An Analysis of Economic Infrastructure Investment in South Africa¶
Peter Perkins; Johann Fedderke; John Luiz (2005). An Analysis of Economic Infrastructure Investment in South Africa. South African Journal of Economics, Vol. 73:2. Blackwell Publishing / Economic Society of South Africa.
| Type | journal article |
| Scope | Economics and economic history of apartheid |
| Method | econometric |
| Period studied | 1870–2002 |
| Regions | South Africa |
| Themes | business and state, macro fiscal finance debt, methods data sources, mining migrant labour, transition post apartheid |
| Citation key | perkins2005a |
Argument¶
The paper assembles a long-run database (1870-2002) on South African economic infrastructure: public-sector investment and capital stock, railways, roads, ports, air travel, telephones and electricity. It uses Pesaran-Shin-Smith bounds F-tests to identify the direction of association between infrastructure and real GDP. Growth in GDP mostly drives infrastructure, but public infrastructure investment, fixed capital stock and roads appear to drive GDP, with some simultaneity. Infrastructure developed in phases, and public infrastructure investment fell sharply from the mid-1970s to 2002. The authors argue that the need for investment never goes away and that projects should be chosen by cost-benefit analysis.
Evidence and method¶
Newly collated and updated time series from the SARB database, Statistics South Africa and its predecessors, Union Statistics for Fifty Years, Official Year Books, railway general managers' reports, SANRA, NPA, Telkom, Vodacom and SAA. Also uses secondary literature (De Kock, Frankel, DBSA 1998, Merrifield 2000) and the Department of Transport (1998). Appendix 2 gives series notes.
Place in the debates¶
- Does public infrastructure raise output? (international literature and South African DBSA estimates) (see Did the apartheid economy fail? Growth, productivity, investment and the post-1970s crisis). Takes a middle position. Notes wide variation in estimates, with Aschauer high and Holtz-Eakin and Garcia-Milà near zero, and warns that some DBSA results are spurious (OLS on time series). Finds causation running in both directions, with GDP driving most infrastructure and only roads and aggregate public infrastructure investment and capital stock forcing GDP.
- Was the post-1976 decline in public investment a correction of earlier over-expansion or damaging underinvestment? (see Did the apartheid economy fail? Growth, productivity, investment and the post-1970s crisis). Acknowledges the over-expansion argument and overcapacity in some areas, but leans towards seeing the long decline as probably undesirable. Also links it to growing government consumption and fiscal deficits late in apartheid.
Findings¶
-
Per capita public-sector economic infrastructure investment collapsed after the mid-1970s. (p. 3)
Investment per capita fell from R1 268 in 1976 to R356 in 2002 (1995 prices), a collapse of 72 per cent.
-
Infrastructure investment fell as a share of GDP to below the international benchmark. (p. 3)
Investment fell from 8.1 per cent of GDP to 2.4 per cent of GDP
-
From 1980 the state shifted from investment to consumption spending, funded partly by deficits, possibly to prop up the apartheid system. (p. 4)
After 1980 government consumption expenditure increased at the expense of government investment
-
The budget deficit widened in the late apartheid years. (p. 4)
As a proportion of GDP, the national budget deficit rose from an average of 2.9 per cent in 1980-986 to an average of 4.9 per cent in 1987-1994
-
Rail expansion was tied to the mining discoveries and the revenues they generated, with very rapid early growth. (p. 5)
Between 1875 and 1885, the average annual growth rate in South Africa’s railway lines was 29 per cent.
-
Paved national and provincial roads expanded massively after 1940, supporting rapid growth in road traffic. (p. 7)
Between 1940 and 2001, South Africa’s national and provincial paved roads increased from around 2 000km to around 63 000km.
-
Port capacity was expanded sharply in the 1970s with Richards Bay and Saldanha. (p. 9)
between 1976 and 1979 the volume doubled from 38 million harbour tons to 76 million harbour tons, the result of opening Richards Bay and Saldanha harbours.
-
Gold-mining demand drove early electricity supply. (p. 11)
in 1923 the electricity sold by VFTPC to the mines exceeded the combined electricity consumption of London, Sheffield and Birmingham.
-
Apartheid left under-serviced areas, which Eskom's 1990s electrification programme targeted. (p. 11)
In the 1990s Eskom embarked on an electrification programme to expand its services to areas which had been neglected during apartheid
-
The main econometric result is that GDP generally forces infrastructure, rather than the reverse. (p. 13)
In South Africa, growth in GDP tends to drive growth in individual physical measures of infrastructure-related goods and services, rather than vice versa.
-
Roads are the exception, with a very high correlation with GDP and a forcing relationship to GDP. (p. 13)
the correlation coefficient between paved roads and real GDP is 0.996 for the period 1938-2001
Evidence ledger¶
Sources used¶
| Source | Kind | Period | How used |
|---|---|---|---|
| South African Reserve Bank database (national accounts: GDP, gross fixed capital formation and fixed capital stock of public-sector economic infrastructure, 1995 prices) | national accounts macro | 1946-2002 (GDP); infrastructure investment and capital stock series c.1960-2002 | Supplies real GDP and public-sector infrastructure investment and capital stock, the key variables in the PSS F-tests and in the decline of investment since 1976. |
| South African Reserve Bank Annual Economic Reports, 1996-2000 | official statistical series | 1996-2000 | Cited to explain the 1990s recovery and subsequent slump in infrastructure investment (Telkom, Eskom, SAA). |
| Statistics South Africa and predecessors: South African Statistics (2002 and other years), including SSA 1980 value-added series | official statistical series | 1911-2002 | Source for rail, road, vehicle, port, air, phone and electricity series, population, and for back-extrapolating nominal GDP to 1911. |
| Union Statistics for Fifty Years, 1910-1960 jubilee issue | official statistical series | 1910-1960 | Historical series for railways, roads, electricity, population and wholesale/retail price indices used to deflate early GDP. |
| Official Year Book of the Union (1916, 1917, 1925, 1933) | official statistical series | 1916-1933 editions | Source for early rail lines, roads, ports, telephone lines and the dates of first telephone exchanges. |
| Population censuses of 1904 and 1911 | population census | 1904, 1911 | Anchor population figures used to extrapolate the series back for per capita measures. |
| Railway administrations' Reports of the General Manager (Cape of Good Hope, Colony of Natal, Natal Government Railways, Central South African Railways, South African Railways and Harbours), various years | company records or reports | 1873-1916 and later | Sources for locomotives, coaching stock, goods stock, passenger journeys and revenue-earning traffic in the early period. |
| South African Railways and Harbours / South African Transport Services / Spoornet annual reports | company records or reports | 1910-2003 | Rail and port series including rolling stock, carrying capacity, passenger journeys and freight; 1903-1908 passenger/freight estimated from graphs in SARH 1910. |
| CSIR, SANRA and NATIS road and vehicle data | official statistical series | 1920-2002 | Series for total and paved road distance and for passenger and goods vehicle numbers; paved series for 1994-2000 interpolated. |
| National Ports Authority data and Ports of Southern Africa and Mauritius (Young, 2003) | official statistical series | 1911-2001 | Cargo handled in harbour tons, used as a proxy for port capacity; Durban Container Terminal congestion reported. |
| South African Airways, SATS and SSA air passenger figures; SAA audited group results year ended 31 March 2003 | company records or reports | 1950-2003 | Passenger numbers used as a proxy for air transport infrastructure; fleet data cited. |
| Telkom, Vodacom and SSA telephone line data (plus Canning 1998 database) | company records or reports | 1912-2003 | Fixed and mobile line series; 1934-1960 interpolated from telephones in use. |
| Eskom and Union Statistics/SSA electricity generation data; DBSA (1998) history of electricity | official statistical series | 1920-2003 | Electricity generated series and Eskom capacity and peak demand figures supporting the electricity trend and capacity warning. |
| Department of Transport (1998), Moving South Africa: a transport strategy for 2020 (draft), including freight customer interviews | commission or official report | 1998 | Used for estimates of funding shortfalls in roads and ports and overloading damage, and for customer dissatisfaction at ports. |
| SANRA Annual Report 2002 and Horizon 2010; National Treasury Medium Term Budget Policy Statement 2003 | commission or official report | 2002-2004 | Evidence on road overloading damage (R600m a year) and road spending trends. |
| DBSA (1998), Infrastructure: a foundation for development (summary of earlier elasticity estimates) | secondary literature compilation | 1998 | Reports prior estimates of the output elasticity of infrastructure (0.15-0.33) and electricity history. |
| De Kock (1936) The Economic Development of South Africa | secondary literature compilation | 1836-1936 | Historical narrative on Kimberley, railway building and early roads and their link to diamond revenues. |
New data¶
Authors compiled and updated a long-run database of South African infrastructure series (railways, roads, vehicles, ports, air passengers, phone lines, electricity, plus national accounts and population), 1870-2002, standardized in units, with technical notes in Appendix 2.
Inference¶
Quantification: Econometric (causal design).
Descriptive time-series trends and growth-rate tables by sub-period, plus Pesaran-Shin-Smith bounds F-tests (ARDL error-correction form, pairwise with real GDP) to identify directions of long-run association; historical narrative supplements.
Key claims¶
| Claim | Basis | Strength |
|---|---|---|
| Public-sector infrastructure investment per capita collapsed 72% between 1976 and 2002, falling from 8.1% to 2.4% of GDP. (p. 213) | SARB national accounts data on public-sector infrastructure investment, deflated to 1995 prices and expressed per capita. | Systematic data |
| For most infrastructure types (rail lines, ports, air passengers, phone lines) GDP drives infrastructure, not vice versa; roads are an exception where roads drive GDP. (p. 223) | PSS F-tests of GDP against individual infrastructure series (Tables 3, 5, 7, 9) and a correlation of 0.996 for paved roads and GDP. | Partial data |
| Early railway expansion was driven by diamond and gold discoveries which created both demand and revenue. (p. 215) | Narrative reliance on De Kock (1936) and railway growth rates from the railway reports. | Illustrative or anecdotal |
| Ports and electricity face capacity constraints: port congestion and possible inadequate generation capacity for peak demand by 2007. (p. 221) | Department of Transport 1998 report, Young (2003) and Eskom capacity and peak demand figures. | Partial data |
| The post-1976 decline in infrastructure investment was part of a broader shift from government investment to consumption and rising deficits, possibly tied to the unsustainable apartheid political system. (p. 214) | Merrifield (2000) and budget deficit figures cited in a footnote; no direct data analysis. | Asserted |
The author's caveats¶
Notes that roads data quality is questionable and unpaved road data unreliable; there is a structural break in 1956/57 (reclassification of 25 000km), no consistent provincial road data after 1993, no lane-number provision, some series are interpolated or estimated from graphs, and cargo and passengers are only indirect measures of port and air infrastructure. Concedes that PSS tests are preliminary and that earlier series were discontinued.
Gaps in the evidence¶
The analysis uses only bivariate bounds tests with GDP, with no controls for omitted variables (e.g. private investment, mining and gold cycles, sanctions, policy regimes), so causal claims about infrastructure and growth are weakly identified. Series are physical quantities rather than quality, utilization or cost measures, and apartheid-specific features (homeland and racial allocation of infrastructure, regional breakdowns) are not analysed. Provincial or regional infrastructure data and project-level cost-benefit evidence existed in departmental and parastatal archives and could test the argument better; sectoral multivariate models would also be feasible with the compiled data.
Replicability: Data described but not public.