Sanctions on South Africa: What Did They Do?¶
Philip I. Levy (1999). Sanctions on South Africa: What Did They Do?. Yale University Economic Growth Center Discussion Paper No. 796. Economic Growth Center, Yale University.
| Type | working paper |
| Scope | Economics and economic history of apartheid |
| Method | synthesis review |
| Period studied | 1948–1994 |
| Regions | South Africa |
| Themes | sanctions divestment, macro fiscal finance debt, labour markets wages, politics ideology elections, transition post apartheid, resistance social movements |
| Citation key | levy1999 |
Argument¶
Levy argues that the mid-1980s government-imposed trade sanctions had very small direct economic effects and probably played little role in ending apartheid. The most damaging international actions were private (bank loan withdrawals, corporate disinvestment), and they followed South Africa's own political and economic deterioration. The slowdown in growth began in the early 1970s, well before sanctions. He offers an alternative, overdetermined account: internal labour-market distortions and rising economic costs, effective black political opposition, and the fall of communism, which removed the Afrikaner government's fear of a communist ANC. Sanctions' main effect was probably psychological, and he concedes that their role cannot be conclusively ruled out.
Evidence and method¶
Secondary literature and statistics drawn from it: Hufbauer, Schott and Elliott (1990) on sanctions timeline and cost estimates, Jones and Müller (1992), Lewis (1990), Lipton (1989), Waldmeir (1997), and Lowenberg (1997). It also uses public statements by Mandela, Tutu and Boesak, and a Chase Manhattan executive quote. No original archival or survey data.
Place in the debates¶
- Sanctions effectiveness in ending apartheid (see Sanctions, disinvestment and the anti-apartheid movement). Sceptical: the trade sanctions' contribution was probably trivial and possibly counterproductive; the role was mainly psychological and cannot be fully excluded.
- Private (financial/disinvestment) vs public sanctions as the source of economic pressure (see Sanctions, disinvestment and the anti-apartheid movement). The capital reversal was far more painful than trade sanctions, but it was driven by private lenders responding to unrest and instability, so it cannot be counted as evidence of sanctions' efficacy.
- Causes of the end of apartheid (internal economic inefficiency vs resistance vs external/Cold War factors) (see Why apartheid ended: reform, transition and elite bargaining). Overdetermined: labour-market distortions, black political opposition and the fall of the Soviet bloc are given priority over sanctions.
Findings¶
- The growth break came well before the 1985-87 sanctions. (p. 4)
-
South Africa depended heavily on foreign capital and was exposed through short-term debt. (p. 6)
By the mid-1980s, South Africa’s external debt was roughly $24 billion, of which two-thirds was short-term.
-
The 1985 financial crisis was triggered by private lenders, before multilateral sanctions. (p. 6)
Shortly thereafter, Chase Manhattan Bank declared it would not renew its short-term loans, touching off a liquidity crisis as other lenders followed suit.
-
Capital withdrawal was driven by private risk assessment and cannot be counted as evidence for sanctions. (p. 6)
this privately induced financial crisis — the repercussions of which were substantially greater than any of the public sanctions that ensued — cannot be used as evidence of sanctions’ effectiveness.
-
Export volumes rose despite trade sanctions, which South Africa circumvented. (p. 7)
In sum, from 1985 to 1989, export volumes rose by 26 percent, although terms of trade suffered.
-
The estimated marginal cost of the trade sanctions was small. (p. 7)
One estimate of the marginal cost to South Africa of the mid-1980s trade sanctions was $354 million annually, or 0.5 percent of GNP.
-
Sanctions omitted gold, South Africa's largest export, so their reach was limited. (p. 7)
gold was easily South Africa’s largest export. In 1985 it accounted for 42.6 percent of the value of merchandise exports.
-
Labour-market reforms began in the 1970s, well before serious sanctions, pointing to internal economic pressures. (p. 11)
This is supported by the steady relaxation of apartheid labor laws, beginning in the 1970s, well before serious sanctions.
-
Levy credits the collapse of the communist threat with enabling a political deal with the ANC. (p. 11)
Once the communist bloc fell apart and a withdrawal of Cuban troops from Angola was negotiated, a political deal with the ANC became conceivable.
-
The paper concludes that sanctions' positive role is implausible, though not disprovable. (p. 12)
the best one can do is argue that it is implausible that trade sanctions played a significant and positive role.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| Jones and Müller, The South African Economy, 1910-90 (real GDP growth series 1974-87 and earlier decades) | secondary literature compilation | c.1950s-1987 | Supplies average real GDP growth rates (4.9% pre-1974 vs 1.8% 1974-87) used to show the growth break predated the 1985-87 sanctions. |
| Lipton (1989), 'The Challenge of Sanctions', South African Journal of Economics (annual GDP growth 1986-88) | secondary literature compilation | 1986-1988 | Provides annual GDP growth figures (0.5%, 2.6%, 3.2%) showing growth accelerated after sanctions. |
| Jones and Müller current account data (deficits of 2-3% of GDP, 1940s-1984) | secondary literature compilation | 1940s-1984 | Establishes South Africa's long dependence on foreign capital inflows to finance current account deficits. |
| Lewis (1990), The Economics of Apartheid (net capital outflow 1976-80; dividend outflows vs new FDI; Iranian oil) | secondary literature compilation | 1970s-1980s | Used for capital outflow of 2.3% of GDP in 1976-80 and for profit/dividend outflows exceeding new FDI. |
| Waldmeir (1997), Anatomy of a Miracle (external debt, Chase quote, export volumes, disinvestment, Afrikaner anti-communism) | secondary literature compilation | 1970s-1990 | Provides external debt figures ($24bn, two-thirds short-term), the Chase executive quote, export volume growth and disinvestment behaviour. |
| Hufbauer, Schott and Elliott (1990), Economic Sanctions Reconsidered, 2nd ed. (sanctions timeline, cost estimate, scoring, GNP, Mandela quote) | secondary literature compilation | 1985-1990 | Source of the sanctions timeline, the $354 million (0.5% of GNP) cost estimate, 1984 GNP, and the effectiveness scoring that is critiqued. |
| Jones and Müller (1992) export composition data, 1985 | secondary literature compilation | 1985 | Shows gold was 42.6% of merchandise exports in 1985, so sanctions omitting gold limited direct impact. |
| Comprehensive Anti-Apartheid Act (CAAA) 1986 and EC/Japan sanctions measures | legislation gazette | 1985-1987 | Describes the content and scope of the sanctions (import bans, lending restrictions, omissions) as the object of assessment. |
| Public statements by black leaders and South African actors (Tutu, Boesak, Mandela 1990) | other | 1985-1990 | Statements are treated as the main evidence in favour of sanctions' (psychological) effectiveness. |
Inference¶
Quantification: Illustrative figures.
Qualitative, narrative case study built on secondary sources: timing comparison (growth break in 1974 vs sanctions in 1985-87; private capital flight preceding public sanctions), back-of-envelope sanction cost from Hufbauer et al., and counterfactual reasoning/theory favouring an alternative 'overdetermined' account (labour distortions, internal opposition, collapse of communism). No econometrics or new data.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| The decisive break in South African economic performance predated the 1985-87 sanctions, so sanctions cannot explain the slowdown. (p. 4) | GDP growth averages from Jones and Müller (1992) and annual rates from Lipton (1989). | Partial data |
| The direct economic impact of governmental trade sanctions was small (about 0.5 percent of GNP), and South Africa circumvented them. (p. 7) | Hufbauer, Schott and Elliott's cost estimate and Waldmeir on export volumes. | Partial data |
| The financial crisis was caused by private lenders' decisions that preceded public sanctions, so it cannot be counted as evidence of sanctions' effectiveness. (p. 6) | Chronology from Hufbauer et al. and Waldmeir, including the Chase executive quotation. | Illustrative or anecdotal |
| The end of apartheid is better explained by labour-market distortions, internal political opposition, and the fall of communism than by sanctions. (p. 11) | Narrative chronology and secondary accounts (Waldmeir, Lowenberg); timing argument only. | Asserted |
| Sanctions' principal effect was psychological, and this was insufficient to swing the balance against apartheid. (p. 12) | Statements by black leaders and South African officials; reasoning about Afrikaner incentives; no direct data. | Asserted |
| Capital outflows were far more painful economically than trade sanctions, but are largely attributable to private rather than public measures. (p. 8) | Lewis (1990) on dividend outflows exceeding new FDI, and Waldmeir on disinvestment. | Partial data |
The author's caveats¶
Acknowledges that the end of apartheid was overdetermined, so sanctions' ineffectiveness cannot be proven. Notes that Lipton's GDP figures may understate growth because of informal-sector activity. Acknowledges that statements by South African leaders denying sanctions' role may be strategic and that the psychological impact is hard to resolve.
Gaps in the evidence¶
The paper relies entirely on secondary compilations and lacks direct data on trade flows, capital flows, and the timing and composition of disinvestment, which official balance-of-payments and Reserve Bank statistics could provide. It does not use government archives, cabinet or National Party records, or interviews to test whether sanctions influenced decision-makers, material that has since become available through archives and memoirs. A cross-country or counterfactual econometric design, or event-study evidence on financial markets and the rand around sanction dates, could test the timing argument better.
Replicability: Data poorly documented.