Jump to ContentJump to Main Navigation
Manufacturing TransformationComparative Studies of Industrial Development in Africa and Emerging Asia$

Carol Newman, John Page, John Rand, Abebe Shimeles, Måns Söderbom, and Finn Tarp

Print publication date: 2016

Print ISBN-13: 9780198776987

Published to Oxford Scholarship Online: August 2016

DOI: 10.1093/acprof:oso/9780198776987.001.0001

Show Summary Details
Page of

PRINTED FROM OXFORD SCHOLARSHIP ONLINE (www.oxfordscholarship.com). (c) Copyright Oxford University Press, 2019. All Rights Reserved. Under the terms of the licence agreement, an individual user may print out a PDF of a single chapter of a monograph in OSO for personal use (for details see www.oxfordscholarship.com/page/privacy-policy).date: 24 May 2019

The Pursuit of Industry

The Pursuit of Industry

Policies and Outcomes

Chapter:
(p.1) 1 The Pursuit of Industry
Source:
Manufacturing Transformation
Author(s):

Carol Newman

John Page

John Rand

Abebe Shimeles

Måns Söderbom

Finn Tarp

Publisher:
Oxford University Press
DOI:10.1093/acprof:oso/9780198776987.003.0001

Abstract and Keywords

This chapter looks at one of the most significant changes in the global economy over the last forty years, the shift of industry from high-income countries to developing countries. Between 1992 and 2012, the share of world manufacturing output produced by developing countries nearly doubled, rising to more than a third of global production. Growth of manufactured exports has greatly exceeded the growth of manufacturing output, and developing countries have gained world market shares in both simple and complex manufactured products. Today, East Asia, led by China, produces more than one-fifth of global manufacturing value added. Africa’s experience with industrialization over the same period has been disappointing. In 2010 the average share of manufacturing in GDP in sub-Saharan Africa was 10 per cent, unchanged from the 1970s.

Keywords:   Africa, industry, developing countries, manufacturing output, global production

1.1 Introduction

One of the most significant changes in the global economy over the last forty years is the shift of industry from high income to developing countries. Between 1992 and 2012, the share of world manufacturing output produced by developing countries nearly doubled, rising to more than a third of global production. Growth of manufactured exports has greatly exceeded the growth of manufacturing output, and developing countries have gained world market shares in both simple and complex manufactured products. Today, East Asia, led by China, produces more than one-fifth of global manufacturing value added (UNIDO 2013).

Africa’s experience with industrialization over the same period has been disappointing. In 2010 the average share of manufacturing in GDP in sub-Saharan Africa (SSA) was 10 per cent, unchanged from the 1970s. Africa’s share of global manufacturing has fallen from about 3 per cent in 1970 to less than 2 per cent in 2010. Manufacturing output per person is about a third of the average for all developing countries and manufactured exports per person are about 10 per cent of the global average for low income countries (Page 2012).

This book contains eleven country studies—eight from SSA, one from North Africa, and two from newly industrializing East Asia—sponsored by the African Development Bank, the Brookings Institution, and UNU-WIDER under a research project called Learning to Compete. The project was undertaken in an effort to understand why Africa has so little industry. The research (p.2) programme sponsored the eleven country studies that appear in this volume and an additional twenty quantitative studies based on firm-level data from the same countries.1 The country studies were undertaken to provide a comparative perspective on the role that public policy has played in Africa’s industrialization.

Table 1.1 The Africa country cases in context, 2013

GDP per capita

GDP per capita rank

Population share

GDP share

Share of manufacturing value added

GDP growth 1995–2012

Year of independence

Ethiopia

1,354

36

11.26

4.57

1.22

7.75

1941

Ghana

3,974

13

3.10

3.33

1.55

6.06

1957

Kenya

2,265

22

5.31

4.52

3.16

3.71

1963

Mozambique

1,045

40

3.09

1.89

1.52

7.37

1975

Nigeria

5,601

12

20.78

30.78

6.63

6.44

1960

Senegal

2,269

21

1.69

1.91

1.66

4.04

1960

Tanzania

1,775

26

5.89

4.03

2.47

6.06

1961

Uganda

1,410

33

4.50

2.62

1.27

6.96

1962

Note: Column 2 is rank out of 46 SSA countries excluding South Africa. Somalia and Réunion are also excluded from the computation of total GDP due to a lack of data. Columns 3, 4, and 5 are computed as the share of the total for sub-Saharan Africa excluding South Africa. Total population for SSA excludes Réunion, total GDP for SSA excludes Somalia and Réunion, and the total manufacturing value added for SSA excludes Cape Verde, Equatorial Guinea, Réunion, and Somalia due to a lack of data.

Source: World Bank (2014) World Development Indicators; UNIDO Statistics (2014) Manufacturing Value Added Database; authors’ calculations.

The eight SSAn countries included in this volume—Ethiopia, Ghana, Kenya, Mozambique, Nigeria, Senegal, Tanzania, and Uganda—are some of the stars of Africa’s growth turnaround. Six of the eight have been among its fastest growing economies since 2000. Together they represent 54 per cent of the region’s GDP and 56 per cent of its population (Table 1.1). Their manufacturing sectors make about one-fifth of SSA’s manufacturing value added (excluding South Africa). These are Africa’s emerging markets, but they are not emerging industrial economies. Senegal has the highest share of manufacturing in GDP at about 18 per cent. Nigeria has the lowest at less than two per cent. On average they are quite similar to Africa as a whole. Their share of manufacturing in GDP is 9.5 per cent, and the policies all eight countries adopted for industrial development closely parallel those of the region more broadly.

Table 1.2 Structural characteristics of the case study countries

Country

Share of labour force

Share of GDP

Agriculture

Manufacturing

Other industry

Services

Agriculture

Manufacturing

Other industry

Services

Cambodia 1994 and Vietnam 1990 (US$800–1,200)

74.0

7.1

2.9

16.0

43.5

11.4

7.6

37.5

Ethiopia 2005

75.1

6.3

2.5

16.1

45.3

4.2

6.2

44.3

Mozambique 2010

71.0

3.6

0.7

24.7

24.2

14.2

10.8

50.8

Cambodia 2003 and Vietnam 1992 (US$1,200–2,000)

68.5

8.5

2.0

21.0

34.0

15.9

10.6

39.5

Tanzania 2010

73.0

2.6

3.3

21.1

27.6

9.8

14.8

47.8

Uganda 2009

69.3

3.9

0.5

26.3

23.4

9.4

10.4

56.6

Cambodia 2010 and Vietnam 2000 (US$2,000–3,000)

59.5

9.5

2.0

22.3

29.5

17.3

9.2

44.0

Ghana 2010

42.0

10.8

4.5

42.7

29.7

6.7

12.2

51.4

Kenya 2010

48.4

12.8

3.6

35.2

24.8

11.2

7.2

56.8

Senegal 2010

51.5

9.9

4.0

34.6

17.4

18.4

9.7

59.5

Vietnam 2010 (US$5,000–)

51.0

13.5

0.9

30.5

19.0

21.9

16.1

43.0

Nigeria 2010

58.8

4.0

2.0

35.2

30.4

1.9

44.4

23.3

Tunisia 2010

18.0

30.0

3.0

49.0

Sources: World Bank (2014) World Development Indicators; McMillan and Rodrik (2011) Database; Groningen Africa Sector Database (2013).

The comparator countries took quite different policy approaches to industrialization and had very different industrial development outcomes. Tunisia (p.3) in North Africa gained its independence at about the same time as SSA, but its approach to industrialization diverged early on from that found south of the Sahara. Its manufacturing growth has exceeded that for SSA for three decades. Cambodia and Vietnam had per capita income levels and structural characteristics similar to African economies as recently as 2005 in Cambodia and 2001 in Vietnam (Table 1.2). After an early period of state-led industrialization, both countries have followed industrial development strategies that are similar to those of other emerging East Asian countries with considerable success. Since 1990 manufacturing growth has averaged more than 10 per cent per year in both countries.

The chapter is organized in the following way. Section 1.2 makes the case for renewed attention to industrialization in Africa. Sections 1.3 and 1.4 provide overviews of industrial development policies and outcomes in Africa and in emerging Asia. Section 1.5 presents comparative results from the country studies in four areas identified in the literature as key drivers of industrial development: infrastructure, human capital, and institutions; manufactured exports; agglomerations; and foreign direct investment (FDI). Section 1.6 concludes.

1.2 Industrialization, Structural Transformation, and Growth

One of the enduring ‘stylized facts’ of economic development is that low income countries have large differences in output per worker across sectors. Structural transformation—the shift of resources from lower productivity to higher productivity sectors—is, therefore, often a key driver of growth (Lewis 1954; Chenery 1986). During the course of structural transformation, economy-wide changes in output per worker are the result of labour productivity growth within individual sectors and the change in labour productivity due to labour reallocations across sectors.2 When strong within sector productivity growth combines with rapid movement of labour into higher productivity sectors—the pattern of structural transformation seen in East Asia over the last twenty-five years—very rapid growth of output per worker is possible (McMillan and Rodrik 2011).

Because it has the greatest differences across sectors in output per worker, Africa is the developing region with the most to gain from structural transformation. Recent research, however, finds that this potential has not been fully realized. In fact from 1990 until 1999 structural transformation was ‘growth reducing’. Africa’s higher productivity sectors failed to generate (p.4) (p.5) enough jobs to absorb a rapidly growing labour force, and the share of workers employed in high productivity sectors declined, offsetting positive within sector productivity growth and reducing aggregate growth of output per worker (de Vries et al. 2013; McMillan et al. 2014).

While labour in Africa began to move out of agriculture into more productive employment after 2000, the movement that has taken place has been into ‘market services’, mainly retail trade and distribution (de Vries et al. 2013). This structural shift from agriculture to services differs from the development experience of other regions, and it represents movement from very low productivity to slightly higher productivity jobs. Output per worker in services in Africa is only about two times higher than output per worker in agriculture. Moreover, services have been absorbing workers faster than output in the sector has been increasing. The relative productivity of African market services fell from 3.0 times the economy-wide average in 1990 to 1.8 in 2010, suggesting that the marginal productivity of new services workers is low and possibly negative (de Vries et al. 2013).

Historically, industry is the sector into which resources have first moved in the course of structural transformation (Chenery 1986). Between 1950 and 2006, about half of the catch-up by developing countries to advanced economy levels of output per worker was explained by rising productivity within industry combined with structural transformation out of agriculture (Duarte and Restuccia 2010). Industry is the pre-eminent destination sector at early stages of development because it is a high productivity sector capable of absorbing large numbers of moderately skilled workers. Average labour productivity in manufacturing in Africa is more than six times that in agriculture (McMillan and Hartgen 2014). It is also a powerful engine of within sector productivity growth. There is evidence that modern manufacturing industries—unlike agriculture or services—converge to global best practice productivity levels ‘unconditionally’, regardless of geographical disadvantages, poor institutions, or bad policies (Rodrik 2013).

Developing countries with the most rapidly sustained growth have typically reallocated the most labour into manufacturing. Where manufacturing has stagnated and structural transformation has mainly involved reallocation of workers into lower productivity sectors, aggregate per capita income growth has lagged (Duarte and Restuccia 2010). This is not good news for Africa: the share of manufacturing in GDP for Africa as a whole has been declining since the mid-1970s. Many of the region’s recent growth success stories—Ethiopia, Ghana, Kenya, Tanzania, and Uganda for example—have shares of manufacturing in GDP that are well below the values predicted by their GDP per person (Dinh et al. 2012). Lack of industrial development combined with declining within sector productivity in services raises the risk that structural transformation in Africa may once again become growth reducing.

(p.6) 1.3 Industrial Policies and Outcomes in Africa: From a Dominant State to the Investment Climate

Despite major differences in history, language, and political structure, the SSAn countries in this volume share a striking uniformity in their approach to industrialization. Broadly, their policies to encourage industrial development fall into three phases: state-led; import substitution; the Washington Consensus and structural adjustment, and reform of the ‘investment climate’, the physical, institutional, and regulatory environment for firms. Industrial performance has largely followed three phases as well: an early boom, decline, and stagnation.

1.3.1 State Ownership and Import Substitution, 1960–85

When Africa gained independence, post-colonial governments saw industrialization as a key driver of economic development. Leaders in English-, French-, and Portuguese-speaking Africa shared similar views, strongly shaped by a desire to modernize their mainly agrarian economies and reduce dependence on the former colonial powers.3 The centrepiece of the industrialization effort was the development of large-scale, often capital-intensive manufacturing industries owned and managed by the state. Protection of the domestic market against imports was viewed as necessary and was particularly appealing to post-colonial leaders as a way of securing ‘economic independence’. The state became the central actor in the industrialization story for a variety of reasons. Political ideology and nationalism were certainly key motivations (Ndulu 2007). Newly independent African governments invested heavily in manufacturing, setting up state-owned enterprises (SOEs) for the domestic production of consumer goods, building materials, and the processing of primary products.

1.3.1.1 Early Industrialization

Arguably, Ghana set the tone. When it gained independence from Britain in 1957, import substitution, supported by high levels of protection, was used to create a new industrial structure and reduce dependence on Britain. The state invested heavily in the manufacturing of consumer and producers goods. The electrical, electronic, and machinery industries in particular were viewed as essential to provide the inputs needed to industrialize further. Other countries followed similar industrialization strategies upon gaining independence. Senegal pursued state-led import substituting industrialization in the late (p.7) 1960s. Tanzania’s 1967 Arusha Declaration stressed socialism and self-reliance. Major industrial firms were nationalized and the public sector assumed a leading role. Industrialization and state ownership featured prominently in Uganda’s Second Five-Year Plan. Nigeria’s growing revenues from oil exports led the government to undertake ambitious investments in iron and steel, cement, salt, sugar, fertilizer, and pulp and paper.

Kenya adopted a post-independence industrialization strategy in the 1960s that relied on import substitution, but state ownership and management of the industrial sector was limited to a few ‘strategic industries’. Two governments followed more explicitly central planning approaches. In Ethiopia the Marxist Dergue government nationalized most privately owned medium- and large-scale manufacturing enterprises and intensified the import substitution strategy that had been in place since the time of the emperor. Beginning in 1975 the Frelimo government in Mozambique introduced a set of policies designed to make the public sector the dominant economic actor.

Tunisia embraced import substitution and state ownership at the beginning of the 1960s, but by 1970 industrial growth had proved disappointing. In response the government adopted an infitâh policy that combined import substitution and export promotion.4 The economy was divided into an ‘offshore’ sector geared towards exports and an ‘onshore’ sector that was shielded from competition and regulated by the state. Institutional and policy reforms were introduced to attract FDI, mainly from Europe. The offshore sector was dominated by foreign owned firms engaged in assembly-type manufacturing for export to the European market. The onshore private sector primarily consisted of small factory units producing simple consumer goods under high protection. Heavy industry, transport, water, and electricity were state owned.

1.3.1.2 The Industrialization Drive Falters

Table 1.3 Average annual growth of value added in manufacturing, 1965–2010

1965–70

1970–5

1975–80

1980–5

1985–90

1990–5

1995–2000

2000–5

2005–10

Ethiopia

8.45

3.20

4.56

3.91

1.51

1.03

3.92

5.00

9.48

Ghana

8.22

1.43

−4.74

−4.36

7.53

−7.38

4.68

4.54

2.60

Kenya

7.43

7.67

11.46

3.84

5.75

2.52

−0.03

3.09

4.38

Mozambique

−1.31

18.97

15.17

3.01

Nigeria

5.52

12.20

13.59

−0.99

4.10

−1.07

0.25

8.85

8.43

Senegal

3.17

1.27

2.64

4.05

3.19

2.90

3.11

2.02

Tanzania

9.95

4.73

2.36

−5.01

2.43

−0.02

5.73

8.09

8.61

Uganda

2.14

6.38

12.81

13.45

6.13

7.13

Tunisia

−0.74

20.50

13.63

6.39

0.50

5.66

5.78

2.79

4.88

Cambodia

8.86

21.04

13.86

8.70

Vietnam

9.50

2.42

10.35

11.26

11.66

9.33

Notes: All growth rates in constant prices. Blank cells indicate that data are not available.

Sources: 1965–90: de Vries et al. (2013); McMillan and Rodrik (2011); 1990–2010: UNIDO Statistics (2014) Manufacturing Value Added Database; Uganda 1980–90, Tunisia 1965–2010, Vietnam 1985–90; World Bank (2014), World Development Indicators; authors’ calculations.

Africa’s state-led push for industrial development had considerable success in the 1960s. Manufacturing grew substantially faster than overall output between 1960 and 1970 and the share of manufacturing in total output increased. Between 1965 and 1970 industrial growth averaged more than 7 per cent per year in all of the African case study countries (except Mozambique which was still in a fierce liberation war). By 1970, however, the industrialization drive was beginning to lose steam, and by 1975 growth of the manufacturing sector had begun to lag GDP growth (Table 1.3). The growth deceleration was particularly sharp in Ethiopia, Ghana, Senegal, and Tanzania. (p.8)

(p.9) By the 1980s the state-led industrialization effort had reached its limits in most countries. Between 1980 and 1985 manufacturing output began to decline in Ghana, Nigeria, and Tanzania. Contrary to the intent of the import substitution strategy, dependence on imports actually increased due to the heavy reliance of industry on imported capital and intermediate goods. Public investment exceeded the fiscal capacity of the state, and the state’s capacity to manage the enterprises. The efficiency of production, measured in terms of international prices, was low, and in some cases goods were produced at negative value added in international prices. There was substantial excess capacity in public manufacturing enterprises, many of which were heavily constrained by lack of imported intermediates and working capital.

1.3.2 The Washington Consensus, 1985–2000

In the early 1980s Latin America—the intellectual home of import substitution and its most important laboratory—was engulfed in macroeconomic turbulence. In response the international financial institutions (IFIs) developed a new set of policy-based lending instruments.5 These ‘structural adjustment loans’ (SALs) provided quick disbursing funds for budget and balance of payments support in return for an agreed set of economic reforms. A consensus emerged—at least among the US Treasury, the Federal Reserve, the International Monetary Fund (IMF), and the World Bank—on a set of ten policy reforms considered appropriate for developing countries. Williamson (1990) famously termed these the ‘Washington Consensus’. Six of the ten directly affected the policy environment for industry: liberalization of interest rates; liberalization of exchange rates; liberalization of trade; liberalization of foreign investment; privatization; deregulation.

1.3.2.1 Structural Adjustment

The Washington Consensus quickly found its way to Africa. The same external shocks that struck Latin America had left Africa’s economies with flagging economic growth and chronic foreign exchange shortages. Governments attempted to sustain growth through expansionary macroeconomic policies leading to widespread loss of fiscal and monetary control. Exchange rates became seriously overvalued, and most governments responded to the lack of foreign exchange by introducing exchange controls and rationing. According to one estimate, the median African currency was 82 per cent overvalued in purchasing power parity (PPP) terms in 1980 (Easterly 2009). Growth ground to a halt and Africa entered structural adjustment. Between 1985 and 2000 more (p.10) than thirty African countries, including all those in the country studies, would undertake structural adjustment programmes (World Bank 2000).

The initial focus of public policy advice and conditionality by the IFIs in Africa was on macroeconomic stabilization (World Bank 1992). Policy changes designed to improve resource allocation—liberalization of trade and finance and regulatory reform—followed closely thereafter. Across the continent governments liberalized trade and engaged in some deregulation of the domestic market. Privatization became a major objective and was often pushed, even in weak regulatory environments (Megginson and Netter 2001). Divestiture of state-owned enterprises was viewed as important both because it reduced the drain on the budget imposed by poor investment choices and because the state had proved to be a poor entrepreneur (Nellis 1986).

The reform programmes eventually restored macroeconomic balance. Fiscal deficits in the thirty-one countries covered by the Special Programme of Assistance for Africa had dropped to an average of 5.3 per cent of GDP in 1997 (World Bank 2000). The currency in the median African country was at PPP parity or undervalued in the early 1990s, and the black-market premium for foreign exchange had virtually disappeared (Easterly 2009). Quantitative trade restrictions were replaced by tariffs, and trade weighted average tariff rates fell from 30–40 per cent in 1980 to 15 per cent or less by 2000 (World Bank 2000). Privatization was more controversial and less widely embraced. In many countries the principal motivation to privatize was to placate the IFIs (Nellis 2003).

1.3.2.2 A Short-Lived Industrial Recovery

Perhaps no episode in Africa’s contemporary economic history has raised as much debate as structural adjustment. The dramatic about-face in economic policies and more than a decade of very poor development outcomes sparked considerable academic and popular criticism.6 The early policy adjustments in combination with increased inflows of foreign aid provided a stimulus to industrial production in some countries, as firms used capacity that had been heavily constrained by lack of imported intermediates. Between 1980 and 1985 and 1985 and 1990 manufacturing growth shifted from negative to positive in Ghana, Nigeria, and Tanzania, and accelerated in Kenya, Senegal, and Uganda while it fell in Ethiopia and Tunisia.

The partial recovery of manufacturing was short-lived, however. Increased competition from imports and rising production costs due to reforms in the foreign exchange and financial markets put considerable pressure on manufacturing enterprises. Import competition, lack of technical expertise, and the (p.11) shortage of working capital resulted in most government-owned firms operating at as little as 10 per cent of capacity. By 1990–5 manufacturing output was falling in Ghana, Mozambique, Nigeria, and Tanzania, and growth of manufacturing had declined in every other country except Uganda. The textile and clothing sector was especially hard hit. In Tanzania twenty-two out of twenty-four textile factories had closed by 1990, and in Nigeria employment in the textile and garments sector fell from 700,000 in 1980 to 40,000 in 1995. Tunisia in contrast maintained manufacturing growth rates of more than 5.5 per cent per year throughout the 1990s, despite embarking on its own structural adjustment programme.

1.3.3 Investment Climate Reform and New Directions, 2000–

Africa entered the twenty-first century in substantially better macroeconomic shape than it had been in the last decades of the twentieth. The region began to experience positive per capita income growth around 1995, a trend that would accelerate through the 2000s. Improved economic performance led to a retreat from structural adjustment lending, and the Millennium Development Goals set a new agenda for aid to Africa, one mainly centred on human development.

1.3.3.1 Investment Climate Reforms

In the area of industrial development, the World Bank and many bilateral donors shifted their focus after 2000 to the ‘investment climate’. As defined by the World Bank, the investment climate included: (1) macroeconomic stability; (2) openness; (3) good governance and strong institutions; (4) the quality of the labour force and infrastructure (Stern 2001, 2002). Led by the donors, investment climate reforms became widespread, often becoming key components of budget support programmes. Around one-quarter of official development assistance (some US$21 billion per year) currently supports investment climate reforms (OECD 2014).

The sub-Saharan case-study countries have all undertaken investment climate reform programmes in the last decade. Ghana has focused on trade policy and regulatory reforms. In Kenya reforms were undertaken to liberalize the regulatory regime. Mozambique adopted a new Industrial Policy and Strategy in which a significant role was assigned to promoting private investment. Nigeria’s 2004 National Economic Empowerment and Development Strategy was explicitly targeted at making the industrial sector internationally competitive. The 2005 Senegal Accelerated Growth Strategy set as its main objective establishing a ‘business environment consistent with international good practice’. In 2010 Tanzania introduced an Integrated Industrial Development Strategy aimed at creating a ‘competitive business environment’.

(p.12) 1.3.3.2 New Directions

In addition to implementing the investment climate reform agenda, a number of countries have adopted more activist approaches to industrial development. In 1998 the Ethiopian government launched a strategy aimed at promoting labour-intensive manufactured exports. Kenya’s Vision 2030 also emphasizes manufactured export growth. Most of the region’s strategy and planning documents list a range of instruments intended to encourage private investment in targeted sectors. Ethiopia has attempted to coordinate private investment in textiles and garments, meat, leather and leather products, and agro-processing industries. Ghana’s national industrial policy includes a number of highly sector-specific objectives. A prominent feature of Mozambique’s industrialization strategy has been the promotion of large mining, manufacturing, and energy projects, known as ‘mega-projects’.

Tunisia was the only African country studied in which the government undertook initiatives aimed at improving the competitiveness of individual industries and enterprises.7 An industrial upgrading programme, Programme de mise à niveau, was launched in 1996 followed by the Industrial Modernization Programme (PMI). These programmes were intended to provide technical assistance, training, financial subsidies, and infrastructure upgrades for firms to help them face international competition arising from the preferential trade agreement with the European Union under the Euro-Med initiative.

1.3.3.3 Not Yet A Turning Point

For Africa as a whole neither the widespread adoption of investment climate reforms nor the new directions taken by some governments have reversed the four decade decline in industry. Manufacturing growth has remained below the growth rate of GDP. Since 2000 industrial performance among the SSAn countries covered in the country studies has been uneven. There has been some acceleration in the growth of manufacturing in Ethiopia, Kenya, Tanzania, and Uganda (Table 1.3). Manufacturing growth in Ghana and Senegal has remained low and has lagged behind the overall growth of the economy. Nigeria was an exception; manufacturing grew at about 8 per cent per year between 2000 and 2010.

Table 1.4 Manufacturing value added per worker in sub-Saharan African countries, 1995–2010

1995

2000

2005

2010

Ethiopia

100

85

74

64

Ghana

100

123

123

123

Kenya

100

65

53

56

Mozambique

100

36

Nigeria

100

139

192

267

Senegal

100

82

73

66

Tanzania

100

100

101

107

Uganda

100

87

130

Note: Indexed to 100 in first year of series.

Sources: de Vries et al. (2013); McMillan et al. (2014); authors’ calculations.

Table 1.4 shows trends in value added per worker in manufacturing between 1995 and 2010.8 Only Nigeria has registered rapid long-run manufacturing labour productivity growth. In Ethiopia, Kenya, and Senegal output per (p.13) worker declined between 1995 and 2010. In Ghana and Tanzania productivity growth was near zero. Uganda had a manufacturing productivity growth rate of about 2.7 per cent. In Tunisia growth in value added per worker in manufacturing averaged about 5 per cent per year between 1985 and 2005 (Ben Jelili and Goaied 2009).

1.4 Policies and Outcomes in Emerging Asia

Like Africa, Cambodia and Vietnam reacted to their colonial history by embracing state-led industrial development. In both cases early industrialization efforts were derailed by political instability, conflict, and war. Both countries also went through important macroeconomic adjustments in the 1990s accompanied by episodes of liberalization, and in the second half of the 1990s Cambodia and Vietnam followed the examples of other Asian economies by introducing policy and institutional reforms and public investments aimed at promoting manufactured exports.

1.4.1 Early Industrialization, 1960–85

When Cambodia gained independence from France in 1953, it focused on building an industrial base through a mixed-market model. Following a period of civil war and political instability, two Marxist regimes—the Khmer Rouge and the People’s Republic of Kampuchea (PRK)—held power from 1975 to 1989. The PRK regime collectivized agriculture, nationalized all sectors of the economy, and adopted a policy of self-reliance. International trade, except with a few allied communist countries, was banned. The state controlled finance, transport, and foreign commerce. By the time the PRK left power in 1989 the Cambodian economy had been reduced to subsistence agriculture. (p.14) Industry was decimated; only about 1 per cent of manufacturing establishments had survived.

Vietnam’s approach to industrialization did not begin very differently from that of Cambodia. The country’s First Five-Year Plan (1961–5) prioritized heavy industry and state entrepreneurship. The plan was disrupted in 1964 by the US–Vietnam war, and when the war ended in 1975, the government adopted a unified and centrally planned economic system. All industrial firms became state-owned enterprises, governed directly by ministries and provincial authorities. Input and output levels in manufacturing were set administratively. Between 1980 and 1985, industrial production increased at an annual rate of 9.5 per cent, but the economy was burdened by low labour productivity, rising inflation, and high unemployment. By 1985 Vietnam was on the brink of a socio-economic crisis.

1.4.2 Structural Reform and the Export Push, 1985–

Vietnam’s comprehensive reform programme, Doi Moi (renovation), began in 1986. Doi Moi was intended to create a ‘socialist-oriented market economy’ that would combine state leadership (mostly at the planning stage) with market incentives. Private businesses and foreign owned enterprises were encouraged. The new policy framework included reform of SOEs and a law on foreign investment. In 1989 Vietnam adopted a radical and comprehensive reform package aimed at stabilizing and opening up the economy, increasing the autonomy of firms, and promoting competition.

One of the most important policy decisions that Vietnam made under Doi Moi was to shift from import substitution to an export-oriented industrialization strategy. Macroeconomic and trade policy reforms were targeted at reducing anti-export bias. There was a gradual reduction in export and import controls. Investment loans, post-investment interest subsidies, and export credit guarantees were introduced, together with short-term credit for exporters. Export Processing Zones and Industrial Zones were used to provide the infrastructure needed for new export industries. The Ministry of Planning and Investment was made responsible for the implementation of investment policy, including coordination with the private sector through its Foreign Investment Agency (FIA), Enterprise Development Agency, and Department of Economic Zones.

Spurred by the fall of the Soviet Union and the Doi Moi reforms in Vietnam, Cambodia began a gradual process of economic reform in the early 1990s. The most significant reforms included gradual privatization of state-owned companies and the de-collectivization of agriculture. Trade policies were liberalized and most quantitative restrictions and import licences were eliminated. In the late 1990s Cambodia took a set of steps towards industrial (p.15) export promotion. Barriers to export and to the import of intermediate inputs were reduced or eliminated. Macroeconomic policy was aligned with the objectives of attracting foreign investment and export growth. The government provided generous incentives to FDI, and investments made within special economic zones received favourable treatment. A government–private sector forum was launched to include the private sector in planning and managing the economy.

The response to the reforms was dramatic. In Vietnam manufacturing grew at over 11 per cent per year during the period 1995–2005 and at more than 9 per cent between 2005 and 2010. The average annual growth rate of manufacturing in Cambodia was about 6 per cent per year during 1990–5. It accelerated to 21 per cent per year during 1995–2000. Between 2000 and 2010 manufacturing growth averaged more than 10 per cent per year. During 2000–10, manufacturing productivity growth averaged 2.5 per cent per year in Cambodia and 5 per cent per year in Vietnam (APO 2013).

1.5 Policy and Performance: A Comparative Framework

The literature on industrialization in developing countries suggests that four factors have largely shaped the global distribution of industry through their impact on productivity at the firm level.9 The first is the presence or absence of some ‘basics’: infrastructure, human capital, and institutions. The second is the growth of manufactured exports. The third is industrial agglomerations, and the fourth is FDI. Each of these areas is affected by public policy choices. This section presents comparative evidence from the country studies on relative performance in Africa and emerging Asia with respect to each productivity driver.

1.5.1 The Basics

Initial conditions matter for industrial development. Cross-country evidence shows that a variety of country‐specific factors, including basic infrastructure and human capital, financial depth, and barriers to entry are correlated with industrial development and diversification in low income countries (IMF 2014). Better and more reliable electrical power, lower costs of transport, and workers who are better able to perform their jobs raise the potential productivity of all firms in an economy. A number of econometric studies highlight the productivity penalty that enterprises pay as a result of poor infrastructure (p.16) and skills (Escribano et al. 2010; Foster and Briceño-Garmendia 2010). Regulatory burdens and poorly functioning institutions inhibit productivity growth by reducing the scope for competition.10

Twenty years of stabilization, fiscal austerity, and slow growth left Africa with very large gaps in infrastructure and human capital. Africa started out in the 1960s with stocks of roads that were generally not very different from those in East Asia. The same was true in the 1970s for telephones and in the 1980s for power. By around 2000 it trailed in every infrastructure category (Foster and Briceño-Garmendia 2010). Today, SSA lags at least 20 percentage points behind the average for low income countries on almost all major infrastructure measures (World Bank 2009). In addition the quality of service is low, supplies are unreliable, and disruptions are frequent and unpredictable.

Power emerges as a major constraint in every SSA country case study. Self-reported losses associated with power outages amounted to more than 10 per cent of sales in some countries. Uganda has one of the lowest per capita electricity consumption levels in the world. Manufacturing firms in Tanzania experienced on average almost nine power outages per month, costing about 15.1 per cent of total sales for the firms affected. Transport finishes a close second. Around one-third of firms cite transportation as a major or severe constraint. Ugandan firms on average lost 1.8 per cent of domestic sales and 1.1 per cent of exports due to delays in transportation services. Almost a quarter of the enterprises surveyed in Mozambique considered transportation to be a major obstacle to investment. Currently, it is more expensive to transport cargo within Mozambique than to ship it to a different continent.

Gaps in terms of human capital are equally large. Only 60 per cent of Africa’s 15–24-year-olds have completed primary school, and only 19 per cent have gone beyond lower-secondary school (Filmer and Fox 2014). Between 1990 and 2005, as East Asia increased secondary enrolment rates by twenty-one percentage points and tertiary enrolment rates by 13 percentage points, Africa only managed to raise secondary enrolments by 7 percentage points and tertiary enrolments by one percentage point (World Bank 2007).

The political and economic turmoil of the 1980s and 1990s also took a toll on the region’s institutions. In 2000 SSA trailed all other developing regions in terms of government effectiveness, regulatory quality, rule of law, and control of corruption, in some cases by wide margins (Kauffman et al. 2010). The country case studies point to the many ways in which regulations and regulatory discretion affect firms. In Uganda senior managers of manufacturing firms spend more than thirteen days a year on average dealing with government officials, and 40 per cent of the manufacturing firms surveyed complained that (p.17) regulations were not interpreted consistently. In Mozambique business regulations—and the opportunities for corruption engendered by the regulatory regime—increase firms’ costs and reduce competitiveness.

1.5.2 Manufactured Exports

There is an empirical regularity linking exports to higher firm-level productivity in manufacturing, but there is a long-standing debate concerning whether the most productive firms in an economy select exporting or the act of exporting raises productivity. The answer is complex and linked to a country’s stage of development. While the most productive firms in an economy or a sector are most likely to become exporters, productivity in exporting firms rises faster than that in non-exporters, especially in lower income countries (Harrison and Rodriguez-Clare 2010). There is increasing evidence that learning by exporting takes place in low income country manufacturing firms, mainly as a result of the transfer of knowledge and working practices.11

SSA entered the twenty-first century with few manufactured exports and a small and declining share of the growing global market in manufactures. Manufactured exports per capita for the region as a whole (excluding South Africa) in 2000 were US$28 compared with US$291 for all developing economies. In 2000 the manufactured exports per capita of the eight SSAn countries in our study averaged just US$16.50. The share of manufactured exports in total exports for the region (excluding South Africa) was 25 per cent compared to 77 per cent for all developing countries and 91 per cent for East Asia (UNIDO 2013). The eight SSA case-study countries have had somewhat more success in exporting manufactures than the region as a whole. In 2010 their manufactured exports were on average 37 per cent of total exports. Manufactured exports per capita for the group were US$87 in the same year, nearly twice the SSA regional average (excluding South Africa), but well below the average for all developing economies of US$305.

Cambodia, Tunisia, and Vietnam have all achieved considerable success in manufactured exports. As recently as 2000, manufactured exports per capita in Cambodia were only US$107; by 2010 they had reached US$335. Tunisia’s manufactured exports to EU countries have expanded more than 10 per cent annually since the 1990s. Manufactured exports per capita which were US$522 in 2000 had more than doubled to US$1,381 in 2010. Vietnam’s exports increased at an average annual rate of 26.3 per cent during the period 2005–10. Per capita manufactured exports reached US$764 in 2010. In 2010 the share of manufactured exports in total exports for Cambodia and Vietnam (p.18) was about 70 per cent. In Tunisia manufactured exports made up 85 per cent of total exports.

1.5.3 Industrial Agglomerations

Manufacturing and service industries tend to concentrate in clusters and cities (Fujita et al. 1999). Econometric evidence (Newman et al. 2016) and case studies (Sonobe and Otsuka 2006) document the significant productivity gains to firms from industrial agglomeration in low income countries. This literature further suggests that localization economies derived from the clustering of similar firms are more important as a source of productivity gains in low income countries than urbanization. Productivity gains due to clustering come from a number of sources: a substantial labour market, information and knowledge spillovers, the ability to share common overheads and services, and the opportunity to observe customers and competitors closely (UNIDO 2009).

The country studies give some insight into how the agglomeration of industry is evolving across SSA. Industrial production in Senegal and Mozambique is mainly concentrated in the capital cities of Dakar and Maputo. Ghana’s industrial firms are primarily located in its two major urban centres, Accra, the capital, and Kumasi, the capital of the interior Ashanti region. In Uganda, Kampala has emerged as the major industrial centre, accounting for 41 per cent of manufacturing firms. These urban agglomerations are largely composed of firms engaged in unrelated activities, limiting the scope for localization economies.

There is a significant amount of clustering of similar firms in Cambodia and Vietnam. Both countries have been able to attract large numbers of foreign investors producing a limited range of products into their special economic zones (SEZs). There are two main industrial agglomerations in Vietnam, one located in the north near Hanoi and another in the south near Ho Chi Minh City. Over the past decade new clusters, anchored by large firms, have begun to appear in other areas. In Tunisia the degree of industrial concentration grew dramatically between 1995 and 2010, largely as a result of growth in the offshore sector. Relative to African agglomerations clusters in all three countries show a higher concentration of firms operating along the same value chain or producing similar products.

1.5.4 Foreign Direct Investment

Foreign owned firms in low income countries generally have higher levels of productivity than domestically owned firms. A large empirical literature finds positive productivity spillovers from foreign to domestic firms in lower (p.19) income countries, arising from access to the knowledge and technologies foreign firms have developed in more advanced markets. Not surprisingly, these knowledge spillovers are the result of firm-to-firm interactions and are more likely to occur between firms that are located along value chains than among firms competing in the same market.12

Table 1.5 FDI as a share of gross fixed capital formation (%), 1990–2013

1990–5

1995–2000

2000–5

2005–10

2010–13

Ethiopia

0.7

9.1

16.0

6.3

4.8

Ghana

6.0

7.9

5.7

21.8

23.3

Kenya

1.0

1.2

2.4

3.8

4.1

Mozambique

5.6

19.7

25.0

29.3

156.8

Nigeria

36.9

59.4

38.6

50.2

24.8

Senegal

3.2

8.4

4.5

8.7

8.8

Tanzania

1.1

13.7

17.8

17.1

18.7

Uganda

4.1

11.3

14.7

23.5

19.2

Tunisia

9.7

12.5

20.7

22.8

12.7

Cambodia

20.5

48.2

17.3

37.7

48.3

Vietnam

33.5

28.7

12.6

19.9

21.4

Africa regional average

6.1

11.2

13.0

17.1

12.4

Source: UNCTAD (2013).

Outside of the resource sector, SSA has had some difficulty in attracting foreign investors. Table 1.5 shows FDI as a share of gross fixed capital formation. With the exception of Nigeria, where the oil sector was a major FDI destination, the average share of FDI in total investment was 3.1 per cent during 1990–5 and 10.2 per cent during 1995–2000. Beginning around 2000 the share of FDI in total investment increased to about 15 per cent. Ghana, Mozambique, Tanzania, and Uganda all experienced substantially increased shares of FDI in total investment, but these inflows were driven largely by investments in the extractive sector.

FDI has played a much larger role in the comparator countries. Asian-based FDI has driven Cambodia’s manufactured exports. FDI increased from US$124 million in 1993 to over US$1,500 million in 2012. During the period 1990–2013 FDI represented 34.4 per cent of total investment. Between 1990 and 2013, FDI as a share of total investment in Vietnam averaged 23.2 per cent. Manufacturing has been the largest and fastest growing FDI sector, taking up over 60 per cent of all FDI. Tunisia’s offshore sector was an early magnet for European investors, particularly in textiles and garments. Since 1999 the annual flow of FDI has been about US$400 million, and between 2000 and 2010 FDI made up more than 20 per cent of gross fixed capital formation.

(p.20) 1.6 Conclusions

The idea that Africa should industrialize is certainly not new, but it remains relevant. While it is possible for economies to grow based on abundant land or natural resources—and Africa is well endowed with both—the role of industry in structural transformation makes a persuasive case for continued attention to industrialization. Africa’s current pattern of structural transformation—one of workers moving from agriculture to services—runs the risk that structural changes in employment may once again become growth reducing.

Despite their diversity in many other respects, the eight SSAn countries studied in this volume have a remarkable similarity in their approach to industrial policy and in their industrial performance. Each country, like SSA as a whole, has moved from state-led import substitution, to structural adjustment, to reform of the investment climate. Over the long run, industrial growth has followed a start–stop pattern and has failed to exceed the long-run growth of GDP. Cambodia, Tunisia, and Vietnam took a different path. Each undertook investments and institutional and policy reforms to implement an export-oriented industrialization strategy similar to that used by earlier East Asian industrializers. Tunisia is an African industrial success story. Cambodia and Vietnam have had explosive industrial growth since the mid-1990s.

Comparative results from the case studies suggest that policy choices are largely responsible for the differences in industrialization outcomes. SSA and Cambodia, Tunisia, and Vietnam differ strikingly in four areas—infrastructure, human capital, and institutions; exports; agglomerations; and FDI. These are major drivers of industrial growth through their impact on firm-level productivity. The low growth and fiscal austerity of the structural adjustment period left SSA with major gaps in infrastructure, skills, and institutions. This was probably more a matter of bad luck than poor policy.

In the other three areas, however, policy choices played a key role. Cambodia, Tunisia, and Vietnam all took explicit public actions to encourage exports, develop industrial clusters, and attract FDI. The SSAn economies had an opportunity to make such a policy turn after structural adjustment but did not do so. The result has been highly divergent paths towards industrial development.

References

Bibliography references:

African Center for Economic Transformation (ACET) (2014). 2014 Africa Transformation Report: Growth with Depth. Accra: ACET.

Asia Productivity Organization (APO) (2013). Asia Productivity Data Book. Tokyo: APO.

(p.21) Ben Jelili, R. and Goaied, M. (2009). ‘Entry, Exit, and Productivity in Tunisian Manufacturing Industries, in Khalid Sekkat (ed.) Market Dynamics and Productivity in Developing Countries. New York: Springer, 73–108.

Bleaney, M. and Söderbom, M. (forthcoming). ‘Learning by Exporting’. Journal of African Economies.

Chenery, H. (1986). ‘Growth and Transformation’, in H. Chenery, S. Robinson, and M. Syrquin (eds) Industrialization and Growth: A Comparative Study. Oxford: Oxford University Press, 13–36.

de Vries, G., Timmer, M., and de Vries, K. (2013). ‘Structural transformation in Africa: Static gains, dynamic losses’. GGDC Research Memorandum 136. Groningen, the Netherlands: University of Groningen.

Dinh, H. T., Palmade, V., Chandra, V., and Cossar, F. (2012). Light Manufacturing in Africa: Targeted Policies to Enhance Private Investment and Create Jobs. Washington, DC: World Bank.

Duarte, M. and Restuccia, D. (2010). ‘The Role of the Structural Transformation in Aggregate Productivity’. Quarterly Journal of Economics 121(1): 129–73.

Easterly, W. (2009). ‘Can the West Save Africa?’ Journal of Economic Literature 47(2): 373–447.

Escribano, A., Guasch, J. L., and Pena, J. (2010). ‘Assessing the Impact of Infrastructure Quality on Firm Productivity in Africa: Cross-country Comparisons Based on Investment Climate Surveys from 1999 to 2005’. Policy Research Working Paper 5191. Washington, DC: World Bank.

Filmer, D. and Fox, L. (2014). Youth Employment is Sub-Saharan Africa. Washington, DC: World Bank.

Foster, V. and Briceño-Garmendia, C. (eds) (2010). Africa’s Infrastructure: A Time for Transformation. Washington, DC: World Bank.

Fujita, M., Krugman, P., and Venables, A. J. (1999). The Spatial Economy. Cambridge, MA: MIT Press.

Harrison, A. and Rodriguez-Clare, A. (2010). ‘Foreign Investment, and Industrial Policy for Developing Countries’, in D. Rodrik and M. Rosenzweig (eds) Handbook of Development Economics, vol. 5. The Netherlands: North Holland, 4040–214.

International Monetary Fund (IMF) (2014). ‘Sustaining Long-run Growth and Macroeconomic Stability in Low-income Countries—The Role of Structural Transformation and Diversification’. IMF Policy Paper. Washington, DC: IMF.

Kauffman, D., Kray, A., and Mastruzzi, M. (2010). ‘The Worldwide Governance Indicators: Methodology and Analytical Issues’ Policy Research Working Paper No. 5430 Washington, DC: World Bank.

Killick, T. (1978). Development Economics in Action. New York: St. Martin’s Press.

Lewis, W. A. (1954). ‘Economic Development with Unlimited Supplies of Labour’. Manchester School 22(2): 139–91.

McMillan, M. and Hartgen, K. (2014). ‘What is Driving the Africa Growth Miracle?’. Tunis: African Development Bank.

McMillan, M. and Rodrik, D. (2011). ‘Globalization, Structural Change and Productivity Growth’. NBER Working Paper 17/143. Cambridge, MA: National Bureau of Economic Research.

(p.22) McMillan, M., Rodrik, D., and Verduzco-Gallo, I. (2014). ‘Globalization, Structural Change, and Productivity Growth, with an Update on Africa’. World Development 63: 11–32.

Megginson, W. L. and Netter, J. M. (2001). ‘From State to Market: A Survey of Empirical Studies on Privatization’. Journal of Economic Literature 39(2): 321–89.

Ndulu, B. J. (2007). ‘The Evolution of Global Economic Paradigms and their Influence on African Economic Growth’, in B. J. Ndulu, R. Bates, S. A. O’Connell, P. Collier, and C. Saludo (eds) The Political Economy of Growth in Africa, Volume 1. New York: Cambridge University Press, 315–47.

Nellis, J. (1986). ‘Public Enterprises in sub-Saharan Africa’. Washington, DC: World Bank.

Nellis, J. (2003). ‘Privatization in Africa: What has happened? What is to be done?’ Working Paper Number 25. Washington, DC: Center for Global Development.

Newman, C., Page, J., Rand, J., Shimeles, A., Söderbom, M., and Tarp, F. (2016). Made in Africa: Learning to Compete in Industry. Washington, DC: Brookings Institution Press.

Organisation for Economic Co-operation and Development (OECD) (2014). ‘DAC CRS International Development Statistics (IDS) online databases’ available at www.oecd.org/dac/stats/idsonline.htm, accessed 28 October 2014.

Page, J. (2012). ‘Can Africa Industrialize?’ Journal of African Economies 21 (Special Issue 2): ii86–ii124.

Rodrik, D. (2013). ‘Unconditional Convergence in Manufacturing’, The Quarterly Journal of Economics 128(1): 165–204.

Sonobe, T. and Otsuka, K. (2006). Cluster-based Industrial Development: An East Asian Model. Basingstoke: Palgrave Macmillan.

Stern, N. (2001). A Strategy for Development. Beijing: People’s University.

Stern, N. (2002). The Investment Climate, Governance, and Inclusion in Bangladesh. World Bank Office of the Senior Vice President, Development Economics. Washington, DC: World Bank.

Syverson, C. (2011). ‘What Determines Productivity?’ Journal of Economic Literature 49(2): 326–65.

UNIDO Statistics (2014) Manufacturing Value Added Database, available at www.unido.org/resources/statistics/statistical-databases.html, accessed 28 January 2016.

United Nations Conference on Trade and Development (UNCTAD) (2013). World Investment Report 2013: Global Value Chains: Investment and Trade for Development. Geneva: UNCTAD.

United Nations Industrial Development Organization (UNIDO) (2009). Industrial Development Report, 2009. Vienna: UNIDO.

United Nations Industrial Development Organization (UNIDO) (2013). Industrial Development Report, 2013. Vienna: UNIDO.

Williamson, J. (1990). ‘What Washington Means by Policy Reform’, in J. Williamson (ed.) Latin American Adjustment: How Much Has Happened? Washington, DC: Institute for International Economics, 36–58.

World Bank (1992). ‘World Bank Structural and Sectoral Adjustment Operations: The Second OED Overview’. Washington, DC: World Bank.

(p.23) World Bank (2000). Can Africa Claim the Twenty-first Century? Washington, DC: World Bank.

World Bank (2007). ‘Expanding the Possible in Sub-Saharan Africa: How Tertiary Institutions Can Increase Growth and Competitiveness’. Washington, DC: World Bank.

World Bank (2009). Transforming Africa’s Infrastructure. Washington, DC: World Bank.

World Bank (2014). World Development Indicators, available at <http://data.worldbank.org/products/wdi>, accessed 28 January 2016. (p.24)

Notes:

(1) Working paper versions of the research are published on the UNU-WIDER website at <http://www.wider.unu.edu/research/current-programme/en_GB/L2C-2010> and on the Brookings Institution website at <http://www.brookings.edu/about/projects/africa-growth/learning-to-compete>. An overview of the research programme as a whole can be found in Made in Africa: Learning to Compete in Industry (Brookings Institution Press, 2015).

(2) McMillan and Rodrik (2011) and de Vries et al. (2013) present differing formal decompositions.

(3) See Killick (1978) for a survey of the influence of development economics on the strategies pursued by Africa’s post-independence leaders.

(4) Mauritius was the only SSA country to introduce a similar two-track approach to industrial development, beginning with the development of its free trade zones at about the same time.

(5) While the IMF and the World Bank were the prime movers the regional development banks, including the African Development Bank, also became involved in structural adjustment.

(6) See for example Easterly (2009).

(7) Interestingly, at about this time, Mauritius also was making efforts to improve the productivity of individual firms (ACET 2014).

(8) The estimates for countries other than Mozambique and Uganda are drawn from the Groningen Africa Database (de Vries et al. 2013). Estimates for Mozambique and Uganda were kindly provided by Margaret McMillan from the extended McMillan–Rodrik database described in McMillan et al. (2014).

(9) For a review of the literature on firm-level productivity see Syverson (2011). UNIDO (2009) surveys the evidence on the impact of productivity drivers on choice of industrial location.

(10) For reviews of the literature see the various World Bank Doing Business reports.

(11) See Bleaney and Söderbom (forthcoming).

(12) See Harrison and Rodriguez-Clare (2010).