2.4 percent a year over the decade,^[2] against 1 percent for building footprint. Thirty-four suburbs combine real appreciation faster than the metro average with little new construction and little population change: these are suburbs that are being repriced without evidence of physical or demographic change. We call these the under-pressure list. 

How should this list be read against the rest of the city? Three comparisons are at work, each measured against the metro trend: growth in building footprint, growth in population, and growth in real value. The first two sort suburbs into four built-form regimes. Within each regime, we then ask where real values also grew faster than the metro, on both raw value and value per square metre. 

This repricing suggests a different meaning in each built-form regime: where buildings and population both grew faster than the city, and price increased faster than the city as well, it indicates demand-led repricing. Where buildings grew faster than people, this could be read as investment- or supply-led repricing. Where people crowded into a static building stock, it could be read as densification under pressure. And where values rose with neither new building nor new people, it may indicate gentrification or displacement pressure - the thirty-four suburbs that are of our primary interest. Table 2 summarises this discussion with illustrative examples; and Figure 2 presents a map of how these regimes are distributed spatially.

Table 2 Built-form regimes and repricing, 2012 to 2022. Suburbs with at least 50 dwellings and adequate satellite coverage (n = 568); repricing means real appreciation faster than the metro on both value and value per square metre.

Because rates bills follow valuations and rents follow prices, the under-pressure suburbs could be where the asset-rich, cash-poor squeeze is most likely to be felt by owners; where population is instead crowding into a broadly fixed stock, the same repricing may be felt through rents rather than rates. Going by contemporary discourses, could these be the suburbs where prices are driven by forces residents had no hand in, such as foreign buyers and short-term rentals (Savage, 2026)? The valuation roll carries no income data, so at the moment we can only say that this list indicates where the squeeze is most likely felt, rather than establishing that price-based displacement has occurred; linking incomes and tenure to these suburbs is our immediate next step.

Figure 2 Left, how each suburb changed physically over the decade, i.e., whether its building footprint and population grew faster or slower than the metro's; morphological stasis marks those where neither did. Right, the suburbs where real values also outran the metro. Suburbs that appreciated despite little physical or demographic change form the under-pressure list, shown in red in the right panel as possible gentrification / displacement. 

However, two observations are worth mentioning here. The suburbs most associated with the gentrification debate (such as Woodstock, Salt River, and Bo-Kaap) are gaining population on a broadly fixed stock, which at the suburb level could be read as crowding-in rather than emptying-out. That is a statement about aggregates, not about residents; displacement can proceed household by household without shifting a suburb-level median. Gentrification as commonly understood is compositional: it is a change in who lives in a suburb rather than how many, and it can proceed while the population grows. Reading repricing against built form is what narrows where to look for it. Rising population on a broadly fixed stock is consistent with substitution inside that stock, most plausibly through rentals and subdivision; repricing with neither new building nor new people points instead to owners selling. Resident incomes would confirm this difference, and that is the gap flagged above. The under-pressure suburbs instead appear to lean towards the Cape Flats and the periphery, from Vrygrond and Langa to Windermere and Wetton, rather than the City Bowl names the public discourse tends to assume.

The existing-wealth list (i.e. the top fifth on both 2022 value and value per square metre) and the under-pressure list (i.e. real appreciation above the metro with little new building and little population change) are nearly disjoint: exactly one suburb, Kalk Bay, appears on both. The under-pressure suburbs average R1.0 million in 2022 value; the existing-wealth suburbs average R4.9 million, nearly five times as much. And although the under-pressure suburbs appreciated almost twice as fast, at 3.2 percent a year against 1.7 percent, the existing-wealth suburbs gained more in absolute terms: roughly R72,400 a year in real terms against R22,900, since a slower rate on a sufficiently larger base compounds to a larger absolute gain. How much a suburb is already worth is a better guide to where wealth is accumulating than how fast its value is growing.

Figure 3 The existing-wealth suburbs and the under-pressure suburbs on one map, with each suburb's value and growth compared beneath. Only Kalk Bay appears on both lists; the map colours it red as a pressure suburb, so 53 of the 54 existing-wealth suburbs appear in gold.

Table 3 The two lists compared. Suburbs with at least 50 dwellings (n = 566); values are real, in 2015 rands.

What follows for possible tax design

These two geographies call for different treatment. Where wealth has matured and owners can pay, a wealth levy can apply now. Where value is still forming on cash-constrained incumbents, collection can be deferred until the asset changes hands - as circuit-breaker and deferral programmes in North America and elsewhere have done since the 1970s (Atria et al., 2025; Brunner et al., 2015; Carlson, 2004; Stine, 1988; Van Den Boogaard, 2025). Deferral costs the incumbent nothing while they occupy the dwelling, and the authorities collect the public's share when the gain is realised. Eligibility could also follow tenure rather than geography: a long-standing resident in a repricing suburb defers, while a recent buyer, who has just demonstrated the means to pay the current location premium, pays now.

Two Cape Town-specific observations follow. First, many currently under-pressure suburbs fall outside the tax net altogether: for example, Vrygrond, with a median value around R206,000, sits far below the City's rate-free threshold of roughly R620,000 and pays no rates. So the “squeeze” most likely concerns a mid-value subset of the list, not all of it. It is important that this subset is better identified.

Second, the City's struck-down design bundled rates with fixed charges for services, merging what could have been a wealth levy with a bill for public services. That design has its defenders, who argue that linking fixed service charges to property value was a rational and progressive cross-subsidy, unlawful only because the Municipal Systems Act requires charges to be proportional to consumption, and that the Act should be amended instead (Graham, 2026). Our reading of the data suggests a complementary proposition: a charge linked to property value is, in effect, a levy on property wealth, and the case for it might be easier to make (and to, perhaps, legislate in the current global climate) if it is argued as such, separately from the pricing of services.

Framed this way, a property-wealth tax need not threaten long-standing residents: deferral can protect households whose asset values have outrun their incomes, while the wealth levy could reach suburbs where publicly created value has already accumulated as private wealth and households are benefiting from this accumulated wealth. 

More granular data would improve the design: plot-level valuations, small-area census population, and household income microdata are the obvious next steps. The central finding, however, appears to be robust: at the moment, most of Cape Town's property appreciation is location value produced by the surrounding city, and most of it has accumulated in suburbs where property wealth is already established. Whether a particular owner can (or should) pay a levy on that wealth is a household-level question, not a suburb one. A long-tenured owner on a fixed pension may hold a multi-million-rand asset they could never afford to buy today, while their newest neighbour has just paid the full location premium. Distinguishing the two - by tenure first, and by income where the data allow - is what a deferral could be designed to do.

References

Alstadsæter, A., Johannesen, N., & Zucman, G. (2018). Who owns the wealth in tax havens? Macro evidence and implications for global inequality. Journal of Public Economics, 162, 89–100. https://doi.org/10.1016/j.jpubeco.2018.01.008

Atria, J., Contreras, D., & Méndez, M. L. (2025). Tackling wealth accumulation in a context of social upheaval: The property tax in Chile. Canadian Journal of Development Studies / Revue Canadienne d’études Du Développement, 46(2), 290–313. https://doi.org/10.1080/02255189.2023.2279119

Broitman, D., & Koomen, E. (2015). Residential density change: Densification and urban expansion. Computers, Environment and Urban Systems, 54, 32–46. https://doi.org/10.1016/j.compenvurbsys.2015.05.006

Brown, H. J. (Ed.). (1997). Land use & taxation: Applying the insights of Henry George. Lincoln Institute of Land Policy.

Brunner, E. J., Ross, S. L., & Simonsen, B. K. (2015). Homeowners, renters and the political economy of property taxation. Regional Science and Urban Economics, 53, 38–49. https://doi.org/10.1016/j.regsciurbeco.2015.04.001

Carlson, R. H. (2004). A Brief History of Property Tax. IAAO Conference on Assessment Administration, (February), 3–10.

Causa, O., Woloszko, N., & Leite, D. (2019). Housing, Wealth Accumulation and Wealth Distribution: Evidence and Stylized facts. OECD Economics Department Working Papers, 1588(1588). https://doi.org/10.1787/86954c10-en

Chatterjee, A., Czajka, L., & Gethin, A. (2022). Wealth Inequality in South Africa, 1993–2017. The World Bank Economic Review, 36(1), 19–36. https://doi.org/10.1093/wber/lhab012

Davis, R. (2026, February 22). How Cape Town’s municipal bills are outrunning its middle-class residents. Daily Maverick. https://www.dailymaverick.co.za/article/2026-02-22-how-cape-towns-municipal-bills-are-outrunning-its-middleclass-residents/

Graham, N. (2026, July 14). When right can be wrong: Analysis of the Cape Town tariff case. LinkedIn. https://www.linkedin.com/pulse/when-right-can-wrong-analysis-cape-town-tariff-case-nick-graham-ettdf

Kretzmann, S. (2026, April 30). Court declares Cape Town’s new fixed charges unconstitutional. GroundUp. https://groundup.org.za/article/court-declares-cape-towns-new-fixed-charges-unconstitutional

Naidoo, L. (2026, March 18). In Cape Town’s historic Bo-Kaap, homes under siege from rich foreign buyers. Al Jazeera. https://www.aljazeera.com/features/2026/3/18/in-cape-towns-historic-bo-kaap-homes-under-siege-from-rich-foreign-buyers

Orthofer, A., Du Plessis, S., & Reid, M. (2019). Private Wealth in a Developing Country: Evidence from South Africa. Review of Income and Wealth, 65(3), 632–656. https://doi.org/10.1111/roiw.12361

Pesaresi, M. (2023). GHS-BUILT-S R2023A - GHS built-up surface grid, derived from Sentinel2 composite and Landsat, multitemporal (1975-2030) [Data set]. European Commission, Joint Research Centre (JRC). https://doi.org/10.2905/9F06F36F-4B11-47EC-ABB0-4F8B7B1D72EA

Saez, E. (2017). Income and Wealth Inequality: Evidence and Policy Implications. Contemporary Economic Policy, 35(1), 7–25. https://doi.org/10.1111/coep.12210

Savage, R. (2026, March 24). Airbnb in firing line as Cape Town’s housing crisis catches up with middle class. The Guardian. https://www.theguardian.com/world/2026/mar/24/airbnb-cape-town-housing-crisis-middle-class

Stine, WilliamF. (1988). Estimating property tax base elasticity over time: Evidence on the revenue maximizing politician. Public Choice, 58(1), 35–44. https://doi.org/10.1007/BF00183327

Van Den Boogaard, V. (2025). Tax reform coalitions for urban development: The politics of property tax reform in Sierra Leone. Environment & Urbanization, 37(1), 95–115. https://doi.org/10.1177/09562478251317993

Wenzer, K. C. (2016). Land-Value Taxation (0 edn). Routledge. https://doi.org/10.4324/9781315501574

Zucman, G. (2019). Global Wealth Inequality. Annual Review of Economics, (11), 109–138. https://doi.org /10.1146/annurev-economics080218-025852