Capital flight under apartheid

Literature › Works

Zavareh Rustomjee (1991). Capital flight under apartheid. Transformation 15.

Type	journal article
Scope	Economics and economic history of apartheid
Method	descriptive quantitative
Period studied	1970–1988
Regions	South Africa, USA (trade case study 1985), Latin America (comparison)
Themes	sanctions divestment, macro fiscal finance debt, business and state, transition post apartheid, methods data sources, manufacturing industrial policy
Citation key	rustomjee1991

Argument

Rustomjee estimates that South Africa lost about $37.1 billion in current prices ($55.3 billion in 1988 prices) between 1970 and 1988 through unrecorded capital outflows. He measures this with mirror-trade statistics, comparing South African reported trade with partners' reports, and adds the balance-of-payments errors and omissions item. Export underinvoicing is the main conduit. Because trade is highly monopolised, he infers that the large conglomerates (Anglo American, De Beers and others) account for most of the flight. A 1985 SA-USA commodity case study supports this. He concludes that exchange controls failed, that flight began before the clear crises and continued through early-1980s liberalisation, and that a post-apartheid state cannot rely on 'confidence-building' alone. He draws policy implications: stronger trade monitoring, retained capital controls and addressing conglomerate power.

Evidence and method

IMF Direction of Trade Statistics yearbooks (1970-88) compared with South African reported trade. IMF International Financial Statistics, including the errors and omissions item. SARB Quarterly Bulletin for debt figures. Comparative Latin American estimates from Cuddington and Gulati in Lessard and Williamson (1987). Morgan Guaranty (1986) flight estimates. A preliminary 1985 SA-USA commodity breakdown (CCCN). Secondary literature and press reports on conglomerate structure.

Place in the debates

• Causes and scale of capital flight under apartheid: official statistics vs unrecorded flows (see Finance, capital flight, the 1985 debt crisis and banking). Argues official balance-of-payments statistics understate outflows. Unrecorded flight via trade misinvoicing is far larger than other estimates such as Morgan Guaranty's.
• Free-market claim that flight responds to state interventionism / business 'confidence' (see Finance, capital flight, the 1985 debt crisis and banking). Rejects it as a sufficient explanation, noting flight did not abate when National Party policy liberalised in the early 1980s. Is sceptical that confidence-building measures will stop flight.
• Falling profitability as cause of capital outflow (see Finance, capital flight, the 1985 debt crisis and banking). Argues measured falling profitability may be as much a result of flight as its cause, so profitability-based arguments need careful treatment. Implicitly contests Nattrass-type use of recorded profit statistics.
• Sanctions and disinvestment (see Finance, capital flight, the 1985 debt crisis and banking). Questions the good faith of business and financial leaders who argued against sanctions and disinvestment while moving capital out themselves.
• Post-apartheid mixed economy and export-oriented growth (see Post-apartheid economic policy, governance and corruption). Favours retaining or strengthening capital controls, monitoring trade and curbing conglomerate power. Warns that export-led growth without anti-flight mechanisms may exacerbate the drain.

Findings

• Headline estimate of capital syphoned out via trade manipulation, 1970-88. (p. 1)

  between 1970 and 1988, some $55 billion of capital has been systematically syphoned out of the South African economy through manipulation of the trading system

• Export underinvoicing averaged 20%, in line with other developing-country studies. (p. 4)

  Between 1970 and 1988, exports were underinvoiced by an average of 20% of South African exports after allowing for FOB/CIF differences.

• A worked 1988 example of the unrecorded outflow through exports. (p. 4)

  in 1988 while specified goods to a declared value of $13 677 million were exported, their actual value realised on sale in the destination country was $15166 million.

• Import underinvoicing was small, producing minor offsetting inflows, a puzzle he attributes to conglomerate dominance. (p. 5)

  imports were underinvoiced by an average of 4% of South African imports resulting in relatively smaller unrecorded capital inflows to South Africa.

• Cumulative unrecorded outflow compared with 1988 foreign debt. (p. 7)

  The magnitude of unrecorded outflows since 1970 at current prices is $37.1 billion ($55.3 billion in 1988 prices). Note that South Africa's foreign debt stood at $21.2 billion in 1988.

• His estimates exceed the Morgan Guaranty flight estimates for the same sub-periods. (p. 7)

  Hence they estimate flight from South Africa at $13 billion (our estimate $18.8 billion) between 1976 and 1982 and $4 billion (our figure $9.8 billion) between 1983 and 1985.

• In the 1985 SA-USA case study, underinvoicing is concentrated in conglomerate-dominated commodities. (p. 8)

  indicates that underinvoicing exports to the USA amounts to $724 million, about 32% of reported US imports.

• Falling profitability may be partly an effect of flight rather than only its cause. (p. 9)

  the magnitude and consistency of flight suggests falling profitability to be as much a result of massive capital flight, averaging 7.0% of GDP over a period of 18 years, as its cause

• Flight was large relative to GDP and foreign debt, despite exchange controls. (p. 11)

  an average of 7.0% of GDP per annum and more than twice the size of the country's 1988 foreign debt.

• Flight continued during early-1980s financial liberalisation, which undercuts the free-market explanation. (p. 12)

Evidence ledger

Sources used

Source	Kind	Period	How used
IMF Direction of Trade Statistics (DOTS) Yearbooks, 1977, 1982, 1988	international trade or finance data	1970-1988	Compares South African-reported world trade with trading partners' reported trade to estimate export and import misinvoicing as capital flight.
IMF International Financial Statistics (IFS) Yearbooks, various years 1970-1988	international trade or finance data	1970-1988	Supplies recorded capital flows and the Errors and Omissions item, FOB/CIF ratios and the CPI deflator used to build recorded versus unrecorded flows.
IMF Balance of Payments Statistics (monthly and yearbook)	international trade or finance data	not specified	Mentioned as an alternative series whose differences from IFS are attributed to reporting lags; IFS used instead.
RSA Foreign Trade Statistics, 1985	official statistical series	1985	Defines the 'Unclassified Goods'/special categories (military stores, petroleum, gold) excluded from the comparison and notes petroleum's reclassification after 1979.
South African Reserve Bank statistics on capital flows and foreign assets/liabilities, and Quarterly Bulletin (June 1989)	official statistical series	to 1986; debt figure 1984	Used to check that IFS aggregate capital flows correlate with SARB data and as the source of South African foreign debt in Table 1.
SA-USA trade by commodity (CCCN classification), 1985	international trade or finance data	1985	Commodity-level comparison of South African export values with US-reported imports to identify underinvoiced goods.
Cuddington (1986) and Cuddington in Lessard and Williamson (1987) - Latin American errors and omissions flight estimates	secondary literature compilation	1974-1984	Provides comparative Latin American Errors-and-Omissions flight figures for Table 1a.
Gulati (1985) / Gulati in Lessard and Williamson (1987) - trade misinvoicing estimates	secondary literature compilation	1977-1983 (and 1974-1984 in Table 1b)	Provides comparative misinvoicing estimates for Table 1b and the range of export underinvoicing.
Morgan Guaranty Trust Company (1986) 'LDC Capital Flight' estimates for South Africa	secondary literature compilation	1976-1985	Alternative stock-change estimates of South African flight compared with the author's figures.
Press reports (Financial Times, The Times, Financial Mail, UK press on Pensions Investment Research Consultants survey, South African press)	newspapers periodicals	1988-1990	Used for the De Beers/Minorco restructuring, ANC economic policy discussion, investor reluctance and sanctions-busting anecdotes.

New data

Author-computed annual series of South African export and import misinvoicing (1970-1988) from mirror-trade comparison, combined with IFS Errors and Omissions into an unrecorded flow series, plus a preliminary 1985 SA-USA commodity-level underinvoicing breakdown; calculations available from the author on request.

Inference

Quantification: Statistical analysis.

Mirror-trade discrepancy analysis: South African-reported trade is compared with partner-reported IMF DOTS data, and the gap is taken as misinvoicing and hence capital flight. This is added to Errors and Omissions, compared with recorded flows over time and with Latin American estimates, and supplemented by a one-year SA-USA commodity case study and assertions about conglomerate dominance.

Key claims

Claim	Basis	Strength
Roughly $37.1bn ($55.3bn in 1988 prices) of unrecorded capital left South Africa between 1970 and 1988, about 7.0% of GDP per year. (p. 7)	Sum of mirror-trade misinvoicing from IMF DOTS plus IFS Errors and Omissions, deflated by the IMF CPI.	Partial data
Exports were underinvoiced by an average of 20% over 1970-1988, and imports by about 4%, giving net outflow through trade. (p. 4)	Comparison of South African-reported figures with partner-reported DOTS data, adjusted for FOB/CIF.	Partial data
In 1985 underinvoicing of exports to the USA was about $724 million, around 32% of reported US imports, concentrated in precious stones, chemicals and minerals. (p. 8)	Preliminary CCCN commodity-level comparison of SA-USA trade, 1985.	Partial data
The bulk of flight was probably channelled through the dominant conglomerates. (p. 8)	Inference from the monopolised structure of the commodities found to be underinvoiced; no firm-level data.	Asserted
South African flight, measured via Errors and Omissions, is within the Latin American range, but flight via trade misinvoicing is far higher relative to debt. (p. 7)	Tables 1a and 1b comparing the author's figures with Cuddington's and Gulati's country estimates.	Partial data
Falling measured profitability is partly a result of capital flight rather than only its cause, and flight did not abate during early-1980s liberalisation. (p. 9)	Interpretation of the flight series against official profitability and recorded flows; no profitability data analysed.	Asserted
Recorded flows show net inflows in 1981-84 while unrecorded flows were larger outflows, so official statistics mask flight. (p. 7)	Figure 2 superimposing unrecorded series on recorded IFS flows.	Partial data

The author's caveats

The author notes that single-invoice transfer pricing is excluded, so estimates are conservative. Special categories (about 35% of trade) and unspecified destinations are excluded; timing lags, destination misreporting and unspecified partners may contaminate the residual. The partner-statistics-are-accurate assumption is stated, the 1985 SA-USA breakdown is described as preliminary, and the SARB statistics were unpublished after 1986.

Gaps in the evidence

The method treats the entire mirror-trade discrepancy as flight without testing alternative explanations such as valuation, timing, re-exports, gold/diamond marketing conventions, or SACU treatment. Firm-level customs, exchange-control or Reserve Bank transaction records would be needed to attribute flight to conglomerates, and these were largely inaccessible then. The conglomerate and determinants claims rest on assertion; disaggregated bilateral comparisons for other partners, and later released records, could test them.

Replicability: Data described but not public.