An analysis of economic infrastructure investment in South Africa

Abstract :

This paper discusses a time-series data set of economic infrastructure goods and services in South Africa, including infrastructure investment in the national accounts, railways, roads, ports, air transport, telephones and electricity. Social infrastructure, such as schools and hospitals, is not considered in this paper. The data were collected from a range of sources, with a view to extending the series as far back as possible in order to identify long-run economic infrastructure trends. The relevance of these to the country’s economic policymakers is also considered, firstly by investigating infrastructure in relation to the rest of the economy, secondly by reviewing the empirical literature on infrastructure (in particular the relationship between infrastructure and long-run economic growth), and thirdly by analysing the long-run trends in South Africa’s infrastructure in relation to the country’s gross domestic product (GDP) since 1911. The latter is a preliminary analysis which does not attempt to extract all the explanatory information from the data set, but provides a basis for policy implications and may provide a useful background for further quantitative research.

The paper finds that South Africa’s economic infrastructure developed rapidly from the mid- 18708 to the mid-1970s (an important role was played by the growing mining industry), but this was followed by a sharp slowdown from the late-1970s to 2002 (the country’s savings and total investment rates fell, the share of government consumption in the economy grew, and, during the 1990s, fiscal consolidation became a priority). The levels of infrastructure goods and services are generally highly correlated with the level of annual real GDP. The slowdown in infrastructure development from the late 1970s coincided with falling real GDP per capita during 1982-1993 and slowly rising real GDP per capita during 1994-2002.

To examine whether infrastructure promotes GDP growth or GDP growth promotes infrastructure, and to assess the strength of these associations, a series of Pesaran, Shin and Smith (PSS) F-tests and associated autoregressive distributed lag (ARDL) regressions are reported. These indicate significant long-run relationships between infrastructure and GDP in South Africa. More specifically, the results indicate that South Africa’s GDP growth tends to drive growth in individual measures of infrastructure-related goods and services rather than vice versa. The evidence suggests that the effects are strong, although this could be the result of under-specification bias. Paved roads are an exception, as they appear to have a strong effect on South Africa’s GDP growth, and at the national accounts level infrastructure investment seems to drive GDP growth. Accordingly, two forms of constraint that infrastructure may exercise on economic growth may be distinguished. Firstly, if policymakers fail to provide additional infrastructure in response to the greater demand for infrastructure generated by GDP growth, further GDP growth could be hampered by bottlenecks. Secondly, underinvestment in certain types of infrastructure may leave potential areas of economic growth unexploited.

The finding that infrastructure matters in terms of South Africa’s long-run economic growth is subject to two considerations, namely the type of infrastructure in question, and the stage of the country’s development. Four policy implications which follow from the ARDL regression results are as follows. Firstly, given the finding that South Africa’s GDP growth tends to drive growth in individual measures of infrastructure-related goods and services, policymakers should monitor the growing pressures placed on existing infrastructure by GDP growth and identify areas of congestion or wear and tear. Secondly, policymakers should respond to infrastructure- related pressure points by budgeting and implementing the required investment, since aggregate infrastructure investment (at the national accounts level) has a positive impact on the long-run trend in GDP. Thirdly, at different stages of economic development, it may be appropriate to discontinue the expansion of a certain type of infrastructure and allow it to level out, a phenomenon referred to in this paper as a ‘plateau’ effect. This would be the case where GDP growth fails to cause congestion effects related to the particular type of infrastructure in question. Fourthly, paved roads appear to be particularly effective in promoting long-run economic growth, provided that their construction is supported by rigorous cost-benefit analysis. While these results rest exclusively on the data set for South Africa, they could potentially apply to other countries as well.

The South African government is showing renewed interest and commitment with regard to investing in the country’s economic infrastructure in response to backlogs, following two-and-a- half decades (1977-2002) of an infrastructure slowdown. The findings presented in this paper are fully supportive of this shift in policy.



Author Perkins P
Date Accessioned 2016-09-22T11:20:46Z
Date Available 2016-09-22T11:20:46Z
Date Created
Identifier URL 2003
Language English
Subject 2
Alternative Title
Degree Type Masters degree
Degree Description  MCom