Living standards, health and social policy¶
What do health, height, pension and welfare evidence show about African living standards and the purposes of apartheid social policy?
Works engaging the debate: 12
Positions¶
- Biological indicators reveal unmeasured homeland poverty
- Fetal-origins effects
- Social pensions as apartheid instruments vs welfarist reform
- Health financing trade-offs
Evidence base¶
The debate rests on a few distinct bodies of evidence that rarely overlap. The biological-indicator and fetal-origins strand is built almost entirely on post-apartheid microdata read backwards. The three Mariotti papers (Mariotti (2012a)) (Mariotti (2012b)) (Mariotti (2012c)) use measured adult heights in NIDS Wave 1 (2008). Treatment is assigned by birth district council and the 1976-77 birth cohort in four recruiting homelands (Transkei, Ciskei, Bophuthatswana, KwaZulu). Treated cells are small: about 82 men and about 137 women. The labour-demand shock itself is documented from SAIRR annual surveys of mine employment and wages for 1969-79 and from Crush, Jeeves and Yudelman. Pre-shock homeland poverty comes from Tomlinson Commission figures from about 1953, Horrell, Wilson (1972) and Matanzima's speeches, none later than 1973. Dinkelman (2012) has a much larger evidential base. It uses the 1996 Census 10% individual sample (298,475 African males, with self-reported disability), station-level South African Weather Service rainfall converted to a drought index by district and birth year, and Perry-Castañeda homeland maps to code TBVC versus non-TBVC status. Mobility responses are traced through Census last-move data and 1995 household remittance receipts. Earlier biological evidence comes from occupational and colonial records. Harington et al. (1975) analyses 1,344 cancers in Chamber of Mines health returns over 1964-71, about 2.9 million man-years, by home territory and magisterial district, using Poisson observed/expected tests. Simkins and Van Heyningen (1989) applies indirect demographic estimation to the 1891 and 1904 Cape censuses and to 1895-1903 vital registration, and finds a roughly fifteen-year white/non-white gap in life expectancy.
The income and poverty strand is a methodological dispute over homeland household income in 1960, 1970 and 1980. McGrath and Simkins (1985) audits Simkins's microsimulation against Bureau of Market Research surveys (Pietermaritzburg 1976, Venda 1978, KaNgwane 1980), BMR Minimum Living Level reports and BENSO statistical surveys built on the 1980 census. The social-policy strand is documentary. Devereux (2007) relies on pension statutes from 1928 to 1973, House of Assembly debates (1965, 1973) and SAIRR fact papers to build a table of racial payment ratios, supplemented by Namibian interviews and 1993 World Bank coverage data. Duncan (1993) draws on Pretoria State Archives Public Health (GES) and Native Affairs (NTS) files, the wartime commissions (UG 9-1943, UG 14-1944, the Gluckman report UG 30-1944) and Union Year Book bed and loan counts. Author not named in text (MSc candidate (2012) adds Wheat Board annual reports for 1939-96, McGrath's racial tax shares and Smit's 1950 survey of 2,561 black urban families to argue for a hidden bread-subsidy transfer. The health-financing endpoint (Di McIntyre et al. (1995)) (McIntyre et al. (1995)) is the most systematically quantified. It uses the 1992/93 ReHMIS facility expenditure survey ranked by van Wyk magisterial-district income quintiles, together with Registrar of Medical Schemes returns deflated by the CSS CPI, a private-sector expenditure survey, SARB consumption estimates, 1991 census denominators and 1990 death registrations.
Contested evidence¶
The sharpest conflict is over what income evidence can say about homeland welfare. Simkins's simulations implied that over two-thirds of homeland households improved in real terms. McGrath and Simkins (1985) counters that the assumptions cannot be tested because no representative 1960s or 1970s surveys exist. He cites BMR tables showing that remittances vary within income types and correlate negatively with wages. He revises 1980 homeland agricultural employment from 413,600 to 767,200 using BENSO and census figures. He also argues that applying an urban Minimum Living Level overstates the level of poverty, though perhaps not its trend. The two sides even dispute a single chi-squared statistic (20.69): Simkins argues it was misapplied to the wrong units. The height literature implicitly enters this dispute from the other side. The 6 cm gain for 1976-77 male births in Mariotti (2012a) is read as evidence of severe pre-shock stunting that income aggregates miss. But that reading rests on an ad hoc calculation (about 24% of households gaining at least R14.22 a month against farm incomes of R40-120 a year) and on 1953 income data. The negative estimates for women and for cohorts born after the shock are unexplained. Exogeneity of recruitment location and the absence of in-migration under the pass laws are asserted, not tested. Dinkelman (2012) cites Mariotti and supports the view that apartheid-era shocks left lasting biological scars. However, it uses a different outcome (disability rather than height), a different source (the 1996 Census rather than NIDS 2008) and a different identifying contrast (TBVC versus non-TBVC mobility rather than labour demand). Its remittance mechanism also rests on 1995 rather than apartheid-era data. Harington et al. (1975) asserts that home environment rather than mine work caused the cancers. It has no age-structured denominators and no data on length of service or exposure, and the patterns could also be read as evidence on homeland diet and living conditions. On pensions, the same statutes and Hansard extracts support rival readings. Devereux (2007) treats differentiated urban, town and rural native rates and their later abolition as instruments of influx control. The same paper reports African take-up rising from about 54% of eligibles in 1947 to 81% of all pensioners by 1993, and racial ratios narrowing from the 1970s, which a welfarist reading would emphasise. Duncan (1993) reads the 1940s commissions against the legislation actually passed as cynical appearance-making. Neither has Cabinet or Treasury deliberations, so motive is inferred from selective speeches. The bread-subsidy redistribution claim in Author not named in text (MSc candidate (2012) depends on aggregate tax shares and consumption shares rather than incidence data. Health outcome comparisons in Di McIntyre et al. (1995) (African IMR 54 against white 7) are acknowledged to rest on incomplete death registration.
Underused and missing sources¶
Several sources that would test central claims are named in the evidence ledgers' own gaps but are rarely used. TEBA/NRC recruiting records by magisterial district would allow direct treatment assignment in the Mariotti design and a test of its exogeneity claim. Chamber of Mines health and medical records are used only by Harington et al. (1975) and, glancingly, (McIntyre et al. (1995)); individual service histories would separate home-area from workplace causation. Representative homeland household income surveys for 1960 and 1970 are missing from the Simkins-McGrath exchange, as are sensitivity runs and region-specific price indices. The pension and welfare literature would also be strengthened by several archival series. These include Department of Bantu Administration, South West Africa Administration, Cabinet and Treasury files, and racially disaggregated budget series for pensions, housing and health. Cause-of-death data by race and district would sharpen mortality comparisons from Simkins and Van Heyningen (1989) to Di McIntyre et al. (1995). Constituency returns for 1943 and 1948, and household expenditure surveys, are absent from the bread argument. The district, homeland and TBVC shapefiles (1961 magisterial districts, MDB, GADM, Harvard) would allow birth-district or rainfall-station assignment on consistent historical boundaries rather than 2011 district councils. The SAIRR Survey extracts for 1961-76 cover the mine employment, wage and pension-rate series on which (Mariotti (2012a)) and Devereux (2007) rely. The held CPI series permits consistent deflation of pension rates, mine wages and medical-scheme contributions.
Strength of key claims¶
| Basis | Share of key claims |
|---|---|
| Systematic data | 21% |
| Partial data | 62% |
| Illustrative or anecdotal | 14% |
| Asserted | 3% |
Comparative figures for all debates are on the Literature overview.
Best-evidenced works¶
Dinkelman (2012), McIntyre et al. (1995), Harington et al. (1975), Simkins and Van Heyningen (1989), Duncan (1993)
Most-used primary sources¶
Holdings are matched by machine against the archive's file manifest; a dash means no match was found, not that the project certainly lacks the source.
Works and their positions¶
| Work | Position |
|---|---|
| Author not named in text (MSc candidate (2012) | Challenges Wylie's account of NP neglect by showing a costly bread subsidy and the enriched loaf that in effect favoured black consumers. The subsidy is framed as a by-product of serving white consumers and farmers, not a pro-poor aim. |
| Devereux (2007) | Rejects a purely welfarist reading. Argues the pension served instrumental apartheid aims (de-urbanisation, influx control, homelands resettlement, co-opting elites), though it also reflected Hofmeyr's liberal equity concerns and Smuts' 'enlightened self-interest'. |
| Devereux (2007) | Weighs two explanations: bolstering separate development/influx control and 'winning hearts and minds'. The regional coverage data cast doubt on the second (northern, SWAPO-leaning regions had the lowest coverage), and the motive is undocumented. |
| Devereux (2007) | Favours diffusion theory over the convergence hypothesis, with elements of welfare industrialisation and interest-group theory, to explain the 1920s pension for poor whites. |
| Devereux (2007) | Argues sustainability is as much political as fiscal. He cites Beattie and McGillivray that there is 'no economic law' preventing greater old-age spending, and warns that means testing risks eroding middle-class political support. |
| Di McIntyre et al. (1995) | Neutral but critical: documents rapid medical scheme cost escalation and regulatory gaps, treats the tax subsidy and civil-servant contributions as problems, and flags social health insurance as an option needing cost control. |
| Di McIntyre et al. (1995) | Argues the extension of primary care must be financed largely by savings and efficiency in tertiary and academic hospitals, plus gap funding from government and donors. |
| Di McIntyre et al. (1995) | Sees limited scope to raise fees on the poor without harming access, but greater potential from private and insured patients in specialist hospitals. |
| Dinkelman (2012) | Extends the literature by arguing that mobility also insures long-term health outcomes, implying an even larger role for mobility than previously emphasised. |
| Dinkelman (2012) | Finds that mobility restrictions imposed measurable long-run health and human-capital costs in the homelands, which is a hidden cost of the labour-control system. |
| Dinkelman (2012) | Confirms that drought in utero and at ages 1-2 raises adult disability, especially for males, consistent with the fragile-male literature. |
| Duncan (1993) | Welfare measures were largely window-dressing and legitimation, aimed at controlling unrest and protecting whites from disease. Some reforming civil servants did push for more. |
| Duncan (1993) | Mining capital, backed by the Mines and Native Affairs departments, secured exclusion of migrant mineworkers from unemployment insurance. This protected low-wage migrancy and the influx-control system. |
| Duncan (1993) | The Smuts government had already abandoned its plans for improving African workers' conditions before 1948. The National Party inherited a worsening problem rather than reversing a reform trajectory. |
| Harington et al. (1975) | Holds that cancers reflect home-area exposure (diet, childhood, social factors) and that mining has little or nothing to do with them. |
| Mariotti (2012a) | Supports it: the effect is concentrated in men born at the shock, with no effect on those born slightly before or after. |
| Mariotti (2012a) | Finds no effect on existing children, which is consistent with Duflo (2003) that male income is not necessarily passed on to children. |
| Mariotti (2012a) | Accepts that the mining industry relied on cheap foreign labour and that homeland poverty was severe. Interprets stunting as the physical consequence of that poverty. |
| Mariotti (2012b) | Supports it: only men born during the shock years show a height gain, with no effect on slightly older or younger cohorts. |
| Mariotti (2012b) | Finds results consistent with male income gains not reaching existing children, and no benefit for girls born during the shock. |
| Mariotti (2012b) | Argues that heights show homeland poverty caused physical stunting that had previously gone unmeasured. |
| Mariotti (2012c) | Supports it: only men exposed in utero in 1976-77 show large height gains, with no effect for earlier or later cohorts. |
| Mariotti (2012c) | Takes the standard account that foreign workers were preferred because they accepted low wages, and that homeland conditions had deteriorated. Argues even low mine wages meant a large rise in household income for destitute homeland families. |
| Mariotti (2012c) | Contends living standards were very poor and unmeasured, and that the height gap with Whites reflects environment, not genetics. |
| McGrath and Simkins (1985) | McGrath says the trends are unknowable from Simkins's simulations, though metropolitan evidence is compatible with Simkins's broad direction. Simkins argues that over two-thirds of homeland households improved rapidly in real terms and that inequality rose. |
| McGrath and Simkins (1985) | McGrath says remittance and commuter assumptions are arbitrary and probably vary with correlated wages. Simkins argues that homeland households benefited through remittances and commuting from integration with the metropolitan economy. |
| McIntyre et al. (1995) | Neutral and descriptive. It documents the cost spiral in medical schemes and the tax subsidy to them, and treats SHI as an option that needs cost-control safeguards, without recommending it. |
| McIntyre et al. (1995) | Argues the public system is heavily biased to tertiary and academic hospitals and that much of the primary care expansion must be funded by hospital savings. It also accepts that some government and donor 'gap-funding' is needed. |
| McIntyre et al. (1995) | Takes slow growth, a deficit of 7.8% of GDP and rising debt service as binding limits on health budgets, so health must compete with housing, education and welfare. |
| Simkins and Van Heyningen (1989) | Uses demographic magnitudes as welfare indicators and finds whites were the main beneficiaries of public-health gains. Real living costs changed little across 1891-1904. |