The Impact of Economic Sanctions on South African Exports¶
Simon J. Evenett (2002). The Impact of Economic Sanctions on South African Exports. Scottish Journal of Political Economy, Vol. 49, No. 5. Blackwell Publishers / Scottish Economic Society.
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | econometric |
| Period studied | 1978–1999 |
| Regions | South Africa, European Community (France, Italy, Great Britain), Nordic countries (Denmark, Norway, Sweden), Japan, United States |
| Themes | sanctions divestment, methods data sources, comparative theory, transition post apartheid |
| Citation key | evenett2002a |
Argument¶
Evenett estimates the effect of eight industrialised economies' sanctions on their imports from South Africa. He uses an annual gravity equation for 1978-1999 and compares South Africa with 30 other middle-income exporters. A sanctions dummy is estimated separately for the EC, Nordic countries, Japan and the US, before, during and after the 1986-91 sanctions. On the full sample, EC sanctions look most damaging. This result depends on outlier bilateral pairs with very large or very small trade flows. Once the outliers are trimmed, the US Comprehensive Anti-Apartheid Act emerges as the most robustly harmful, cutting imports by about a third. Nordic and Japanese effects are mostly statistically insignificant. US exports do not recover after 1991, which suggests hysteresis. The paper also argues that outliers inflate gravity estimates of distance elasticity.
Evidence and method¶
Bilateral import data for 8 importers and 31 middle-income exporters, 1978-1999 (248 pairs per year). Source is the World Bank Trade and Production Database (Olarreaga and Nicita 2001), derived from UN COMTRADE. National incomes come from World Development Indicators, deflated to 1995 US dollars, with great-circle distances between capitals. The sanctions chronology (Table 1) comes from Hufbauer, Schott and Elliott (2002).
Place in the debates¶
- Effectiveness of trade sanctions against apartheid South Africa (see Sanctions, disinvestment and the anti-apartheid movement). Finds measurable trade effects, but only robustly for US sanctions. EC effects are an artefact of outliers, and Nordic and Japanese effects are insignificant. This is more nuanced than Levy's claim that sanctions had a trivial role, though the paper estimates trade effects only, not political effects.
- Whether sanctions effects are reversible once sanctions are lifted (see Sanctions, disinvestment and the anti-apartheid movement). Suggests hysteresis. US exports stayed low or fell after 1991, and the EC showed only a small recovery, so removing sanctions need not restore pre-sanction trade.
- Plausibility of gravity-equation distance elasticities (Grossman 1998) (see Data reliability and econometric method). Argues that outliers inflate estimated distance elasticities and sanctions dummies. Trimming reduces the distance elasticity by about half.
Findings¶
-
Headline result: after robustness checks, US sanctions had the largest effect, cutting imports by about a third. (p. 1)
robustness checks reveal that the United States’ Comprehensive Anti-Apartheid Act played the largest role, reducing bilateral imports by a third
-
Naive estimates using all observations point to EC sanctions as the most damaging, but this depends on outliers. (p. 9)
In Table 3, which includes all 248 bilateral pairs in each annual sample, only the sanctions parameters for the EC nations are negative and statistically significant throughout 1977–1999.
-
Trimming outliers cuts the estimated EC effect to about 27%, against roughly half in the untrimmed sample. (p. 16)
However, once outliers are excluded, the EC sanctions regime is found to reduce South African exports by 27%; still a sizeable impact but not a 50% cut.
-
The US effect is the largest once outliers are removed. (p. 16)
In contrast, US sanctions are found to reduce the apartheid regime’s exports the most; essentially by a third.
-
Nordic and Japanese sanctions show no robust statistically significant effect. (p. 14)
outliers are not responsible for the finding that, by and large, the Nordic and Japanese regimes had no statistically significant effect on South African exports.
-
US exports did not recover after sanctions were removed, whereas the EC may have seen some bounce-back. (p. 2)
The estimates also suggest that South African exports have not recovered after the removal of US sanctions in 1991.
-
Persistent post-sanctions damage suggests hysteresis in export relationships. (p. 16)
If prior export success lays the foundation for future export growth, then it appears that the US sanctions regime may well have had adverse consequences for South African exports long after that sanctions regime was dismantled.
-
Outliers inflate the gravity distance elasticity, which falls by about half once they are excluded. (p. 2)
the distance elasticity falls 50% once the outliers are excluded, going some way to remedy the implausibly large size of typical gravity-inspired estimates of this elasticity
-
Corporate divestment may have operated as an additional channel reducing South African trade. (p. 5)
To the extent that multinationals without their former South African subsidiaries ceased or reduced sourcing parts, components, and raw materials from South Africa, then dis-investment may have influenced South African trade flows too.
-
Positions the paper against Levy's thesis that sanctions played a trivial role. (p. 1)
Levy (1999) presents a provocative thesis that state-led trade-based economic sanctions played only a ‘trivial’ role in toppling the apartheid regime.
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| World Bank Trade and Production Database (Olarreaga and Nicita, 2001), bilateral import data 1978-1999 | international trade or finance data | 1978-1999 | Supplies the bilateral import flows (dependent variable) for the gravity regressions estimating sanctions effects on South African exports. |
| UN COMTRADE database | international trade or finance data | 1978-1999 | Original underlying source of the bilateral trade data in the World Bank database. |
| World Bank World Development Indicators (CD-ROM and country income classification) | international trade or finance data | 1978-1999 | Provides national income (GDP), the US price deflator, and income-group classifications used to select the comparison set of middle-income exporters and the gravity variables. |
| Great Circle distance data between capitals (Haveman gravity data, Macalester web resource) | maps spatial data | time-invariant | Supplies the distance variable that proxies for transport costs/trade barriers in the gravity equation. |
| Hufbauer, Schott and Elliott (2002), Case 62-2 and Case 85-1, notes on sanctions against South Africa (mimeo, IIE) | secondary literature compilation | 1962-1993 | Supplies the chronology of sanctions (Table 1) that defines the pre-sanctions, sanctions and post-sanctions periods and which countries are treated. |
| Redding and Venables (2000) gravity formulation | secondary literature compilation | n/a | Theoretical gravity specification adapted to include a sanctioned share of output as the basis of the estimation equation. |
New data¶
Author assembled annual cross-sectional samples 1978-1999 of the full bilateral trade matrix of 8 industrial importers by 31 middle-income exporters (248 pairs per year), combined with GDP, deflator and distance data, from existing World Bank and COMTRADE sources.
Inference¶
Quantification: Econometric (causal design).
Year-by-year cross-sectional gravity regressions with importer and exporter dummies, distance and a South Africa-importer-specific sanctions dummy. Effects are read by comparing the dummy across pre-sanctions (1978-85), sanctions (1986-91) and post-sanctions (1992-99) periods. Robustness is checked by trimming outliers of the dependent variable at cutoffs c = 0.01 to 0.10. Medians of annual estimates are converted to implied percentage reductions in exports.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Including outliers, EC sanctions appear to have the largest negative effect on South African exports, with the sanctions parameter negative and significant in most years. (p. 565) | Table 3 annual gravity estimates on the full 248-pair sample (c=0) from the World Bank/COMTRADE trade data. | Systematic data |
| The EC result is not robust: with outliers trimmed the EC sanctions effect shrinks and loses significance in most years. (p. 567) | Table 4 re-estimation of the EC parameter at cutoffs c = 0.01-0.10. | Systematic data |
| US sanctions (the CAAA) had the largest and most robust effect, cutting South African exports to the US by about a third. (p. 572) | Tables 4 and 6: trimmed-sample estimates for the US dummy, with significance rising over time and an implied median reduction of about 33% in 1986-91. | Partial data |
| Nordic and Japanese sanctions had no statistically significant effect on South African exports. (p. 570) | Table 4 sensitivity analysis for the Nordic and Japanese dummies across cutoffs. | Systematic data |
| South African exports to the US did not recover after sanctions were removed in 1991, whereas EC imports show a small rebound, suggesting persistent hysteresis in trade patterns. (p. 572) | Table 6 implied median percentage reductions for 1986-91 versus 1992-99 at c=0.10. | Partial data |
| Outliers inflate the estimated distance elasticity; trimming the extreme 10% of observations on each side cuts it by about a third to a half. (p. 570) | Table 5 median distance parameters across cutoffs from the annual gravity regressions. | Systematic data |
The author's caveats¶
The author notes that the sanctions variable is a country-pair dummy that captures both state sanctions and private-sector reluctance to buy South African goods. He also notes that distance is only an imperfect proxy for transport costs, that many annual parameters are imprecisely estimated, and that the distance-outlier results are not reported (available on request). He treats the zero-trade convention as a possible source of outliers.
Gaps in the evidence¶
The analysis uses aggregate bilateral imports and a single dummy per importer, so it cannot separate sanctioned sectors (iron, steel, coal, agriculture) from unsanctioned ones; product-level COMTRADE data would test this and existed at the time. The trimming rule is ad hoc and there are no panel or difference-in-differences controls for South African political risk, the exchange rate, the gold-price and mineral cycles, or third-country diversion, so the causal attribution to sanctions is weak. Data on divestment, firm-level sourcing decisions and sanctions evasion (for example via re-labelled exports) would help but are largely unavailable.
Replicability: Data described but not public.