The Impact of Sanctions on South Africa and Its Periphery¶
Charles M. Becker (1988). The Impact of Sanctions on South Africa and Its Periphery. African Studies Review, Vol. 31, No. 2. African Studies Association.
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | theoretical or conceptual |
| Period studied | 1976–1987 |
| Regions | South Africa, Botswana, Lesotho, Swaziland, Zimbabwe, Mozambique, Malawi, Zambia, Bantustans |
| Themes | sanctions divestment, macro fiscal finance debt, labour markets wages, mining migrant labour, bantustans removals influx, business and state, comparative theory, manufacturing industrial policy |
| Citation key | becker1988 |
Argument¶
Becker argues that the Western sanctions in force by 1986-87 are largely symbolic and economically trivial, partly shaped by domestic protectionist interests. He treats South Africa as a segmented dual economy rather than a single-good economy, so different sanctions hit different groups differently. Sanctions that cut demand in capital- and skill-intensive sectors (gold boycott, high-tech embargo, loan and new-investment bans) induce rand depreciation, which shifts output toward labour-intensive export sectors and so cushions unskilled blacks. Sanctions that favour import-substituting manufacturing, or an oil embargo, would hurt blacks and the frontline states badly. Effective sanctions must therefore be paired with infrastructural aid to South Africa's neighbours, who are highly dependent on South Africa.
Evidence and method¶
Non-mathematical, model-informed argument drawing on South African Central Statistical Services input-output tables (1978, 1985), Official Yearbooks, Foreign Trade Statistics, Reserve Bank data, sectoral factor-share tables, World Bank tables, Hanlon (1986), and earlier input-output sanctions studies (Spandau 1979, Richardson 1986). Also uses Becker's own earlier models (1985, 1987a, 1987b).
Place in the debates¶
- Sanctions effectiveness and who bears the cost (blacks vs whites) (see Sanctions, disinvestment and the anti-apartheid movement). Challenges the claim that sanctions must hurt poor blacks most. Argues that tailored sanctions (gold, high-tech, finance) fall mainly on whites and skilled workers and cause only moderate losses to unskilled blacks. Concedes that no stringent package is costless, since the frontline states and black South Africans would still suffer.
- Divestment and disinvestment as an anti-apartheid tool (see Sanctions, disinvestment and the anti-apartheid movement). Sceptical: divestment mostly transfers assets to white South Africans, and Sullivan-principle firms' exits may leave black workers worse off. It is useful chiefly for high-tech firms, and indirectly through expectations and rand depreciation.
- Whether apartheid is a social compact serving all white groups, and where its weak points lie (see Sanctions, disinvestment and the anti-apartheid movement). Apartheid is a compact serving virtually all white groups. Semi-skilled white workers will not yield to sanctions, but skilled workers, farmers and industrialists may erode in support as turnover costs rise and sanctions bite.
- Input-output assessments of boycott impact (see Sanctions, disinvestment and the anti-apartheid movement). Criticises Spandau and Richardson for ignoring imported intermediates, foreign-exchange constraints and changes in final demand. Holds that these flaws probably understate the impact of export boycotts on whites.
Findings¶
-
Existing Western sanctions are symbolic, politically motivated, and in part protectionist. (p. 4)
In reality, the sanctions can be seen more as a statement of determination and moral condemnation, for the economic impact is certainly trivial.
-
US coal ban illustrates sanctions chosen for domestic interests: coal imports were only about 1% of US imports from South Africa, and the US competes with South Africa in European coal markets. (p. 4)
The unrestricted 1985 level of such imports, $43 million, was just over 1% of total US imports from South Africa
-
Sanctions can be designed to harm whites while sparing unskilled blacks by raising labour demand. (p. 5)
It is possible to tailor sanctions that harm white South Africans, but which may well cause only modest reductions in the real incomes of unskilled blacks.
-
Homeland population share rose under apartheid segregation, yet Bantustan economies remain tiny and agricultural. (p. 8)
the South African government's efforts to segregate society actually led to a rise from 37 percent in 1960 to 53 percent in 1980 of the proportion of Africans living in the homelands
-
The underdeveloped (homeland) sector supplied cheap black labour, kept modern-sector wages low and served the regime politically. (p. 8)
the underdeveloped sector has played the key role of supplier of low-cost black labor, thus keeping modern sector wages low.
-
Neighbouring states depend heavily on South African trade and customs revenue. (p. 10)
BLS countries experience extreme dependence, obtaining about 90 percent of their imports from South Africa.
-
A gold boycott would shrink the economy but, via devaluation, shift activity toward labour-intensive sectors and cushion unskilled blacks. (p. 14)
A contraction in gold mining that leads to devaluation therefore can be expected to increase the overall labor intensity of the South African economy
-
The 1985 credit crisis and rand collapse show the investment/credit channel at work. (p. 22)
the credit crisis caused the rand to decline from 0.74 dollars per rand in 1984 to a low of 0.38 in September 1985
-
Cheap-labour benefits of the pass system are offset by turnover costs, eroding industrialists' stake in apartheid. (p. 25)
The benefit of cheap labor created by the pass system and its successors is partially counteracted by the extremely high turnover rates it creates.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Republic of South Africa Central Statistical Services, Input-Output Tables, 1978 and 1985 (preliminary, RAM method) | official statistical series | 1978, 1985 | Table 2 gives sectoral import, export and re-export shares of output, used to characterise sectoral trade structure and import intensity. |
| Official Yearbook of the Republic of South Africa, 1979 and 1986 (Bantustan and South African GDP by sector) | official statistical series | 1976-1980 | Table 3 compares sectoral GDP of the Bantustans with South Africa to show the very low homeland share and the dual economy. |
| Official Yearbook 1979 and 1986 and South African Reserve Bank Supplement to the Quarterly Bulletin (factor intensities) | official statistical series | 1973, 1978, 1984 | Table 7 estimates labour (by race) and non-labour shares of value added by sector, used to argue that sanctions-induced sectoral shifts raise labour intensity. |
| Republic of South Africa, Foreign Trade Statistics, 1985 (1984 import sources by country) | international trade or finance data | 1984 | Table 5 lists sources of South African imports by nation, used to judge the likelihood of concerted sanctions among major trading partners. |
| Republic of South Africa (1984), Foreign Trade Statistics (trade flows by commodity section, 1983-84, with UK, Japan, US and West Germany detail) | international trade or finance data | 1983-1984 | Table 6 gives imports and exports by commodity and partner, used to show the composition of trade and the share of manufactures in US imports. |
| Hanlon (1986), Beggar Your Neighbors, statistical appendix | secondary literature compilation | c. 1970s-1980s | Supplies data on the frontline states' dependence on South Africa for imports, labour remittances, SACU revenue and South African ownership of assets. |
| Sanction chronology sources: Congressional Quarterly 1986 (27 Sept.) and Hufbauer and Schott (1985) | secondary literature compilation | 1963-1987 | Table 1 documents the sanctions imposed by the US, Japan, Britain, EC and UN, used to argue that they are mainly symbolic. |
| Richardson (1986) and Spandau (1979) input-output studies of export boycotts | secondary literature compilation | 1979-1986 | Input-output simulation results on the effect of gold and export boycotts on output and employment are cited and then critiqued as incomplete. |
| Kaempfer and Lowenberg (1986b) oil figures | secondary literature compilation | 1980s | Used to estimate that imports are at least 60-70 percent of South African oil consumption. |
| Consumer expenditure surveys (Africa Insights 1984) and survey of registered black workers by Dept. of Cooperation and Development (Africa Insights 1985) | household or labour survey microdata | 1980s | Establishes black households' consumption pattern (less import-intensive than white) and migrant remittance data. |
| World Bank, World Tables (1983) and Population Growth and Policies in Sub-Saharan Africa (1986) | international trade or finance data | 1980s | Provides sectoral GDP comparisons (transport share) and infant mortality and birth rates used for the 25,000-80,000 avoidable infant deaths calculation. |
| Barclay's Business Briefs, Johannesburg (January 1986) | newspapers periodicals | 1981-1986 | Source for South Africa's foreign debt rise ($18bn to $23.7bn) and for the weak manufacturing export response to the 1983 exchange decontrol. |
| Internal World Bank documents on Malawi; Christiansen and Kydd (1983) | commission or official report | 1966-1977 | Supports the claim that Malawi absorbed returned miners while smallholder output per head rose. |
| US News and World Report, 11 August 1986 | newspapers periodicals | 1986 | Quoted to show that a blanket embargo on Western exports could be circumvented through false bills of lading. |
| Exchange rate and foreign trade figures (rand per dollar 1981-1987; 1985 mining proceeds; 1985 US coal imports) | official statistical series | 1981-1987 | Supplies rand depreciation figures and 1985 mineral export proceeds to document the 1985 credit crisis and the weight of gold. |
Inference¶
Quantification: Descriptive tables.
Theoretical, qualitative analysis based on a dual-economy (Hansen-Lewis) general-equilibrium framework, with sector-by-sector reasoning about the incidence of sanctions; descriptive tables of trade, GDP and factor shares are used illustratively, and the results of other authors' input-output studies are critiqued. No formal estimation is performed.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Existing Western sanctions are largely symbolic, with trivial economic impact, and were chosen partly for domestic political or protectionist reasons. (p. 63) | Table 1 sanctions chronology and the 1985 US coal import figure ($43m, about 1% of US imports from South Africa). | Illustrative or anecdotal |
| The South African economy is a segmented dual economy; the Bantustan economy is tiny and extremely poor relative to the modern sector, and the homeland share of Africans rose from 37 to 53 percent between 1960 and 1980. (p. 67) | Table 3 (Official Yearbook GDP by sector), Table 4 per capita GDP and Table 2 trade structure. | Partial data |
| Sanctions that raise demand for labour-intensive sectors (a gold boycott or high-tech embargo causing rand depreciation) would hurt whites far more than unskilled blacks. (p. 73) | Table 7 labour shares by sector and race, Table 2 import intensities, consumer expenditure surveys and the theoretical model; no direct estimation of the effect. | Partial data |
| The frontline states and BLS countries are extremely dependent on South Africa for imports, remittances, SACU revenue and oil, so sanctions would hit them harder than South Africa. (p. 69) | Hanlon (1986) statistical appendix; SACU revenue shares and remittance shares reported; Foreign Trade Statistics. | Partial data |
| An oil embargo would be catastrophic for the region, and gold and diamond/coal boycotts differ widely in feasibility (gold most boycottable, diamonds and coal easily circumvented). (p. 77) | Input-output studies by Richardson and Spandau, Kaempfer and Lowenberg oil figures, Bailey (1981), and trade and production comparisons. | Partial data |
| Divestment and new-investment bans have limited direct effect but could depreciate the rand; the 1985 credit crisis already produced a sharp rand decline. (p. 81) | Exchange rate series (1981-1987), debt figures from Barclay's Business Briefs, and theoretical argument. | Partial data |
| Apartheid causes large excess infant mortality, with 25,000-80,000 babies' lives saved annually if South Africa matched neighbouring states' rates. (p. 84) | World Bank (1986) infant mortality and birth rate figures with the author's assumptions about white infant mortality and population. | Partial data |
The author's caveats¶
Notes that figures were unavailable for modern agriculture or any underdeveloped sectors in Table 7, that only empirical work can determine the magnitudes, and that the existing input-output studies are flawed. Also acknowledges that the discussion is non-mathematical and that the question of the effect of the rand's collapse on exports demands empirical research.
Gaps in the evidence¶
The central distributional claim (unskilled blacks lose little) rests on sectoral labour shares and theory, with no estimated elasticities, general-equilibrium simulation or labour-market microdata on wages and employment by race. Missing data on agriculture and homeland labour shares, trade elasticities, rand pass-through and actual export response to the 1985 depreciation would test it; some such data existed in the 1980s (October Household Surveys did not yet exist, but Reserve Bank and census series did). Evidence on sanction circumvention and on frontline-state dependence is aggregated from Hanlon and official trade statistics rather than firm- or border-level records, which is weak for judging leakage.
Replicability: Data described but not public.