The role of economic

infrastructure in
economic growth:
building on experience
Economic infrastructure may be compared to the foundation of a building. It plays a supporting
role, facilitating the multitude of productive economic activities that constitute the bulk of the
economy, or gross domestic product.

Broadly speaking, economic infrastructure comprises investments and related services

that raise the productivity of other types of physical capital, e.g. transport, power,
water systems, communication; and social infrastructure comprises investments
and services that raise the productivity of human capital, e.g. education and health.
The subject of this paper is the relationship between economic infrastructure and
economic growth in South Africa. The relationship between social infrastructure
and economic growth is just as complex and no less important, but should be the
subject of a separate analysis.

The development of economic and social infrastructure in South Africa has a long
Peter Perkins and troubled history. Excellent in parts yet hopelessly inadequate in others, and
holds an MCom riddled with the discriminatory practices and inequalities that were the hallmarks
in economics and of the country’s apartheid past, the South African infrastructural experience does
has worked as an not lend itself to generalisations or easy assessment.
economist at the
Chamber of Mines History aside, if South Africans today perceive there to be infrastructural backlogs
of South Africa, JCI everywhere they look, they are not alone. Lamenting the poor condition of America’s
Ltd, Standard Bank, infrastructure in The New York Times in August and October, economist Paul
and Fleming Martin Krugman described America as being ‘on the unlit, unpaved road to nowhere’1 and
Securities. He lectured its roads, railways and sewer systems as ‘antiquated and increasingly inadequate’2. In
economics at Wits October, The Economist3 had the following to say about infrastructure in Britain:
and currently works for
Statistics South Africa. The dreadfulness of Britain’s infrastructure has become legendary. Visitors to
The views expressed in
continental Europe and elsewhere return with awed tales of cheap, clean trains
this paper are his alone.
that run on time, zippy new roads and properly functioning airports, which are
crossly compared with Britain’s pot-holes, delays and check-in scrums.

Why should there be such concern over the state of economic infrastructure, whether
in South Africa or elsewhere? Apart from the sheer convenience of having enough
electricity at an affordable price, clean water, efficient public transport, free-flowing
traffic, and telephones that work, few would dispute the notion that there are close

24
 T h e r ol e of ec onomic inf r a s t r u c t u r e i n e c o n o m i c g r o w t h : b u i ld i n g o n e x p e r i e n c e

links between economic infrastructure and economic growth. The evidence for such
links is strong, whether in the form of quantitative or qualitative studies, whether
over time or across regions or countries. In the following section I shall confine
myself to just two examples.

Correlation between infrastructural investment and economic

growth
The first example of the relationship between economic infrastructure and the
economy is a cross-country comparison of electricity usage and gross national income
(GNI). Figure 1 plots the per capita GNI of 124 countries against each country’s
per capita electricity usage in 2007. Each point on the graph represents a country.
The correlation between income and electricity is strongly positive. Excluding the
outliers Norway and Iceland, whose high rates of per capita electricity consumption
lie well beyond the scale of Figure 1 (25 000 and 37 000 kilowatt hours respectively),
the correlation is 89%. Low-income countries generate and use relatively low levels
of electricity. High-income countries generate and use relatively high levels of
electricity. It is reasonable to suppose that a similar pattern exists for other types of
infrastructure and infrastructural services.
Figure 1 – Cross-county comparison of electricity consumption and gross national income, per capita, 2007
Electricity consumption per capita, kilowatt hours

Data source: World Development Indicators

The second example is the South African experience between 1960 and 2009.
Figure 2 shows the relationship between gross domestic product (GDP) (bold line)
and public-sector economic infrastructural investment4 (light line), both measured
per capita and in real terms (i.e. adjusted for inflation). The two series follow similar
trends. Average annual growth rates for selected periods are shown in Table 1.

25
p eter p er kins

Figure 2 – South Africa: real GDP and public-sector economic infrastructural

investment, per capita, rands, 2005 prices, moving average

40 000 2 500

Economic infrastructure investment per capita

35 000 2 000
GDP per capita

30 000 1 500

25 000 1 000

20 000 500
1960 1965 1970 1975 1980 1985 1990 1995 2000 2005

GDP per capita Economic infrastructure investment per capita

Data source: South African Reserve Bank

Table 1 – South Africa: real GDP and public-sector economic infrastructural

investment, per capita, average annual growth rates

GDP per capita, Economic infrastructure

Period
% p.a. per capita, % p.a.
1961–1976 2.2 6.0
1977–1978 -1.0 -15.7
1979–1982 1.5 4.0
1983–1993 -1.4 -8.1
1994–2002 0.9 0.3
2003–2008 3.2 19.3
2009 -2.7 19.5
Data source: South African Reserve Bank

From the early 1960s to mid-1970s, both series exhibited positive growth. Between
the late 1970s and early 1980s performance was mixed, but then, too, the two series
moved broadly in the same direction, first falling and then rising. Between the early
1980s and early 1990s, a period during which the disastrous economic consequences
of the apartheid system became increasingly clear, both GDP and infrastructural
investment declined in real per capita terms. Happily, between the mid-1990s and
the early 2000s the long-term downward trends in both series were halted. Then
followed a seven-year period of rapid expansion in infrastructural investment, on
average 19.3% per annum (in real per capita terms) during the six years 2003–2008,
with a similar performance in 2009. Real per capita GDP grew by 3.2% per annum
during 2003–2008, but was negative in 2009 largely on account of the global
economic recession.

The extended decline in living standards during the 1980s and early 1990s was
particularly tragic when considering that for many other emerging market

26
 T h e r ol e of ec onomic inf r a s t r u c t u r e i n e c o n o m i c g r o w t h : b u i ld i n g o n e x p e r i e n c e

economies this was a period of rapid economic growth methods of infrastructure provision, in particular the
and development. South Africa’s investment malaise generation of electricity? How much emphasis should
was by no means confined to economic infrastructure be placed on economic infrastructure compared with
during the 12 years 1982–1993: the level of annual social infrastructure (health and education)? Can the
total investment (as measured by gross fixed capital experiences of one country or region provide easy
formation including the private sector) declined in real lessons or models for other countries or regions? It lies
terms, and as a proportion of GDP averaged 20 per far beyond the scope of this paper to provide answers
cent (25 per cent during 1982–1985 and 18 per cent to these and other tricky questions faced by policy
during 1986–1993). South Africa’s savings rate also makers in the field of infrastructure. Partial answers to
declined over this period. By contrast, high-growth some of them in the South African context may lie in
emerging market economies in Asia maintained high the historical development of South Africa’s economic
rates of savings and investment (Table 2). It was not infrastructure. This is discussed in the following
until 2005/6 (a quarter of a century) that South Africa section. The section following that addresses more
returned to its 1981 high in terms of per capita real explicitly the question of causality between economic
GDP (Figure 2). infrastructure and economic growth in the context of
the South African experience.
Table 2 – Rates of investment and economic growth
during 1982–1993 Historical development of economic
infrastructure in South Africa
GDP (real) % Investment as a
Country The building of South Africa’s economic infrastructure
p.a. % of GDP
during the 19th and 20th centuries was dominated by
South Africa 0.7 20.3
the state. Not exclusively, but certainly for the most
Malaysia 6.8 31.8 part, it was the state that owned and operated railway
Singapore 7.3 38.0 lines, roads, harbours, airports, water systems, power
South Korea 8.4 32.1 stations and communication networks.5 As with social
infrastructure, access to economic infrastructural
Thailand 8.2 32.6
services was in most cases determined along racial
Data source: International Financial Statistics lines, heavily skewed in favour of the minority
white population and away from the majority black
Whereas the correlation between infrastructural population. The damage so done is incalculable, except
investment and economic growth may be to say that it must have been enormous; no assessment
demonstrated with ease, causality between the two of it will be attempted here.
is more difficult to show. Is there any causality at all,
or is their correlation merely coincidental? If there The history of rail in South Africa began in the 1860s,
is causality, does infrastructural investment lead when the Natal Railway Company and the Cape Town
economic growth, or does economic growth lead Railway and Dock Company opened the first railway
infrastructural investment, or does the causality run lines in and around Durban and Cape Town. It was
in both directions depending on circumstances? not long before these developments were purchased
by the Natal and Cape governments, and thereafter
These are not the only difficult questions that the growth of rail in South Africa became largely a
arise regarding the provision of infrastructure and government affair. By the late 1870s, four main lines
infrastructural services. If infrastructure does promote were reaching inland from Cape Town, Durban,
economic growth and development, just how much Port Elizabeth and East London, whose seaport
infrastructure is optimal? Do different types of infrastructure predated rail by many years.
infrastructure have different effects on the economy?
Should infrastructure be provided exclusively by The discoveries of diamonds in 1867 and gold in 1886
the public sector, or is there a role for the private provided considerable impetus to the development of
sector as well? To the extent that infrastructure is rail. Following the discovery of diamonds, Kimberley
or should be provided by the public sector, is it best grew rapidly in wealth and population, and yet remained
funded through general tax revenues, user charges, constrained by poor transport and communication
foreign direct investment, or borrowing? What are services. Diamonds generated both the need and the
the implications for the environment of different resources for a railway link to Cape Town, a venture

27
p eter p er kins

that was achieved in 1885. Gold played an equivalent role in the case of Johannesburg, which by 1896 was
connected to the four main ports in the Cape and Natal as well as Lourenço Marques (now Maputo). In the
Transvaal Republic, the construction and operation of railway infrastructure was undertaken by the Netherlands
Railway Company until the government took over this role in 1902.

South Africa’s network of railway lines was largely in place by 1930, and the growth in railway infrastructure
thereafter was mainly in the form of rolling stock. Locomotives, coaching stock and goods stock continued to
increase in number quite steadily between 1930 and 1980, after which they fell victim to South Africa’s sharp
fall-off in infrastructural investment (Figure 2). Not only did they fall in number, but their remaining lifespan as
productive assets continued to fall as well, with negative consequences for reliability. The rise and decline of rail
in South Africa is summarised in Table 3.

Table 3 – Rail infrastructure and services (South Africa)

Goods stock
Coaching Goods Passenger Revenue-earning
Railway lines Loco-motives carrying
stock stock journeys traffic
capacity
route km number number number million tonnes million million tonnes
1880 1 621
1910 11 331 1 405 2 071 22 576 0.355 33.7 9.7
1930 18 445 2 193 3 668 37 546 0.783 80.5 20.4
1980 20 353 4 907 10 704 188 799 6.291 691.3 174.9
1995 21 079 3 574 6 740 135 155 6.161 416.0 176.0
2003 20 796 3 253 6 588 114 135 5.593 468.2 179.5
Data sources: Statistics South Africa; Central South African Railways (Report of the general manager of railways); Official Year
Book of the Union; South African Railways and Harbours (Report of the general manager of railways and harbours / Annual
reports); Union Statistics for Fifty Years; Spoornet
The development of rail in 20th-century South Africa was accompanied by the development of roads. The first
trip by road between Cape Town and Johannesburg is reported to have taken place in 1905, an 11-day journey
plagued by dust, pot-holes and farm gates.6 The expansion of national and provincial roads and the traffic they
supported are shown in Table 4.

Table 4 – National and provincial roads, and vehicles (South Africa)

National and provincial National and provincial
Passenger vehicles Goods vehicles
paved and unpaved roads * paved roads *
kilometres kilometres million million
1915 75 279
1930 115 076 0.170 0.016
1940 142 573 2 235 0.341 0.049
1950 145 063 7 057 0.500 0.124
1960 183 316 ** 17 592 1.000 0.212
1970 185 523 33 120 1.674 0.394
1980 183 844 45 948 2.621 0.874
1990 181 290 53 446 3.927 1.273
* Intercity roads (urban roads excluded)
** Includes 25 000 km of roads previously classified as tertiary roads
Data sources: Statistics South Africa; National Traffic Information System; Official Year Book of the Union; South African
National Roads Agency; Union Statistics for Fifty Years

South Africa’s ports and airports handled ever-growing volumes of cargo and passengers during the 20th century.
A summary is provided in Table 5. The sharp increase in cargo handled by the ports in the 1970s resulted in large
measure from the opening of Richards Bay and Saldanha.

28
 T h e r ol e of ec onomic inf r a s t r u c t u r e i n e c o n o m i c g r o w t h : b u i ld i n g o n e x p e r i e n c e

Table 5 – Cargo handled by ports, and air passengers (South Africa)

SAA passengers
Cargo handled International air
(domestic and
by ports passengers
international)
harbour tons (m) million million
1910 4.5
1930 7.2
1950 10.4 0.160
1970 25.4 1.499 0.746
1980 78.7 3.983 1.608
1990 107.5 5.181 1.952
2000 162.7 5.856 5.770
Data sources: Statistics South Africa; National Ports Authority; Official Year Book of the Union;
South African Airways; South African Transport Services (Annual reports)

Telegraphic and telephonic communications were introduced in the 1870s and 1880s.
Twentieth-century progress in the number of fixed (land) telephone lines is shown
in Table 6. Prior to 1991 telephone infrastructure and services were provided by the
Just as the early
Department of Posts and Telecommunications, which then became Telkom, initially development of rail
still wholly owned by the South African government but subsequently listed on the was closely associated
Johannesburg and New York stock exchanges in 2003. Cellular phones appeared in with mining, so too
South Africa in the 1990s, and rapidly overtook the number of fixed phone lines.
The cellular phone market was driven in large measure by private enterprise, though
was the expansion
with Telkom also owning a substantial share of the market. of electricity
generation capacity.
Table 6 – Fixed phone lines and electricity generation (South Africa)
Not only did the
Fixed phone lines Electricity generated
(Telkom) (Eskom and other producers)
mines require
million gigawatt hours large quantities of
1920 0.030 1 277 electricity to operate,
1930 0.070 2 454 but South Africa
1940 0.142 7 168 had vast reserves of
1950 0.277 11 187
1960 0.633 22 561
coal with which to
1970 0.879 50 791 produce electricity at
1980 1.508 98 951 a relatively low cost.
1990 3.080 165 384
2000 5.493 210 577
Data sources: Statistics South Africa; Official Year Book of the Union; Union Statistics for Fifty
Years; Telkom

The bulk of South Africa’s electricity is provided by Eskom (Table 6). Just as the early
development of rail was closely associated with mining, so too was the expansion
of electricity generation capacity. Not only did the mines require large quantities
of electricity to operate, but South Africa had vast reserves of coal with which to
produce electricity at a relatively low cost. Industry and households alike benefited
from a cheap and reliable supply of electricity for decades, and Eskom came to be
regarded by many as a model of what any successful state-owned enterprise should
be (apart, that is, from the lopsided provision of services along racial lines). However,
as we entered the 21st century, there were growing concerns about future generation
capacity, concerns that were compounded by a lack of clarity over the respective
roles of Eskom, various government departments and the private sector in building

29
p eter p er kins

capacity to meet future demand. By early 2008, Eskom’s ability to meet demand had
become severely stretched and the country was subjected to a series of blackouts, a
repeat of which has thus far been averted through demand-side management and a
slowdown in the economy.

Causality between infrastructural investment and economic

growth
If the preceding account of South Africa’s historical development of its infrastructure
creates an impression of good planning, efficient progress and excellence, it should
be hastily dispelled, or at least heavily qualified. The decline in railway infrastructure
after 1980 was increasingly met with complaints of poor and unreliable service, and
much of the freight previously transported by rail (or which was better suited to
rail) was shifted to road. With the additional volumes of freight on the roads, and
growth in road passenger vehicles brought about in part by an inadequate public
transportation system, the roads became increasingly
congested and difficult to maintain in the late 20th
It is difficult to quantify the extent to which century. Congestion at some of the ports became
a growing problem. Before the commercialisation
the provision of economic infrastructure
of Telkom in the 1990s, although the telephone
may promote economic growth, or just how network was generally reliable, customer service (new
much congestion effects in infrastructure installations and repair work) was slow and frustrating.
may hamper economic growth; and equally Electricity fared better, though those chickens, too,
came home to roost in early 2008, along with steep
difficult to quantify the impact of a growing
increases in electricity prices not long after.
economy (with associated growth in tax
revenue) on the provision of infrastructure. It is difficult to quantify the extent to which the
provision of economic infrastructure may promote
economic growth, or just how much congestion
effects in infrastructure may hamper economic growth; and equally difficult to
quantify the impact of a growing economy (with associated growth in tax revenue)
on the provision of infrastructure. Internationally the empirical literature on the
subject has produced mixed results, with some studies showing relatively strong
and positive effects on economic growth from additional infrastructure provision,
but others showing much weaker or even negligible effects (see Perkins et al., 2005).
Before summarising some of the empirical estimates that have been attempted for
South Africa, a simple model of the relationship between growth and infrastructure
may be outlined as follows.

The theoretical case for infrastructural investment having a positive impact on

economic growth is provided by Barro’s growth model7 in which output (per
worker) is a function of both private-sector investment and public-sector provision
of productive services, an important example of which is infrastructural services. The
rationale for treating these expenditures separately is that they are not substitutes.
The large-scale nature of infrastructural projects (with associated high expenditure
and high risk), the difficulty of collecting user charges (not in all but in some cases),
and the presence of positive externalities are all obstacles to the optimal provision
of productive services such as infrastructure if left to the private sector. In Barro’s
model, public-sector services raise the marginal product of private-sector capital,
which in turn raises the rate of economic growth.8

If small or modest amounts of infrastructural investment have a favourable impact

on economic growth, does it follow that ever greater amounts of infrastructural

30
 T h e r ol e of ec onomic inf r a s t r u c t u r e i n e c o n o m i c g r o w t h : b u i ld i n g o n e x p e r i e n c e

investment would have ever more favourable effects on the economy? Most certainly
not. In the Barro model, starting from a relatively low level of productive services
provided by the public sector, economic growth increases as the level of public-
sector productive services increases. However, the improvement in economic growth
is moderated by two effects, namely an ever-decreasing marginal product of public-
sector expenditure, and a tax effect that is growth-negative. Thus there exists an
optimal level of public-sector expenditure on productive services at which the rate
of economic growth is optimised, beyond which the negative tax effect dominates
the productivity effect and the rate of economic growth declines (eventually turning
negative).9

Such a result in theory, namely that there is some

level of infrastructural investment that is optimal The challenge which researchers and policy
for economic growth, should come as no surprise. At
makers face is to estimate just the right type
one end of the spectrum, an economy without power
stations (for example) is unlikely to progress. At the and amount of infrastructural investment
other end, to build more power-generation capacity that should be undertaken, along with
than could ever be used would clearly be wasteful and
appropriate timing. The fact that such
therefore harmful to economic growth. The challenge
which researchers and policy makers face is to estimate decisions must take place in the face of great
just the right type and amount of infrastructural uncertainty about the future makes them all
investment that should be undertaken, along with
the more difficult.
appropriate timing. The fact that such decisions must
take place in the face of great uncertainty about the
future makes them all the more difficult.

The relationship between infrastructural investment and economic growth in South

Africa has been examined by scholars.10 Empirical investigation was undertaken
using long-term time series of national accounts data and individual measures
of infrastructure and infrastructure-related variables, namely those for transport,
communication and power discussed above.

In Perkins et al.11 public-sector investment in economic infrastructure was found to

have a positive effect on GDP growth, as were roads and road passenger vehicles.
GDP growth was found to have a positive effect on railway lines, rail coaching
stock, rail passengers, cargo handled at ports, SAA passengers, and fixed phone
lines. Two-way relationships (or more accurately potential simultaneity) were
found between GDP growth and: rail locomotives, rail goods stock, road goods
vehicles, and electricity generation. More generally, Perkins et al.12 concluded that
‘the relationship between economic infrastructure and economic growth appears
to run in both directions. Economic growth provides both the need for, and the
resources to fund, various types of infrastructure. Provided that infrastructure
projects take place in response to appropriate cost-benefit analyses, they are more
likely to promote GDP growth than hinder it. Alternatively, the failure to provide
appropriate infrastructure services may hamper GDP growth.’

The same data set was investigated in Fedderke et al.13 The main results to emerge
from a variety of specifications were that electricity generation had a positive and
direct impact on GDP and that public-sector infrastructural investment had the
effect of promoting private-sector investment in physical capital, which in turn had a
positive impact on GDP. ‘The empirical results are generally supportive of the South
African fiscal authorities’ renewed interest in public-sector investment since 2002’14.

31
p eter p er kins

Conclusion
Over the long term, namely the half century 1960–2009 shown in Figure 2, public-
sector investment in economic infrastructure averaged approximately four per
cent of South Africa’s GDP. Considering the amount of economic activity which
that four per cent facilitated and supported, and apart from investment in social
infrastructure such as health and education, it would be difficult to find another
category of expenditure of comparable size that is of equal importance in promoting
economic growth.

Nevertheless, investments in infrastructure must be undertaken with care.

Constructing a highway or railway between two previously unconnected centres
of economic activity may reasonably be expected to have a beneficial impact on
economic growth. Linking two uninhabited areas with no prospects for economic
development would simply be wasteful. In other words, the growth-enhancing
effects of economic infrastructure are not automatic; they must be considered with
care in each situation, and weighed against the costs.

Since the mid-1990s, there has been growing

Since the mid-1990s, there has been growing recognition of the importance of investing in economic
infrastructure in South Africa. The extended period of
recognition of the importance of investing in decline that started in the early 1980s (or even before)
economic infrastructure in South Africa. The was brought to an end, the level of investment was
extended period of decline that started in the early stabilised, and between 2002 and 2009 there was an
1980s (or even before) was brought to an end, the upsurge across a range of investment types. Highways
have been expanded and upgraded country-wide; the
level of investment was stabilised, and between capacity of the ports has been increased, including the
2002 and 2009 there was an upsurge across a new port of Ngqura (Coega) in the Eastern Cape;
range of investment types. major airports have been transformed, with Durban
getting an entirely new airport; in Gauteng a high-
speed rail service between O.R. Tambo International
Airport and Sandton was completed in 2010, with further extensions to be completed
in 2011; bus rapid transport systems have been implemented; and Eskom has
embarked on a long-term investment programme to build new electricity generation
capacity. In communications, cellular phones and the internet have arguably been
the outstanding feature of economic infrastructural development since the late
1990s, not only in South Africa but across the globe. The private sector has played
a leading role in the information and communication technology revolution, with
governments playing an important regulatory role.

Much remains to be done. In electricity alone the demands are enormous. Eskom
plans to double its generation capacity from approximately 40 000 MW in 2010 to
80 000 MW in 2026. Considering that a substantial portion of the existing 40 000
MW capacity will expire before then (power stations depreciate), this is a daunting
prospect.

If South Africa’s economy is to move to a high-growth path, sufficiently high

to bring down unemployment in any meaningful way, sustained investment in
its economic infrastructure will be integral to achieving that outcome. Given
the multitude of demands on the public purse and the temptation to neglect
infrastructural investment when times are tough, and the difficulty of taking long-

32
 T h e r ol e of ec onomic inf r a s t r u c t u r e i n e c o n o m i c g r o w t h : b u i ld i n g o n e x p e r i e n c e

term decisions in the presence of enormous uncertainty, this is easier said than
done. But it cannot be avoided. Beyond recognition of the importance of economic
infrastructure for economic growth, crucial to the continued renewal and expansion
of our economic infrastructure will be careful assessment based on a number of
issues. These include cost-benefit analysis; efficient spending of available funds and
effective implementation of projects from beginning to end; finding sustainable
solutions to funding shortfalls; public consultation and debate; building partnerships
between the private and public sectors; co-ordination between all participants and
other stakeholders, all of whose roles should be made clear; and above all, clear
planning and strong leadership. The objective for policy makers across all levels of
government must be to plan for and provide economic infrastructure that is not too
much, not too little, but just the right amount, just the right type, and at just the
right time and just the right price. It is a difficult challenge indeed.

notes
1 2010a
2 2010b
3 2010
4 Gross fixed capital formation.
5 An important, or rather crucial, private-sector contribution to South Africa’s infrastructural development that ought to be
acknowledged is that of the private-sector civil engineering and construction firms contracted by the state to undertake the
work of putting infrastructure in place.
6 Van Lingen et al., 1960: 141
7 Barro, 1990
8 See Fedderke et al. (2006) for a detailed explanation of the Barro model.
9 In a Cobb-Douglas representation of the Barro model with constant returns to scale, the optimal point is reached when the
marginal product of public-sector expenditure falls to one.
10 Perkins et al. (2005) and Fedderke et al. (2006)
11 (2005). PSS (Pesaran, Shin and Smith) F-tests were used to identify directions of association between economic infrastructure
and GDP.
12 2005: 223
13 (2006) using a vector error-correction mechanism framework.
14 Fedderke et al., 2006: 1052

References
Barro, RJ. 1990. ‘Government spending in a simple model of endogenous growth.’ Journal of Political Economy, 98(5):
S102–S125.
Fedderke, JW, Perkins, P and Luiz, JM. 2006. ‘Infrastructural investment in long-run economic growth: South Africa 1875-2001.’
World Development, 34(6): 1037-1059.
Krugman, P. 2010a. ‘America Goes Dark.’ The New York Times, 8 August. [Link]
opinion/[Link]?ref=paulkrugman, accessed 20 December 2010.
Krugman, P. 2010b. ‘The End of the Tunnel.’ The New York Times, 7 October. [Link]
opinion/[Link]?_r=1&ref=paulkrugman, accessed 20 December 2010.
Perkins, P, Fedderke, JW and Luiz, JM. 2005. ‘An analysis of economic infrastructure investment in South Africa.’ South African
Journal of Economics, 73(2): 211-228.
The Economist, 2010. ‘And now for the good news.’ 30 October: 37.
Van Lingen, A et al. 1960. A century of transport, 1860–1960: a record of achievement of the South African rail, road, air and
harbour services. Johannesburg: Da Gama Publications.

33