South Africa’s real income per person rose by about 42 percent between 1961 and 2025, and almost all of that gain came in two climbs, 1961 to 1981 and 1993 to 2013, each followed by a fall, so that in 2025 income per person stood about 4 percent above its 1981 level and about 7 percent below its 2013 peak. [IC]
Overview
This brief gives a snapshot of South Africa’s development over 65 years, from the declaration of the Republic on 31 May 1961 [IC] to October 2026. The period spans three governing arrangements: the National Party (1961 to 1994), the African National Congress (1994 to 2024) and the Government of National Unity, whose statement of intent was published on 14 June 2024 [IC]. The brief reads the record through the wiki’s two lenses on growth. The S-Curve Thesis and the Retention Counter-Tradition places economies on a growth curve with an acceleration phase and a deceleration phase. What GDP Leaves Out sets the GDP figure beside the conditions households live in. The purpose here is to show the shape of the arc. A companion research run tested how much of the slowdown since 2008 is structural and how much traces to governance; its main findings appear in one section below.
The shape is two climbs and two descents.
- Real GDP per person rose from US$4,032 in 1961 to US$5,499 in 1981 (constant 2015 US$),
- fell to US$4,194 in 1993,
- rose to a peak of US$6,171 in 2013, and
- stood at US$5,713 in 2025 (World Bank World Development Indicators) [IC].
Over the same span the population grew from 16.9 million to 64.7 million [IC]. Total output grew about 2.7 percent a year across the 65 years and output per person about 0.55 percent a year [IC]. A household in 2025 had, on average, roughly the real income per person of a household in 1981.
Other curves run beneath the GDP line, and several of them turned earlier.
- Gold output peaked in the early 1970s above 1,000 tonnes a year [IC].
- Fixed investment peaked at 29.2 percent of GDP in 1976 and was 13.9 percent in 2025 [IC].
- Manufacturing peaked at 24.9 percent of GDP in 1981 and was 12.3 percent in 2025 [IC].
- Life expectancy fell by about nine years between 1991 and 2005 and then recovered by about twelve [IC].
- Government debt fell to 26.0 percent of GDP in 2008/09 and rose to 78.9 percent in 2025/26 [IC].
- On some household measures South Africa did worse than its GDP; on life expectancy it did far better.
The brief is written for practitioners reading the arc of their own economy. The method transfers: set GDP per person beside the resource base, investment, health, jobs, trust and public debt over the longest span the data allow, and mark each change of government on the same timeline.
The arc in one table
| Indicator | 1961 | 1970 | 1981 | 1993 | 2007 | 2013 | 2019 | 2024 |
|---|---|---|---|---|---|---|---|---|
| Real GDP per person (constant 2015 US$) | 4,032 | 5,136 | 5,499 | 4,194 | 5,891 | 6,171 | 6,033 | 5,715 |
| Population (millions) | 16.9 | 22.3 | 30.5 | 43.3 | 50.5 | 54.7 | 59.6 | 64.0 |
| Gross fixed capital formation (% of GDP) | 16.8 | 22.4 | 27.3 | 14.7 | 19.0 | 18.6 | 15.5 | 14.5 |
| Manufacturing value added (% of GDP) | 20.5 | 22.5 | 24.9 | 21.4 | 15.4 | 12.4 | 12.4 | 12.8 |
| Natural resource rents (% of GDP) | n.a. | 1.7 | 11.0 | 4.5 | 6.4 | 6.1 | 4.2 | 7.3 (2021) |
| Life expectancy at birth (years) | 53.1 | 55.7 | 60.4 | 62.6 | 54.9 | 62.4 | 66.1 | 66.3 |
| Total fertility (births per woman) | 6.08 | 5.63 | 4.71 | 3.37 | 2.55 | 2.43 | 2.26 | 2.21 |
| Unemployment, ILO modeled estimate (%) | n.a. | n.a. | n.a. | 23.4 | 22.3 | 24.6 | 28.5 | 32.3 |
Source: World Bank, World Development Indicators, indicators NY.GDP.PCAP.KD, SP.POP.TOTL, NE.GDI.FTOT.ZS, NV.IND.MANF.ZS, NY.GDP.TOTL.RT.ZS, SP.DYN.LE00.IN, SP.DYN.TFRT.IN and SL.UEM.TOTL.ZS, retrieved 22 October 2026 (series last updated between 13 July and 8 October 2026) [IC]. The years mark the start of the Republic, the gold peak, the first income peak, the last full year of National Party government, the eve of the 2008 crisis, the second income peak, the year before COVID-19, and the latest full year.
Notes on the series. The World Bank series cover South Africa within its present borders, back-cast for the years before 1994; apartheid-era national accounts reported the nominally independent homelands separately from 1976, so older official series and later reconstructions can differ [TBV]. The population series is a United Nations estimate; Statistics South Africa’s own mid-2025 estimate is 63.1 million, against 64.7 million in the World Bank series for 2025, a difference of method and vintage [IC]. The ILO modeled unemployment rate begins in 1991 and runs below Statistics South Africa’s official rate, which was 33.6 percent in the second quarter of 2026 [IC].
The six periods
| Period | Governing arrangement | Real GDP growth a year | GDP per person a year | Dated markers |
|---|---|---|---|---|
| 1961 to 1973 | National Party | 5.3% | +2.2% | Republic declared 31 May 1961 [IC]; gold output peaks in the early 1970s above 1,000 t a year, more than 70% of world production [IC] |
| 1974 to 1993 | National Party | 1.8% | −1.1% | Income per person peaks in 1981 and falls 24% by 1993 [IC]; fixed investment falls from 27.3% of GDP (1981) to 14.7% (1993) [IC]; IMF staff estimate potential growth “slightly below 1 percent” in 1985 to 1994 [IC] |
| 1994 to 2007 | ANC | 3.6% | +2.5% | Energy White Paper (December 1998) forecasts the generation surplus “fully utilised by about 2007” [IC]; government debt falls from 48.2% of GDP (1995/96) to 26.0% (2008/09) [IC]; life expectancy falls to 53.9 years in 2005 [IC] |
| 2008 to 2019 | ANC | 1.6% | +0.2% | Load-shedding recorded in 2007 and 2008 and in most years from 2014 [IC]; potential growth falls from 3.9% before the 2008 crisis to 1.4% after [IC]; income per person peaks in 2013 [IC]; “state capture” era dated 2009 to 2018 by IMF staff [IC] |
| 2020 to May 2024 | ANC | 0.3% | −1.2% | GDP −6.2% in 2020 [IC]; July 2021 unrest, an estimated 354 dead and over R50 billion lost [IC]; grey-listed by the Financial Action Task Force, February 2023 [IC]; load-shedding at record levels in 2023 [IC]; ended from April 2024 [IC] |
| June 2024 to October 2026 | Government of National Unity | 0.8% (2024 to 2025) | −0.4% | Grey-list exit 24 October 2025 [IC]; S&P upgrade 14 November 2025, the first by a major agency in over 16 years [IC]; gross loan debt 78.9% of GDP in 2025/26 [IC]; GDP −0.2% quarter on quarter in the second quarter of 2026 [IC] |
Growth rates are geometric averages of the World Bank annual series over each period; per-person rates use the endpoints of the GDP per person series (1961 to 1973, 1973 to 1993, 1993 to 2007, 2007 to 2019, 2019 to 2023, 2023 to 2025) [IC]. The IMF’s own figure for 1994 to 2007 is the same, 3.6 percent a year [IC].
1961 to 1973: the gold-and-manufacturing climb
GDP grew 5.3 percent a year and output per person 2.2 percent a year [IC]. Gold output rose to its peak in the early 1970s, when South African mines supplied more than 70 percent of world production (LBMA 2023) [IC]. Manufacturing rose from 20.5 to 22.5 percent of GDP and fixed investment from 16.8 to 22.4 percent [IC]. Seen against other countries, the climb was modest: Charles Feinstein’s thirty-country comparison, as reported by Nattrass and Seekings (2010), found that the comparison group’s per-person growth over 1950 to 1973 was “almost double South Africa’s” [IC]. The gains of this period were built on the apartheid labor and residential order and went chiefly to the white minority [TBV for distribution figures, which this brief did not compile].
1974 to 1993: the first descent
Growth slowed to 1.8 percent a year and output per person fell 1.1 percent a year [IC]. Investment peaked at 29.2 percent of GDP in 1976 and fell to 14.7 percent by 1993 [IC]. Income per person reached its first peak in 1981, helped by a gold price that lifted natural resource rents to 13.5 percent of GDP in 1980, and then fell by almost a quarter over twelve years [IC]. IMF staff estimate that potential growth in 1985 to 1994 was “very modest, at slightly below 1 percent,” which they attribute to “the high economic costs associated with isolation of the apartheid regime” (Klein 2011) [IC]. Life expectancy kept rising through this period, to about 63 years at the start of the 1990s [IC]. Government debt rose from 32.4 percent of GDP in 1990 to 44.8 percent in 1994 in the World Bank series [IC].
1994 to 2007: the second climb
GDP grew 3.6 percent a year and output per person 2.5 percent a year [IC]. Government debt fell from a peak of 48.2 percent of GDP in 1995/96 to 26.0 percent in 2008/09 (SARB 2019) [IC]. Manufacturing fell from 21.4 to 15.4 percent of GDP as the economy opened to trade [IC]. Two curves moved against the GDP line. Life expectancy fell from 62.6 years in 1993 to 53.9 years in 2005 as HIV/AIDS mortality rose [IC]. The December 1998 White Paper on the Energy Policy of the Republic of South Africa stated that “Eskom’s present generation capacity surplus will be fully utilised by about 2007” (DME 1998) [IC]. The boom years of 2004 to 2007, with growth between 4.6 and 5.6 percent a year, ran while that surplus was being used up [IC].
2008 to 2019: the plateau
GDP grew 1.6 percent a year and output per person 0.2 percent a year [IC]. Income per person reached its second peak in 2013 and then declined [IC]. Load-shedding, the scheduled rotation of power cuts, shed 176 GWh in 2007 and 476 GWh in 2008, ran 121 hours in 2014 and 852 hours in 2015, and returned at 530 hours in 2019 (CSIR data, compiled by IOL 2024) [IC]. A South African Reserve Bank staff note estimates that potential growth averaged 3.9 percent before the 2008 crisis and 1.4 percent after, and concludes that “most of the post-GFC growth slowdown can be attributed to domestic factors” (Janse van Rensburg and Morema 2023) [IC]. Government debt rose from 26.0 percent of GDP in 2008/09 [IC]. Unemployment on the ILO modeled estimate rose from 22.4 percent in 2008 to 28.5 percent in 2019 [IC].
2020 to May 2024: shocks and the low point
GDP grew 0.3 percent a year and output per person fell 1.2 percent a year [IC]. GDP fell 6.2 percent in 2020 with the COVID-19 lockdowns [IC], and life expectancy dropped from 66.1 years in 2019 to 62.0 in 2021 before recovering [IC]. The Expert Panel on the July 2021 unrest reported an estimated 354 dead and over R50 billion lost to the economy (Expert Panel 2022) [IC]. Load-shedding ran 1,169 hours in 2021 and 2,773 hours in 2022 [IC] and was imposed on more than 330 days of 2023 [IC]. The Financial Action Task Force grey-listed South Africa in February 2023 [IC]. Eskom reports that its generation recovery plan began in March 2023 and that uninterrupted supply began on 26 March 2024 (Eskom 2024) [TBV as Eskom’s own account]; the CSIR records that falling demand and a rising plant availability factor “eliminated loadshedding from April 2024 onwards” (CSIR 2025) [IC].
June 2024 to October 2026: recovery markers on a flat line
GDP grew 0.5 percent in 2024 and 1.1 percent in 2025, and output per person was flat [IC]. Brief load-shedding returned in early 2025, including Stage 2 on 19 March 2025 after the loss of five generating units (Engineering News 2025) [IC]. South Africa left the grey list on 24 October 2025; National Treasury described the work as recovery from “the weakening of key law enforcement and other institutions during the state capture era” (National Treasury 2025a) [IC]. S&P raised the foreign currency rating from BB- to BB on 14 November 2025, citing “a reduction in contingent liabilities tied to performance improvements” at Eskom (National Treasury 2025b) [IC]. Gross loan debt reached 78.9 percent of GDP in 2025/26, and debt-service costs took 21.3 percent of revenue (National Treasury 2026) [IC]. GDP contracted 0.2 percent in the second quarter of 2026, with mining, manufacturing and trade down (SAnews 2026) [IC]. Official unemployment was 33.6 percent and the expanded rate, which adds the potential labor force, 43.8 percent (Statistics South Africa 2026) [IC]. The two recovery markers of this period, power supply and the grey-list exit, trace to work begun in 2023 [IC].
Lens A: South Africa on its own S-curve
The mineral curve
The clearest S-curve in the record is gold. Output rose through the 1960s, peaked in the early 1970s above 1,000 tonnes a year, fell to 93 tonnes in 2022 (LBMA 2023) and stood at about 90 tonnes a year in 2026 (Miningmx 2026) [IC]. The decline is about 90 percent and it began more than twenty years before 1994 [IC]. It continued under every government in the period, and record gold prices in 2025 and 2026 did not reverse it [IC].
Fine and Rustomjee (1996) describe the core of the South African economy as a minerals-energy complex: mining, cheap coal-fired electricity and the finance and heavy industry organized around them [IC]. In the wiki’s vocabulary this is South Africa’s throughput economics: value measured by the tonnage and energy that move through the economy. The complex persisted across 1961, 1994 and 2024 [IC]. Its two advantages eroded on different clocks. Gold depletion is geological and long. The electricity shortfall arrived on the date the 1998 White Paper forecast, and the Harvard Growth Lab finds that state failure “squandered the country’s comparative advantage in cheap, coal-fired electricity” (Hausmann et al. 2023, not peer-reviewed) [IC]. The Growth Lab also finds that “the weakening of mining began before 2008,” while global commodity prices remained strong for several years after [IC].
Reinert (2007) distinguishes increasing-returns activities such as manufacturing from diminishing-returns activities such as raw-material extraction. South Africa’s manufacturing share peaked in 1981 and halved by 2013 [IC], leaving an economy weighted toward activities whose returns fall as the best ore is used up. Natural resource rents swing with prices: 13.5 percent of GDP in 1980, 1.7 percent in 1970, 12.0 percent in 2008 [IC].
Retention inside an exclusionary order
The S-curve brief treats tariffs, state investment and directed credit as the tools of retention economics, used by countries building a productive base. South Africa used them heavily before 1994, through import substitution, state corporations such as Iscor and Sasol, the Industrial Development Corporation, and sanctions-era self-sufficiency [IC]. Manufacturing rose to a quarter of GDP by 1981 [IC]. The same order excluded most of the population from property ownership and from skilled work in the cities, and its per-person growth from 1950 to 1973 was about half that of Feinstein’s comparison group [IC]. After 1994 the economy opened to trade and the manufacturing share fell [IC]. The return to industrial policy and localization after about 2010 coincides with the weakest growth in the record; no published estimate isolating its effect was found [TBV].
The South African case complicates the corpus account of retention. Retention built a productive base for a minority and raised a manufacturing share that later halved. The record fits a narrower claim: retention works where the state can administer it and where its gains reach the population that sustains it [structural inference, flagged]. The colonial roots of the labor system are traced in Sound Finance, which quotes the 1893 Mine Managers Association report on using the hut tax to draw labor to the mines.
Deceleration indicators
The corpus lists the indicators of a decelerating curve in Deceleration Indicators. Applied to South Africa:
| Indicator | South African reading | Grade |
|---|---|---|
| Government debt | 48.2% of GDP (1995/96), 26.0% (2008/09), 78.9% (2025/26); projected 76.5% by 2028/29 | [IC] |
| Debt service | 5.4% of GDP and 21.3% of revenue in 2025/26 | [IC] |
| Investment | 29.2% of GDP (1976), 21.6% (2008), 13.9% (2025) | [IC] |
| Fertility | 6.1 births per woman (1961), 2.2 (2024) | [IC] |
| Diminishing returns to complexity | IMF staff link the fall in total factor productivity after 2008 to “high levels of bureaucracy, red tape, governance deficiencies and corruption during the ‘state capture’ era (2009-2018), and inefficient state-owned enterprises” | [IC] for the finding; reading it as [[sources/tainter-collapse-complex-societies |
| Wage-productivity gap, household debt, water stress | Not compiled for this brief | [TBV] |
One reading departs from the corpus. The S-curve brief describes deceleration as a late phase of mature economies. South Africa’s income per person stalled at about US$6,000 in constant 2015 dollars, a level the mature economies passed generations earlier [IC]. If South Africa is decelerating, it is decelerating at middle income [structural inference, flagged].
Turchin’s pressures
The Three Futures cites Peter Turchin’s structural-demographic triggers: popular immiseration, elite overproduction and state fiscal distress (Turchin 2016). Two can be read in South African data. Unemployment rose from about 20 percent in 1994 to 33.5 percent in 2022 (Hausmann et al. 2023) and stood at 33.6 percent in mid-2026 [IC]. Public debt roughly tripled as a share of GDP after 2008/09 [IC]. Against immiseration, the World Bank reports that the poverty rate fell from 46.7 percent in 2015 to 37.9 percent in 2023, while close to 60 percent of South Africans still live below the upper-middle-income poverty line (World Bank 2026) [IC]. No data on elite overproduction were compiled [TBV]. This is the most interpretive reading in Lens A [working hypothesis].
Lens B: what the gauge recorded and what households recorded
Household measures beside GDP
| Measure | Early 1990s | Around 2008 | Around 2014 | 2019 | Latest |
|---|---|---|---|---|---|
| Real GDP per person (constant 2015 US$) | 4,194 (1993) | 6,010 (2008) | 6,155 (2014) | 6,033 | 5,713 (2025) |
| Life expectancy (years) | 62.6 (1993) | 56.0 (2008) | 63.2 (2014) | 66.1 | 66.3 (2024); 66.9 (Stats SA, 2025) |
| Unemployment, ILO modeled (%) | 23.4 (1993) | 22.4 (2008) | 24.9 (2014) | 28.5 | 32.4 (2025) |
| Unemployment, official and expanded (%) | n.a. | n.a. | n.a. | n.a. | 33.6 and 43.8 (Q2 2026) |
| Gini index (consumption or income) | 59.3 (1993) | 62.5 (2008); 65.0 (2005) | 59.6 (2014) | n.a. | 54.1 (2022) |
| Load-shedding | None recorded | 476 GWh shed (2008) | 121 hours (2014) | 530 hours | 2,773 hours (2022); more than 330 days (2023); none from April 2024 to year end |
| Government debt (% of GDP) | 48.2 (1995/96) | 26.0 (2008/09) | n.a. | n.a. | 78.9 (2025/26) |
Sources: World Bank WDI (GDP, life expectancy, ILO unemployment, Gini from survey data); Statistics South Africa mid-year estimates 2025 and QLFS Q2 2026; CSIR data via IOL and CSIR 2025; SARB 2019; National Treasury 2026 [IC]. The Gini values come from household surveys whose design changed over the period, so year-to-year comparisons carry caveats [TBV for comparability].
The South African reading runs partly opposite to the mature-economy pattern in What GDP Leaves Out. There, GDP rose while household measures fell. Here, GDP per person stalled after 2013 while life expectancy rose by about twelve years from its 2005 low [IC]. Statistics South Africa estimates 8.15 million people living with HIV in 2025 [IC]; the recovery in life expectancy coincides with the expansion of HIV treatment [TBV as a causal claim]. Employment and public trust moved with GDP or worse. Trust in the president fell 34 percentage points between 2011/2013 and 2021/2023, the fourth-largest fall among 30 countries surveyed in both periods (Afrobarometer 2024) [IC]. The gauge missed a household gain in health and recorded little of the losses in work and public trust.
Defensive spending
Households and firms buy what public systems fail to supply, and all of it counts as output. At the end of March 2024 South Africa had 608,977 registered active private security officers (Portfolio Committee on Police 2024) [IC], several times the staff of the national police service [TBV for the police figure]. Rooftop solar capacity reached an estimated 7.3 GW by August 2025, more than the renewable capacity contracted to Eskom from independent producers (Financial Gazette 2025) [TBV, estimator not named in the report]. The Growth Lab finds that the few jobs created after 2008 “tended to be in security, household services” [IC]. This is the mechanism Leipert (1989) describes. No estimate of its contribution to measured growth in the worst load-shedding years was found; it likely softened the measured contraction by an unknown amount [TBV].
Disaster and repair
The July 2021 unrest is the costliest single event in the recent record: an estimated 354 dead and over R50 billion lost [IC]. National Treasury estimated looting-related insurance claims at R32 billion and allocated R22 billion to Sasria, the state-owned insurer for riot risk, to meet them (IOL 2022) [IC]. Under national accounting convention, the rebuilding counts as new output and the destroyed stock is never subtracted (BEA) [IC].
Worked example. Storms in May 2026 left the Western Cape with an estimated R9.1 billion in damage, 231,029 people affected, 22,890 houses damaged and agricultural losses above R5.2 billion, according to the provincial cabinet (TimesLIVE 2026) [IC]. The repair will appear in GDP as growth. Place-level treatment of these floods belongs on the BioHubs directory.
South Africa beside comparable economies
| Economy | 2008 | 2013 | 2024 | 2025 | Change 2008 to 2025 |
|---|---|---|---|---|---|
| South Africa | 6,010 | 6,171 | 5,715 | 5,713 | −4.9% |
| Namibia | 4,036 | 4,549 | 4,032 | 4,019 | −0.4% |
| Mexico | 9,826 | 9,739 | 10,284 | 10,257 | +4.4% |
| Botswana | 6,159 | 6,285 | 6,952 | 6,791 | +10.3% |
| Brazil | 8,331 | 9,367 | 9,567 | 9,748 | +17.0% |
| Chile | 11,356 | 13,183 | 14,617 | 14,905 | +31.2% |
| Colombia | 5,049 | 5,911 | 6,865 | 6,976 | +38.2% |
| Malaysia | 7,976 | 8,965 | 11,884 | 12,352 | +54.9% |
| Indonesia | 2,465 | 3,055 | 4,368 | 4,555 | +84.8% |
| Türkiye | 8,448 | 10,269 | 15,395 | 15,883 | +88.0% |
Real GDP per person, constant 2015 US$. Source: World Bank WDI, NY.GDP.PCAP.KD, retrieved 22 October 2026; changes computed for this brief [IC]. The peer group is upper-middle-income commodity exporters and two regional neighbors; Australia and Canada, the natural comparators before 1994, were not compiled [TBV].
South Africa is the only economy in the group with a clear fall in income per person since 2008 [IC]. Namibia was flat, and Mexico and Botswana grew slowly, so part of the slowdown is regional and commodity-wide [IC]. The World Bank states that over the past decade South Africa has “grown much slower than its upper-middle-income peers” (World Bank 2026) [IC].
Reading the arc: structure and governance
The companion research run tested whether the slowdown since 2008 is structural, in the sense that it predates the current governing arrangement, persists across changes of government and appears in comparable economies, or traces to governance choices. Its main documented findings:
- Structural forces are real and old. Gold depletion since the early 1970s, the coal-based power system and the apartheid spatial order predate 1994 and persist after 2024 [IC]. The Growth Lab finds that spatial exclusion in cities “has been entrenched through housing policies” after 1994 as well [IC].
- Multilateral and official analyses attribute most of the post-2008 shortfall to domestic factors. The SARB staff note says so directly [IC]. IMF staff find that total factor productivity subtracted 1.3 percentage points from growth in 2008 to 2023 relative to 1994 to 2007, and link it to the “state capture” era and inefficient state-owned enterprises (IMF 2025) [IC]. The Growth Lab states that “the economic slowdown is not due to macroeconomic problems or external shocks” [IC].
- The official record of governance failure. The Judicial Commission of Inquiry into State Capture estimated that about R57 billion of state spending was “tainted” by state capture, more than 97 percent of it at Transnet and Eskom (State Capture Commission 2022) [IC]. The 1998 power warning and the delayed build are documented decisions [IC]. In the National Party period, isolation held potential growth below 1 percent a year from 1985 to 1994 [IC].
- Turning points do not line up with changes of governing party. Income per person peaked in 1981 and in 2013, each inside a single party’s tenure [IC]. The 2024 recoveries in power supply and the grey-list exit began in 2023 [IC].
The finding the run supports: structural forces set a low ceiling on South African growth, and governance choices within each party’s tenure determined how far below that ceiling the economy ran [academic inference]. The run’s share estimates and its full attribution method sit in the research report outside the wiki.
Explanations this brief leaves aside include institutional-design accounts in the tradition of Acemoglu and Robinson, the commodity cycle treated as a stand-alone explanation, social-cohesion accounts, and labor-market rigidity accounts.
Where the arc departs from the corpus
- Deceleration at middle income. The corpus places the deceleration phase late in the curve of mature economies. South Africa’s income per person stalled at about US$6,000, far below that level [IC for the level; structural inference for the reading].
- The gauge undercounted a gain. Lens B was built on economies where GDP rose and households lost ground. South Africa shows the reverse on life expectancy [IC].
- Retention served a minority. Pre-1994 retention built industry for a racial minority and delivered per-person growth below comparable countries [IC]. The corpus account of retention needs a condition about who gains [structural inference].
Claim grades
| Claim | Tag | Recommended status | Safe for a public essay |
|---|---|---|---|
| Real GDP per person rose about 42% from 1961 to 2025, and in 2025 was about 4% above 1981 | [IC] | documented-fact | Yes |
| Income per person peaked in 1981 and 2013, each followed by a fall | [IC] | documented-fact | Yes |
| Gold output peaked in the early 1970s and has fallen about 90% | [IC] | documented-fact | Yes |
| Fixed investment fell from 29.2% of GDP (1976) to 13.9% (2025) | [IC] | documented-fact | Yes |
| Life expectancy fell about nine years to 2005 and rose about twelve after | [IC] | documented-fact | Yes |
| Government debt fell to 26.0% of GDP (2008/09) and rose to 78.9% (2025/26) | [IC] | documented-fact | Yes |
| The 1998 White Paper forecast the power shortfall for about 2007 | [IC] | documented-fact | Yes |
| South Africa’s income per person fell from 2008 to 2025 while most comparable economies grew | [IC] | documented-fact | Yes |
| Official and multilateral analyses attribute most of the post-2008 slowdown to domestic factors | [IC] | academic-inference | Yes, naming the sources |
| Pre-1994 retention delivered per-person growth below comparable countries | [IC] | academic-inference | With caveat |
| South Africa is decelerating at middle income | [IC] for the level | structural-inference | With caveat |
| Defensive spending softened measured contraction in the load-shedding years | [TBV] | working-hypothesis | No |
| South Africa shows Turchin’s structural-demographic pressures | [TBV] | working-hypothesis | No |
Related
- The S-Curve Thesis and the Retention Counter-Tradition
- What GDP Leaves Out: Lost Prosperity in Mature Economies
- The S-Curve Thesis
- Deceleration Indicators
- Economy versus BioConomy
- Retention Logic
- Sound Finance
- The Three Futures
- Two Systems, One Society
- Glossary: S-curve, throughput, economy
Sources
- Afrobarometer: Adaba, K. A., and Boio, D. (2024, 31 October). Across Africa, public trust in key institutions and leaders is weakening. Afrobarometer Dispatch No. 891.
- Bureau of Economic Analysis (n.d.). How Are GDP and Related Income Measures Affected by a Disaster? US Department of Commerce.
- CSIR (2025). CSIR releases statistics on power generation in South Africa for 2024. Council for Scientific and Industrial Research, Pretoria.
- Department of Minerals and Energy (1998, December). White Paper on the Energy Policy of the Republic of South Africa. Pretoria.
- Engineering News (2025, 19 March). Eskom implements Stage 2 loadshedding after loss of five generation units.
- Eskom (2024, 23 August). Loadshedding suspension continues: 149 days of constant energy supply and R10.60 billion reduction in diesel expenditure achieved.
- Expert Panel into the July 2021 Civil Unrest (2022, released 7 February). Report of the Expert Panel into the July 2021 Civil Unrest. The Presidency, Pretoria.
- Financial Gazette (2025, 17 September). Rooftop solar, now at 7 300MW, overtakes all Eskom’s IPP capacity.
- Fine, B., and Rustomjee, Z. (1996). The Political Economy of South Africa: From Minerals-Energy Complex to Industrialisation. London: Hurst.
- Hausmann, R., O’Brien, T., Fortunato, A., Lochmann, A., Shah, K., Venturi, L., et al. (2023, 15 November). Growth Through Inclusion in South Africa. Growth Lab, Harvard University.
- IMF (2025, January). South Africa: Selected Issues. IMF Country Report No. 25/29.
- IOL (2022, 24 February). Sasria gets R22bn shot in the arm after civil unrest brought the insurer to its knees.
- IOL (2024, 14 February). How South Africa’s load-shedding crisis escalated over 17 years.
- Janse van Rensburg, T., and Morema, K. (2023). Reflections on load-shedding and potential GDP. South African Reserve Bank Occasional Bulletin of Economic Notes OBEN/23/01.
- Klein, N. (2011). Measuring the Potential Output of South Africa. IMF African Department Paper No. 11/2.
- LBMA (2023). Spotlight on the South African Market, chapter 3, Mining. London Bullion Market Association.
- Leipert, C. (1989). National Income and Economic Growth: The Conceptual Side of Defensive Expenditures. Journal of Economic Issues 23(3), 843–856.
- Miningmx (2026, 24 March). Record gold prices fail to lift SA output.
- National Treasury (2025a, 24 October). South Africa exiting the FATF greylist. Media statement.
- National Treasury (2025b, 14 November). Treasury responds to rating action by S&P Global Ratings. Media statement.
- National Treasury (2026). 2026 Budget Review, chapter 7, Debt and financing.
- Nattrass, N., and Seekings, J. (2010). The Economy and Poverty in the Twentieth Century in South Africa. CSSR Working Paper No. 276, University of Cape Town.
- Portfolio Committee on Police (2024, 23 October). Budgetary Review and Recommendations Report on the Private Security Industry Regulatory Authority, 2023/24. Parliament of South Africa.
- Reinert, E. S. (2007). How Rich Countries Got Rich and Why Poor Countries Stay Poor. London: Constable.
- SAnews (2026, 8 September). GDP records 0.2% contraction in second quarter.
- South African History Online (n.d.). Republic of South Africa established.
- South African Reserve Bank (2019). Observations on the evolution of South Africa’s fiscal position. Quarterly Bulletin, box.
- State Capture Commission (2022). Report of the Judicial Commission of Inquiry into Allegations of State Capture. Summary of findings via the Presidency.
- Statistics South Africa (2025, 28 July). Mid-year population estimates 2025. Statistical release P0302.
- Statistics South Africa (2026, August). Quarterly Labour Force Survey, Quarter 2: 2026. Statistical release P0211.
- Tainter, J. A. (1988). The Collapse of Complex Societies. Cambridge University Press.
- TimesLIVE (2026, 11 June). Western Cape faces R9.1bn flood repair bill as farms, roads and communities count the cost.
- Turchin, P. (2016). Ages of Discord. Beresta Books.
- University of Cape Town Libraries (2024, 18 June). Statement of Intent of the 2024 Government of National Unity.
- World Bank (2026). South Africa: Overview.
- World Bank (2026). World Development Indicators. South Africa and peer series, retrieved 22 October 2026.
Provenance
Written on 22 October 2026 from South Africa on the Curve, 1961 to 2026: Testing Structure Against Governance, a deep research report produced the same day under version 0.1 of the instructions south-africa-on-the-curve-deep-research.md. Michael asked for a snapshot of the developmental arc, so this page reframes the report around the 65-year shape of the economy and keeps the party-independence test to one section.
Editorial decisions. The long-run table, the period growth rates and the peer table were rebuilt from World Bank WDI series retrieved on 22 October 2026, replacing period figures the report drew from a 2001 SARB address and from Moll (1991), which could not be retrieved; the WDI series shows a steeper fall in income per person in the 1980s than Moll’s estimate of about 0.5 percent a year, and the page uses WDI throughout for consistency. The peer table extends the report’s 2008 to 2024 window to 2025. Each quotation used here was checked against its source on 22 October 2026, and four report claims were corrected: Afrobarometer ranks South Africa’s fall in presidential trust fourth of 30 countries (the report said among the three largest); the Growth Lab’s statement on external causes is quoted in its own words; the Growth Lab dates the weakening of mining before 2008 without tying it to falling commodity prices; and load-shedding returned briefly in early 2025, so the page says it ended from April 2024 to the end of that year. The expanded unemployment rate is Statistics South Africa’s 43.8 percent (the report rounded to 44 percent). The report’s share estimates for structural and governance causes, its attribution method, its period-by-period attribution table and its treatment of state capture loss estimates beyond the Commission’s figure are left in the full report and kept off the wiki. Excluded: the September 2023 and January 2026 Western Cape storms and other place-level flood detail, named for the BioHubs directory. No individual is named; offices are described as the records describe them. Fine and Rustomjee (1996) and Feinstein’s comparison enter through the report and Nattrass and Seekings (2010); neither book was read directly for this page.