Based on understandings of European capitalist development experiences, economists argue that the classical pattern of structural change began with the reduction of the share of employment and output in the agriculture sector and its transfer to the manufacturing sector and later to the services sector (Chenery Reference Chenery1960; Clark Reference Clark1940; Kuznets Reference Kuznets1955). In the 1950s and 1960s, a consensus prevailed about the importance of structural change as a key factor in promoting catch-up development. However, the neoliberal counter-revolution of the 1970s (Toye Reference Toye1987) marginalised discussions of structural change (or transformation) – and, particularly, manufacturing-based growth. Instead, International Financial Institutions (IFIs), including the World Bank and the International Monetary Fund (IMF), lobbied for the adoption of market-enhancing policies in developing countries while discouraging industrial policies. IFIs encouraged the adoption of market-led reforms, including privatisation, trade liberalisation, financial sector liberalisation and the reduction of exchange and capital controls. IFI proposals were based on the assumptions that reducing market distortions would contribute to a more efficient allocation of resources, which would lead to growth.

Later, the conception of what constitutes ‘development’ broadened beyond a specific focus on growth alone. With the increasing influence of ‘basic needs’ and ‘capability’ approaches, conceptions of ‘development’ were extended to include social indicators (such as health and education improvements) (Sen Reference Sen1985, Reference Sen1999; Streeten Reference Streeten1979). Since the 2008/2009 financial crisis and the ascendance of the Sustainable Development Goals (SDGs), considerations of what governments should focus on when targeting ‘development’ interventions have become even broader, with addressing climate concerns becoming increasingly urgent. The SDGs include 17 goals and 169 targets. Only two goals (Goal 8 and Goal 9) focus on ‘decent work and economic growth’ and ‘industry, innovation and infrastructure’. While this may be interpreted as a sign that the importance of structural transformation is being recognised, it is just one of many priorities in global development policy discussions.

There are reasons to be concerned about why structural change is not at the top of the agenda of development policymaking. Very few Asian, African and Latin American countries have ‘caught up’ with industrialised European countries since the 1950s. Wade (Reference Wade2018) argues that fewer than ten countries have achieved this feat, and most of the countries that have caught up are in East Asia. ‘Catching up’ or achieving sustained economic growth requires countries to achieve structural transformation of their economies. Structural transformation refers to a process where a national economy moves ‘from a set of assets based on primary products, exploited by unskilled labour, to a set of assets based on knowledge, exploited by skilled labour’ (Amsden Reference Amsden2001, p. 2). In common parlance, structural transformation is sometimes wrongly assumed to be simply referring to export diversification. Export diversification is an important facet of structural transformation but it has not been historically sufficient to sustain economic growth and catch-up development. Export diversification, in the short to medium term, can provide more policy autonomy for late-developing economies.Footnote ¹

When most former colonies became independent in the 1950s and 1960s, they heavily depended on primary commodity exports (like coffee, tea, cocoa, oil or minerals). Primary commodity dependency leaves economies vulnerable to global commodity price fluctuations. Late-developing economies do not produce very much of what is consumed domestically. Thus, they need to import products from other countries and usually have high trade deficits. If export prices for the commodities they produce fall, countries have very little foreign exchange to buy those imported products, which include basic consumption goods for their population. Late development is often characterised by the balance-of-payment disequilibria associated with limited export diversification (Stockhammer Reference Stockhammer2023). For structuralists (Fischer Reference Fischer2009), this provided a justification for why aid was necessary for undiversified economies. This structuralist lesson remains pertinent today. In the 2020s, during and after the Covid-19, several countries in the Global South have faced unmanageable trade deficits, leading them to approach the IMF for loans.

Export diversification can occur with very little involvement of domestic firms. Thus, countries may remain vulnerable if foreign investors leave (and export revenues diminish). The increased policy space accompanying export diversification may be short-lived. For structural transformation to be sustained, domestic firms need to acquire knowledge-based assets. Knowledge-based assets are sets of skills that enable owners ‘to produce and distribute a product at or above prevailing market prices (or below market costs)’ (Amsden Reference Amsden2001, p. 3). Firms that acquire knowledge-based assets must also attain managerial and technological skills and then, through processes of learning and trial and error, eventually become competitive in regional and global markets.

Domestic firms usually only acquire knowledge-based assets through learning and investing in technological capabilities in relationships with foreign firms. Thus, governments need to first provide subsidies to attract three-pronged investment: in plants of efficient scale, in distribution networks and in managerial and technological capabilities (Amsden & Hikino Reference Amsden and Hikino2000; Chandler Reference Chandler1990). How East Asian firms – initially Japan, Korea and Taiwan, often described as ‘developmental states’ – acquired knowledge-based assets remains heavily debated. Whitfield (Reference Whitfield2023) summarised this debate, highlighting that the developmental state literature (Amsden Reference Amsden1989; Evans Reference Evans1995; Johnson Reference Johnson1982; Wade Reference Wade1990) focused on the role of the state in directing technology transfer to local firms but downplayed the role of firms themselves. The literature on Global Value Chains/Global Production Networks (GVCs/GPNs) (Morrison et al. Reference Morrison, Pietrobelli and Rabelloti2008; Yeung Reference Yeung2016), as well as other prominent work on technology transfer (Mathews & Cho Reference Mathews and Cho2000), meanwhile, has focused on the role of lead firms in advanced countries in showing how technology transfer occurs but again underplayed the role of latecomer firms. Yet, for many African firms, the process of developing knowledge-based assets remains elusive. Even where some African firms have grown substantially, there are few examples of these firms becoming globally competitive.

For latecomer firms to develop knowledge-based assets, an initial starting point is for governments and national firms to develop aligned interests or effective state–business relationships (Whitfield et al. Reference Whitfield, Therkildsen, Buur and Kjaer2015). The developmental state literature (Amsden Reference Amsden1989; Johnson Reference Johnson1982; Wade Reference Wade1990), which analysed structural transformation and rapid ‘catch-up’ in Japan, South Korea and Taiwan, showed how different configurations of state–business relations enabled the acquisition of knowledge-based assets among latecomer firms.Footnote ² Within the developmental state literature, Evans (Reference Evans1995) provided one influential explanation of why state–business relations in East Asia were productive. Borrowing from Max Weber, Evans argued that the autonomy of East Asian bureaucracies was crucial. East Asian bureaucracies, through departments like Japan’s Ministry of Trade and Industry and Korea’s Economic Planning Board, were highlighted as integral in coordinating overall development strategy. Evans (Reference Evans1995), however, argued that autonomy was not enough. Instead, it was crucial that bureaucracies had ‘embedded autonomy’. He argued that state bureaucracies needed to be autonomous enough to avoid being beholden to individual capitalists while, at the same time, retaining industrial knowledge and maintaining connections with capitalists so that collective decisions could be made about how economic challenges could be confronted. Evans and James Rauch went on to argue that East Asian bureaucracies showed characteristics like Weberian-style bureaucracies (Evans & Rauch Reference Evans and Rauch1999). The ‘Weberianness scale’ measure was based on whether ‘state agencies are characterized by meritocratic recruitment and offer predictable, rewarding long-term careers’, arguing that this was a signal of the autonomy of bureaucrats (Evans & Rauch Reference Evans and Rauch1999, p. 749). In contrast to their study, most empirical evidence in East Asia found the region to be as corrupt and captured by big business as in other regions with less impressive progress in achieving structural transformation (like South Asia) (Kang Reference Kang2002; Khan & Sundaram Reference Khan and Sundaram2000).

Amsden (Reference Amsden2001) argued that reciprocity or what she called ‘reciprocal control mechanisms’ were essential characteristics of effective state–business relationships. When governments provided incentives to firms, the continued provision of those incentives depended on results being achieved. In other words, it was not enough that governments used a stick; they also had to provide carrots. In South Korea, licenses were provided to firms if they met export criteria related to value, geographical diversity and product complexity. In Taiwan, foreign firms were granted incentives and facilities in science parks, provided a percentage of their profits was reinvested in research and development. Local content requirements were placed on foreign firms to transfer technology in Turkey, Brazil, India and China (Amsden & Hikino Reference Amsden and Hikino2000). Similar characteristics of state–business relations were also identified as crucial in achieving industrialisation by prominent economists like Rodrik (Reference Rodrik2008a). Rodrik (Reference Rodrik2004) argued that industrial policy involved a ‘discovery’ process, which relied on governments learning about underlying costs and opportunities and engaging in strategic coordination with firms.

East Asian industrial policies were not always effective. Many errors were made. From Korea’s poorly planned investments in heavy chemicals to Japan’s interventions in videocassette recorder standardisation, there are umpteen examples of industrial policy experiments that failed (Biggs & Levy Reference Biggs, Levy, Perkins and Roemer1991; Moon & Prasad Reference Moon, Prasad, Chan, Clark and Lam1998). Such experiences were not unique to East Asian countries. Enacting industrial policy is a difficult process. It inevitably involves many cases of failure. Learning from that failure and encouraging the acquisition of technological capabilities among local firms during that process is a crucial reason why some East Asian countries achieved structural transformation.

Most neoclassical perspectives ignore the centrality of technological learning and organisational dynamics and, instead, argue that technology flows freely between countries (Pack & Saggi Reference Pack and Saggi2006). In contrast, evolutionary perspectives highlighted that technology is difficult to transfer among firms and countries because it includes bundles of both tacit and explicit information (Nelson & Winter Reference Nelson and Winter1982). Explicit knowledge refers to knowledge that is codified, standardised and easily diffused. Intuitive knowledge or ‘tacit knowledge’ (Polanyi Reference Polanyi1966) refers to knowledge that is not easily put into words (Narula Reference Narula2014). Learning and experimentation in late-developing economies requires both tacit and explicit knowledge, which is necessary to scale the knowledge frontier (Lall Reference Lall1992). Too often, attempts at scaling the knowledge frontier result in a ‘learning trap’ in late-developing countries where local firms do not invest in building capabilities because capability gaps are too large, which leaves most firms stuck in low-productivity activities (Khan Reference Khan2019; Whitfield & Staritz Reference Whitfield and Staritz2021).

Perhaps, the most difficult aspect of industrial policy is disciplining ‘learning rents’, investing in organisational and institutional adaptation to encourage firms to diversify into higher-value activities and markets. Fundamentally, discipling learning rents is a political process that requires examining power relations shaping state–business relationships. Rents play an important role in facilitating the process of learning and innovation, including the copying of existing technologies and adapting industrial policies to local conditions, available capital stock and institutions (Amsden Reference Amsden1989; Khan Reference Khan, Khan and Sundaram2000). In African countries, the most difficult challenge has been to develop effective state–business relations such that firms are incentivised to invest in acquiring technological capabilities (Whitfield et al. Reference Whitfield, Therkildsen, Buur and Kjaer2015). While what is needed to achieve catch-up development is the same as it has always been (structural transformation), the nature of that challenge (both domestically and internationally) has changed significantly after decades of market-led reforms have altered both national economies and the global political economy. The following section describes how the challenge of achieving structural transformation has changed.

The Altered Challenge of Late Development

Across varied schools of economic thought and broader social science, large segments of scholarship retain a view that countries pass through stages of economic growth (Dyson Reference Dyson2010; Lal Reference Lal1985; Marx Reference Marx1976; Rostow Reference Rostow1960). A large share of development studies literature continues to assume a ‘universal’ trajectory of development. However, some prominent development economists have disagreed, arguing that ‘late development’ always followed a unique path. Gerschenkron’s (Reference Gerschenkron1962) key insights on ‘late development’ showed how industrialisation in one country altered the possibility of future structural transformation elsewhere.Footnote ³ Gerschenkron showed how Britain’s industrialisation created competitive challenges for late industrialisers elsewhere (Russia, Japan, Germany), providing an opportunity for importing technologies and creating new incentives for rulers. While Gerschenkron emphasised the opportunities that came with ‘relative backwardness’, Amsden (Reference Amsden1989, Reference Amsden, Putterman and Rueschemeyer1992) argued that the technological gap between industrialised countries and the rest had become much larger in the late twentieth century. International laws on intellectual property constrained the capacity of late developers to copy technology.

While Gerschenkron and Amsden differed on whether ‘lateness’ was an opportunity or a constraint, they both agreed that economic development was inherently context-specific and did not take a universal form. Other scholars have also highlighted how different factor endowments or different colonial legacies have shaped how governments employ their agencies to place their national economies on very different economic trajectories. Amin (Reference Amin1972), for example, categorised Africa into three groups, based on how colonial administrations had organised production and labour systems: (1) ‘Africa of the colonial economy’ or cash crop–dependent economies; (2) Africa of the concessionary economies; (3) Africa of the labour reserves.Footnote ⁴

The Varieties of Capitalism (VoC) approach (Hall & Soskice Reference Hall and Soskice2001) was also established to highlight the varied pathways to achieve economic development globally. However, the VoC approach has been criticised for remaining within the analytical frames of New Institutional Economics (NIE), particularly retaining a focus on transaction costs. The VoC approach has also been criticised for ignoring the role of finance and only focusing on two varieties: liberal market economies and coordinated market economies (Hay Reference Hay2020; Hodgson Reference Hodgson2016). Recognising some of these limitations, other work has focused on broadening analysis of different kinds of economic trajectories (through focusing on hierarchical market economies), as well as adopting more Marxian-inspired analyses, including through ‘comparative capitalisms’ approaches (Hall & Soskice Reference Hall and Soskice2001; Scharpf & Schmidt Reference Scharpf and Schmidt2000; Streeck & Thelen Reference Streeck and Thelen2005). However, these approaches have also been subject to criticisms because they retain a narrow focus, with most analyses only discussing institutional arrangements and public policies, when studying contemporary capitalist development (Brenner et al. Reference Brenner, Peck and Theodore2010; Bruff Reference Bruff2021). These literatures have received very little traction in the study of African political economy.Footnote ⁵

While structural transformation remains central to achieving ‘catch-up’ development, each country will always face different challenges based on its colonial history, factor endowments, geographical location, ideologies of ruling parties and leaders and numerous other factors. However, the challenge of achieving structural transformation has also changed in significant ways. In the 1950s and 1960s, there was consensus within development economics about the central role of state intervention and the importance of industrialisation in fostering growth (Hirschman Reference Hirschman1958; Lewis Reference Lewis1954; Rosenstein-Rodan Reference Rosenstein-Rodan1943). The Prebisch–Singer thesis claimed that the price of raw materials, which accounted for nearly 90 per cent of exports from developing countries, fell over time relative to the price of manufactured goods (Prebisch Reference Prebisch1950; Singer Reference Singer1950). They reasoned that since, as incomes rise, people spend a smaller share of their income on basic goods, manufacturing demand in a growing global economy will be strong. They identified the relatively high-income elasticity of demand for manufactures as compared to demand for primary commodities. The Prebisch–Singer approach provided the intellectual basis for import substitution industrialization (ISI) strategies, which were adopted across many former colonies.

In the 1950s–1970s, ISI policies contributed to a ‘Golden Age’ in the Global South, with higher average growth in those decades than in the successive decades of the twentieth century (Amsden Reference Amsden2007; Mkandawire Reference Mkandawire2005). However, the implementation of ISI came with problems. There was limited evidence of learning among latecomer firms, with few firms developing sufficient technological capabilities to compete in global markets. Governments ‘never fully drew out the central implication, namely, that they should speed up their diversification into nontraditional exports’ (Mkandawire & Soludo Reference Mkandawire and Soludo1999, p. 39). With countries struggling to access foreign exchange because of global commodity price fluctuations in the 1980s, many governments sought loans from IFIs to address trade deficits and access foreign exchange.

ISI’s failures received significant criticism, especially from neoliberal critics and IFIs. Critics vilified the state’s role and political connections between governments and business as wasteful and corrupt (Balassa Reference Balassa1982; Krueger Reference Krueger1974; Lal Reference Lal1985; World Bank 1981). Crucially, since the 1970s and 1980s, influential neoclassical economists have ignored the importance of structural transformation. Instead, they have tried to isolate specific factors to problematically explain the economic underdevelopment of the entire continent or what Easterly and Levine (Reference Easterly and Levine1997) infamously called the ‘African growth tragedy’. Economists have isolated factors ranging from malaria prevalence and other diseases to health concerns, cultural heterogeneity, geography, extractive economic institutions and the slave trade (Acemoglu et al. 2001; Alesina et al. 2016; Gallup & Sachs 2000; Michalopoulos & Papaioannou 2020; Nunn 2008). The distrust of political connections between governments and capitalists remains a feature of mainstream economics and political science literature in the 2020s (Canen & Wantchekon 2022).

Influential strands of dependency theory also contributed to ‘hopeless’ characterisations of African development. Some segments of dependency theory argue that alliances between transnational capital and domestic oligarchies leave countries in positions of permanent capitalist underdevelopment (Frank 1966). The dependency school literature on African studies argues that where growth has taken place on the continent, it has largely been a result of a commodity boom and there has been very limited structural transformation (Sylla 2014; Taylor 2016). Some Marxian approaches to the study of finance have argued that increased financial liberalisation has resulted in increased financialisation or financial subordination, which constrains policy space for late developers. Dependency associated with financial subordination has been most evident in the case of the CFA franc in West Africa, which has left monetary policy outside the realms of government policy in eight West African countries (Pigeaud & Sylla 2021).

Colonialism resulted in structural dependencies for most former colonies, including many African economies. However, there has been significant variation in the economic trajectories taken by different economies across the continent. There have also been varied degrees of progress in manufacturing growth (Cramer et al. 2020), although structural transformation has remained elusive. Variation suggests that African governments have exhibited significant agency and have not been passive participants in the global economy. Static versions of dependency theory, rather than more nuanced versions of dependency (Cardoso & Faletto 1967; O’Donnell 1979), may also fall short in neglecting African agency and the (albeit constrained) capacity of African governments to reshape their economies.

Since the imposition of structural adjustment policies in the Global South, the legacy of more than three decades of market-led reforms has reshaped the challenge of contemporary late development. As countries embraced market-led reforms and reduced trade barriers, supporting domestic firms has become more difficult. Bilateral trade agreements and World Trade Organisation (WTO) requirements have further reduced space for industrial policy (Wade Reference Wade2003). While the WTO imposes explicit constraints, regional and bilateral trade agreements limit policy space even more (Shadlen Reference Shadlen2005; Thrasher & Gallagher Reference Thrasher and Gallagher2008). Recently, the United States’ decision to disable the WTO’s enforcement mechanism by blocking judicial appointments to the appellate body sparked euphoric arguments that this would lead to increased policy space for the Global South (Hopewell 2024). Yet, in African countries, applied tariffs remain far below the ceilings of what the WTO allows or what is referred to as ‘bound tariffs’ (Khan Reference Khan2007). This points to a deeper challenge within African government bureaucracies where internal ideological pressures may pressure bureaucrats and politicians to act in line with international trading norms. The United Nations Economic Commission for Africa (UNECA 2016) identified the IMF and World Bank ‘tariff rationalisation’ processes as significant pressures disciplining African bureaucrats to act in accordance with free market principles. Others note the increasing socialisation of African bureaucracies to the ideals of a market-embracing development project (Harrison Reference Harrison2010; Stein Reference Stein2021). In most African countries, since the adoption of structural adjustment programmes, government officials who advocate for market-led policies and work in ‘austerity ministries’ like the Ministry of Finance have been empowered at the cost of bureaucrats working in ‘expenditure ministries’ like the Ministry of Industry (Mkandawire Reference Mkandawire2014).

Contemporary globalisation, characterised by increased international competition and the dominance of patents, technologies and production by North American, European and East Asian firms, is also often viewed as a constraint for structural transformation in late-developing countries. As the influential GVCs/GPNs scholarship (Gereffi & Korzeniewicz Reference Gereffi and Korzeniewicz1994) suggests, global production and trade have been increasingly fragmented geographically, with lead firms in many industrial sectors exerting control over the governance of production (Dicken Reference Dicken2007; Gereffi Reference Gereffi2001). The GVC and GPN theoretical approaches have analysed governance dynamics across several sectors, lead firm strategies and the implications for supplier firms (Gereffi et al. Reference Gereffi, Humphrey and Sturgeon2005; Henderson et al. Reference Henderson, Dicken, Hess, Coe and Yeung2002). GVCs/GPNs have become the predominant frameworks across a range of disciplines to analyse the challenge of economic diversification (or upgrading) under contemporary globalisation. Exporting through GVCs/GPNs is often highlighted as a solution to weak industrialisation trends (Taglioni & Winkler Reference Taglioni and Winkler2016; World Bank 2020a). Some literature even suggests that industrialisation is easier than ever because information and communication technologies (ICTs) enable easier access to global markets and carry out some stages in the production process (Baldwin Reference Baldwin2016). Segments of the GVC literature suggested that economic upgrading would automatically lead to social gains, with workers benefiting from integration into global production processes. However, scholarship has consistently shown this is not the case (Barrientos et al. Reference Barrientos, Gereffi and Rossi2011; Dussel Peters Reference Dussel Peters2008; Gibbon & Ponte Reference Gibbon and Ponte2005; Selwyn Reference Selwyn2019). Scholars have found that inclusion into GVCs has resulted in ‘immiserising growth’ (Kaplinsky Reference Kaplinsky2000; Kaplinsky & Morris Reference Kaplinsky and Morris2016). Inclusion into GVCs/GPNs results in ‘adverse incorporation’ (Hickey & du Toit Reference Hickey and du Toit2007) of labour and firms into global production and distribution structures.

The GVCs/GPNs literature has grown in varied directions to highlight the unequal effects associated with exporting through GVCs, highlighting challenges associated with upgrading and technological capability acquisition across several sectors (Bair & Werner Reference Bair and Werner2011; Whitfield & Staritz Reference Whitfield and Staritz2021). The literature acknowledges how difficult it is for supplier firms to upgrade economic activities and acquire technological capabilities from subordinate positions in asymmetric and fragmented GVCs/GPNs (Quentin & Campling Reference Quentin and Campling2018; Whitfield & Staritz Reference Whitfield and Staritz2021). Despite these warnings, IFIs and many donors (World Bank 2020a) have taken up a policy suggestion of Gary Gereffi’s early work on GVCs: “development requires linking up with the most significant lead firm in the industry” (Gereffi Reference Gereffi2001, p. 622). Donor advice is out of touch with the pressures facing latecomer firms in African countries. Many domestic firms prioritise producing for domestic and regional markets. Governments and firms are often working at contrasting ends, with donor advice often simply encouraging governments to attract foreign lead firms to invest rather than focusing on policies that encourage governments and latecomer firms to prioritise investments in technological capability acquisition.

Governments also do not have the same financial instruments at their disposal as they did in the 1960s. The space to use the financial sector to promote industrialisation has been significantly curtailed in recent decades. As part of structural adjustment policies, countries were forced to liberalise their financial sectors and reduce capital and exchange controls (Stein Reference Stein2010). Two-thirds of World Bank loans issued to African countries between 1980 and 1995 included conditions relating to financial sector liberalisation (Arriola Reference Arriola and Cheeseman2018). Yet in East Asian countries and in most European countries, nationally owned financial sectors were important instruments in directing investments for structural transformation (Woo Reference Woo1992). IFIs also argue for central banks to restrict their regulatory roles to maintain price stability while also calling for a commitment to central bank independence (CBI). IFI pressures have constrained the capacity of developing countries to employ ‘developmentalist roles’ in their banking sectors. There is limited space to directly use development banks to mobilise resources for structural transformation or to employ developmentalist central banking, which would contribute to diverting resources for diversification (Epstein Reference Epstein2013).

Another significant constraint to structural transformation is the increasing prominence of service sectors as a component of national growth, as well as a centrepiece of many countries’ development strategies. This is significant, since most heterodox scholarship (whether Marxian or Keynesian) has emphasised the role of the manufacturing sector in contributing to sustained latecomer growth. Kaldor’s (Reference Kaldor1967) three growth laws highlight that growth of manufacturing output is positively related to GDP growth (law 1) and productivity growth both within and outside the manufacturing sector (laws 2 and 3).Footnote ⁶ Hirschman’s (Reference Hirschman1958) work on forward and backward linkages showed that all sectors in the economy are linked to each other through backward and forward linkages via input demand and supply. The manufacturing sector is characterised by stronger backward and forward linkages than any other sector, making it the main engine of economic development. Indeed, the historical record shows manufacturing was the predominant source of dynamic growth in Western and East Asian development success stories (Chang Reference Chang2002; Reinert Reference Reinert2007).

There are several estimates of the rapidly growing importance of services in African countries. For example, a recent McKinsey (Reference McKinsey2023) report estimates that services is the major driver of Africa’s economic output, contributing 56 per cent to national GDP and 39 per cent of total employment, as of 2019. Yet estimates of the contribution of services to GDP and overall growth are imprecise (Cramer et al. Reference Cramer, Sender and Oqubay2020; Jerven Reference Jerven2013). One reason for this is that ‘the boundaries between sectoral activities have become increasingly blurred as manufacturing, services and agricultural activities overlap, integrate and become imbricated in one another’ (Cramer et al. Reference Cramer, Di John and Sender2022, p. 496). There is an increasing ‘servicification’ of manufacturing, with an increasing share of the final value of products derived from logistics or branding (Lodefalk Reference Lodefalk2017). Some argue that services are leading to an expansion of manufacturing activities (Dasgupta & Singh Reference Dasgupta and Singh2005). Rodrik (Reference Rodrik2016) argues that services could play the ‘escalator role’ that manufacturing traditionally played. However, he acknowledges that in most of the Global South, service sectors are skill-intensive and do not absorb the labour that exists. More crucially, where sectors like tourism, finance and IT have been prioritised, labour has often not been absorbed and there has been a limited focus on ensuring there are linkages between domestic agricultural and manufacturing sectors with services. For example, many African countries have focused on luxury tourism strategies, which require importing products from abroad to meet luxury certification standards. This negates the possibility of using domestically produced agriculture and industrial products and increases imports unless domestic firms are supported to invest in the technological capabilities to meet certification standards (Behuria & Goodfellow Reference Behuria and Goodfellow2019). Another crucial barrier to industrialisation is the rising prices of urban property in African capital cities, which disincentivises wealthy domestic elites from making productive investments that may promote structural transformation (Gillespie & Schindler Reference Gillespie and Schindler2022).

Traditional developmentalist pathways may seem more restricted. However, there are also some reasons to be hopeful. Grabel (Reference Grabel2017) argues that there are opportunities because of the ‘productive incoherence’ of the global economy. She argues that global financial and economic governance has left cracks for the most adventurous governments to diversify their relationships with external actors and thereby achieve more economic autonomy (Grabel Reference Grabel2017; Hirschman Reference Hirschman1971). Much has also been written about how the variety of new donors supporting development policy in the Global South has led to new opportunities and increased policy space for late developers (Mawdsley Reference Mawdsley2012). Growing, dynamic Asian markets may be future export pathways for late-developing economies. More geopolitical competition may also be a way for some economies to exact more policy space. Yet a fundamental challenge will be whether any policy space extracted through managing geopolitical competition or exporting to a wider range of markets can be used to promote structural transformation.

This book is aligned with a branch of African political economy, which argues that the politics of state–business relationships continues to be among the most significant constraints to structural transformation (Whitfield et al. Reference Whitfield, Therkildsen, Buur and Kjaer2015). It aligns with political economy approaches (Khan & Sundaram Reference Khan and Sundaram2000) to argue that disciplining the provision of rents to firms in line with learning and technological capability acquisition can have positive outcomes and can be a crucial ingredient for structural transformation. The next section describes how existing literature has analysed the politics of structural transformation.

The Politics of Structural Transformation

Analysis of how European countries developed (Moore Jr. Reference Moore1967) emphasised the essential role of domestic capitalists in structural transformation, as well as in democratic transitions. Yet, in many countries outside Europe, domestic capitalists did not become politically significant and were not significant contributors to structural transformation. Work on Latin American experiences, following dependency traditions, highlighted the influence of specific historical circumstances shaping the development of internal capitalist classes (Cardoso & Faletto Reference Cardoso and Faletto1967; O’Donnell Reference O’Donnell1979). A debate emerged, focusing on whether domestic capitalist classes could shape autonomous nationalist development strategies or whether they would be inevitably comprador (or subservient) to the interests of foreign capital. These debates took centre stage in the African political economy literature in the 1970s and 1980s. A divide emerged (most prominently with reference to Kenya), with scholars initially claiming that there were few examples of domestic capitalists and those which existed appeared to be comprador to foreign capital.Footnote ⁷ Leys (Reference Leys1975) argued that there was no sign of an emerging Kenyan bourgeoisie that could challenge the dominant position of foreign capital. However, Leys (Reference Leys1978) was among those scholars who later changed their position to a more optimistic one. Swainson (Reference Swainson1980), for example, argued that a national bourgeoisie had begun to emerge in Kenya. Evidence used by both sides remained ambiguous, and no consensus emerged as to the position of domestic capitalists. Gradually, debates regarding African capitalists faded in significance within African studies.

The study of African capitalist classes waned in the 1990s and 2000s, with some exceptions (Boone Reference Boone1992; Handley Reference Handley2008). More recently, there has been a revival in this literature (Kelsall Reference Kelsall2013; Ochonu Reference Ochonu2018; Whitfield et al. Reference Whitfield, Therkildsen, Buur and Kjaer2015). However, recent research is more concerned with analysing how African business shapes democratic outcomes or social outcomes (Arriola Reference Arriola2013; Handley Reference Handley2019; Pitcher Reference Pitcher2012) rather than analysing their role in the pursuit of structural transformation on the continent. Individual African entrepreneurs have become increasingly visible on the continent, with Nigerian entrepreneur Tony Elemulu among those who have advocated for ‘Africapitalism’, which aims to combine profit motives with the political and economic ideas of African unity (Amaeshi et al. Reference Amaeshi, Okupe and Idemudia2018).Footnote ⁸ However, the Africapitalist literature has been more concerned with glorifying the role of African entrepreneurship rather than analysing whether African businesspeople are contributing to structural transformation.

Scholarship has consistently highlighted that newly independent leadership in African countries perceived their relationships with domestic capitalists to be part of the political calculus of their rule (Boone Reference Boone1992; Iliffe Reference Iliffe1983). Domestic capitalists have been portrayed as a threat to governments given that their wealth could be used to fund opposition politicians (Arriola Reference Arriola2013). Such anxieties about the threat posed by domestic capitalists in financing rivals are experienced by all governments worldwide. However, most African countries have had limited success in maintaining elite cohesion to prioritise structural transformation and manage effective state–business relations. The challenge of managing effective state–business relations is primarily political. It involves giving benefits to some over others. Even those who are given benefits must be disciplined, monitored and incentivised to invest in technological capability acquisition. Crucially, in comparison to successful past developers, African governments face bigger hurdles in creating effective state–business relationships because of the altered global economy and the legacies of market-led reforms imposed during the structural adjustment era that have reshaped domestic economies.

Decades of scholarship have highlighted the importance of elite cohesion in ensuring regime durability (Geddes Reference Geddes1999; Huntington Reference Huntington1968; Slater Reference Slater2010). Influential scholars (Levitsky & Way Reference Levitsky and Way2010; Skocpol Reference Skocpol1979; Tilly Reference Tilly1992) highlighted the importance of revolutionary struggle in building elite cohesion. Scholars, working within these traditions, have discussed how ruling coalitions employed symbolic, coercive and distributive state powers to bind elites together even though capitalist accumulation processes may have led to distributional inequalities among elites and the groups they represent (Slater Reference Slater2010). The work of Doner et al. (Reference Doner, Ritchie and Slater2005) has contributed to focusing attention on why elites may have committed to invest in long-term development goals (like structural transformation) rather than pursuing their own narrow self-interest. Analysing elite cohesion highlights why, in varied political systems, ruling elites may centralise decision-making authority and investments to prioritise structural transformation. Doner et al. (Reference Doner, Ritchie and Slater2005, p. 329) argued that ‘the political origins of developmental states can be located in the simultaneous interplay of three separate constraints’ – broad coalitional commitments, severe security threats and scarce resource endowments. Doner et al. (Reference Doner, Ritchie and Slater2005, p. 340) argued that ‘broad coalitions and external security needs constitute two critical claims on scarce revenues that press ruling elites to become revenue maximisers’. ‘Systemic vulnerability’ motivated political leaders to recognise that coherent bureaucracies and public–private linkages sustained coalitions, state survival and maximised their time in office (Doner et al. Reference Doner, Ritchie and Slater2005). The application of the ‘systemic vulnerability’ thesis to East Asian development has been criticised for ignoring details of empirical examples. It has some relevance in Rwanda and Ethiopia (two of Africa’s fastest growing countries), with both having some characteristics of ‘systemic vulnerability’. However, it does not explain why the ruling Rwandan Patriotic Front (RPF) remains in power and Ethiopia’s Ethiopian People’s Revolutionary Democratic Front (EPRDF) does not.

Scholarship has also discussed how ruling coalitions manage political contestation between elites and other organised groups, as well as inter-elite rivalries. Building on the ‘systemic vulnerability’ thesis, Slater (Reference Slater2010) compared the divergent experiences across Southeast Asia, showing that elite collective action underpinned state strength and the durability of authoritarian regimes whilst elite factionalism reduced the likelihood of elite collective action. He argued that the strongest foundation for elite collective action is ‘protection pacts’ where broad elite coalitions are unified by shared perceptions of endemic threat. This drove elites to collectively promote structural transformation to reduce their vulnerabilities to threats. When elites come together in protection pacts, they effectively intervene in the economy and redistribute resources to sustain their rule. Slater (Reference Slater2010) contrasts ‘protection pacts’ with weaker ‘provision pacts’ where elites perceived pressures from below as manageable and weaker institutions emerge consequently. In contrast, Vu (Reference Vu2007, Reference Vu2010) argues that intra-elite and elite-mass interactions at the time of state formation determine the capacity of governing states to build cohesive governance structures. More fine-grained research into what determines elite actions has highlighted how leaders and ruling parties have strengthened coalitions through patronage to sustain their rule. For example, within the African political economy literature, Boone’s (Reference Boone1992) concept of ‘politicised accumulation’ focuses on how ruling elites used varied strategies of managing distribution and wealth to retain popular support within their coalitions. A range of literature has also described how African governments have either politicised accumulation for crony capitalism, enhanced the position of the bureaucratic bourgeoisie or ensured private capital invested in their election campaigns (Arriola Reference Arriola2013; Bayart Reference Bayart1989; Shivji Reference Shivji1976; Tangri Reference Tangri1998).

Unlike such analyses, political settlements research, first developed by Khan (Reference Khan, Harriss, Hunter and Lewis1995, Reference Khan2010), does not simply focus on how accumulation is politicised to satisfy the self-seeking motives of elites or to maintain elite cohesion or regime stability. Accumulation is also politicised to invest in structural transformation (though such attempts may often fail). To put it simply, politics should not only be considered a constraint to structural transformation. Politics is also the reason why structural transformation is achieved. The political settlements literature was originally developed as a response to the popularity and influence of New Institutional Economics (NIE). NIE focused on how institutions – sets of rules that structure the interactions of agents – were necessary to ensure social interaction was possible (North Reference North1990). NIE recognised that the market did not function optimally in late-developing countries. However, the state’s role was restricted to creating a facilitating environment for the private sector. NIE influenced the Good Governance approach, which was promoted by IFIs and multilateral institutions. IFI proposals encouraged governments of developing countries to adopt market-enhancing institutions (such as secure private property rights, control of corruption) and the ‘right’ political institutions (democracy, rule of law) to promote catch-up development. Khan (Reference Khan, Harriss, Hunter and Lewis1995) introduced the political settlements framework to argue that the impact of institutions depends on the balance of power between classes and groups affected by that institution. Creating new institutions was likely to face a political cost (or ‘transition cost’) from groups that disagreed with such change (Khan Reference Khan, Harriss, Hunter and Lewis1995). NIE failed to recognise that institutions were established through political contestation and were shaped by underlying power relations (Di John Reference Di John2009).

NIE and most of the literature on developmental states had opposing interpretations of how East Asia developed, with the former favouring a more market-enhancing interpretation and the latter arguing that state intervention played a more central role. However, both approaches underplayed the role of politics. Even where politics was acknowledged within the developmental state literature (Doner et al. Reference Doner, Ritchie and Slater2005; Evans Reference Evans1995), there was very little attention to the inherent contestation that is characteristic of capitalist accumulation processes. Khan (Reference Khan2010) argued for highlighting the importance of politics shaping economic policy, finding that informally organised political power was necessary to generate a distribution of benefits between classes and elites. The limited size of the formal productive economy meant that formal institutions could not generate rents in a way that guaranteed a minimum level of stability. Instead, informally distributing such benefits was necessary to create an asset-owning capitalist class and for this class to expand its interests with the help of the state. Hirschman’s (Reference Hirschman1968) work also highlighted the inevitable political contestation that characterised industrial policy processes. He argued that the lack of political power of industrialists can be explained by the lack of industrial exports. Agrarian elites often retain substantial power in society because they contribute to a larger share of exports. In some cases, agrarian elites have used this power to block industrialisation, which would reduce their political and economic power within society.

The political settlements framework has widely influenced donor programming. Research employing the framework has been published across several fields, including African political economy, development studies, international political economy, political science, development economics, urban studies and geography.Footnote ⁹ As political settlements analysis (PSA) became increasingly popular, particularly within donor programming and as donor-funded research centres chose to employ it in their analysis, the application of PSA departed from its structuralist roots. Since then, scholars have aligned it with dominant NIE perspectives (North et al. 2009), as well as tried to find consensus (Kelsall et al. Reference Kelsall, Schulz, Ferguson, Vom Hau, Hickey and Levy2022), while others have tried to reclaim the structuralist roots of the framework (Behuria et al. Reference Behuria, Buur and Gray2017; Gray Reference Gray2018, Reference Gray, Cheeseman, Bernard and Husaini2019; Whitfield et al. Reference Whitfield, Therkildsen, Buur and Kjaer2015). The political settlements framework employed in this book aims to advance the structuralist roots of the framework. In particular, the book hopes to contribute to advance political economy analysis by explaining the significance of multi-scalar processes, particularly transnational dynamics within which internal and external factors have combined under contemporary globalisation. In this way, it clarifies how PSA is distinct from work on development politics that tend to downplay the interaction of internal and external factors.

Rwanda in African Political Economy

Defying pessimistic predictions about African development, there have been several African countries that have been hailed as economic success stories. It is common to use the moniker of ‘developmental states’ within the literature to characterise a wide range of growth spurts across Africa.Footnote ¹⁰ Mauritius and Botswana are often highlighted as the closest examples of African ‘developmental states’, with many drawn to these cases because they are also democratic (Mkandawire Reference Mkandawire2001).Footnote ¹¹ Mauritius did achieve significant structural transformation. It transformed its economy, and many of its domestic firms also developed significant technological capabilities in tourism and apparels (Brautigam Reference Brautigam2005; Whitfield Reference Whitfield2023). Yet the rise of its offshore sector gradually reduced prioritisation of other sectors (particularly agriculture and manufacturing). Crucially, as the offshore sector faces an uncertain future, the government is showing few signs of encouraging existing domestic capital to invest in new productive sectors. This is partly because dominant Franco-Mauritian businesses increasingly concentrated control over the economy. These conglomerates have rarely led investments in new sectors, instead tending to follow others (Behuria Reference Behuria2023). Botswana, in comparison, benefited significantly from diamond exports on which it remains heavily reliant (Good Reference Good2008; Hillbom Reference Hillbom2008). Though Botswana experienced some industrial growth in the 1990s, the government has failed to encourage the growth of a domestic capitalist class outside agriculture. The Botswana government’s capacity to sustain growth is likely to be threatened in the future, especially if diamond prices fall.

Over the last decade, hopeful observers have become increasingly drawn to two new African economic success stories: Ethiopia and Rwanda. Both countries have been characterised by more authoritarian governance than Mauritius or Botswana. When the now-defunct EPRDF assumed power in 1991, the then prime minister Meles Zenawi embarked on a development strategy, which was influenced by East Asian development successes and writings on the developmental state (De Waal Reference De Waal2013; Gebremariam Reference Gebremariam2024; Zenawi Reference Zenawi, Noman, Botchwey, Stein and Stiglitz2012). The EPRDF was widely lauded for its prioritisation of a manufacturing-based development strategy. Yet services still comprised a higher proportion of GDP than manufacturing in the 2010s (Lavers Reference Lavers2023). The Ethiopian government was successful in attracting leading foreign apparels firms to establish factories in Ethiopia, with apparels exports increasing more than ninefold between 2010 and 2017 (Whitfield & Staritz Reference Whitfield, Staritz, Oqubay and Lin2020). Literature on Ethiopia’s industrial strategy (Oqubay Reference Oqubay2015) has shown that across most industrial sectors, few domestic firms had significant technological capabilities and the state was unable to encourage a transfer of technology. Any progress has been undone in recent years with the Tigray war that began in 2020. Lavers (Reference Lavers2023) argues that the EPRDF’s development strategy was unable to meet distributional demands, and this was the reason why the country descended into civil war. However, the failure to meet distributional demands is a concern across most economies globally in the twenty-first century. There are broader questions of whether capitalism is even capable of generating sufficient and secure employment to provide a living wage to the majority in societies (Bernstein Reference Bernstein2010). Also, distributional demands were never met throughout EPRDF’s rule, so Lavers’ arguments do not fully explain why civil war erupted only in 2020.

While Ethiopia’s development success was more directly focused on manufacturing, Rwanda’s strategy has been very different. The ruling RPF, led by President Paul Kagame, has envisioned landlocked, densely populated Rwanda to become a land-linked services hub, focused on finance, tourism, sports and services. As the book will highlight, there is significant evidence of unmet distributional demands in Rwanda. Yet the RPF remains in power. As a result, Lavers’ (Reference Lavers2023) arguments do not seem to travel convincingly. There is something more that needs to be analysed given the nature of late development under contemporary globalisation.

These African development success stories have very little in common with one another. Perhaps, the only factor that Mauritius, Botswana, Rwanda and Ethiopia have in common is they all faced geographical vulnerabilities. They defied the predictions of many mainstream economic geographers (Collier & Gunning Reference Collier and Gunning1999; Krugman Reference Krugman1991; World Bank 2008), who argued that suitable geography is a predictor of development outcomes. Rwanda, Botswana and Ethiopia are landlocked, while Mauritius is an isolated island. Another similarity across these economies is that services represent more of a contribution to GDP than agriculture or manufacturing (as is the case with most African economies). Yet each African country retains a different endowment structure, as well as specific political economic histories that shape the nature of its integration into the global economy. If we acknowledge the unique nature of these experiences, there must be more attention to inductively study the historical political economy of specific country experiences, especially those cases perceived to be successful. A few recent books have attempted such studies recently, including Lavers (Reference Lavers2023) on Ethiopia and Whitfield (Reference Whitfield2018) on Ghana, but such studies remain rare. Indeed, the seminal texts (Amsden Reference Amsden1989; Johnson Reference Johnson1982; Wade Reference Wade1990) on East Asian developmental states all focused on specific countries (Japan, Korea, Taiwan). Surely, African countries warrant the same attention rather than only being applauded for resembling the experiences of other countries. This book takes a ‘possibilist’ approach, following Albert Hirschman, like others do (Cramer et al. Reference Cramer, Sender and Oqubay2020). Possibilism is determined by what a country does and what it becomes because of state interventions within the remit of constraints set by the global political economy (Hirschman Reference Hirschman1968). Like Hirschman (Reference Hirschman1971), the book acknowledges that there is no ‘one best way’ to achieve structural transformation.

This book is a conscious effort to highlight the significance of context sensitivity, the lack of which lies at the heart of so many unfulfilled development endeavours (Lepenies Reference Lepenies2008). The following section briefly describes what has been different about Rwanda’s late development trajectory and what has been misunderstood about its apparent economic success so far. Though there has been some degree of sustained growth and even export diversification, there is less evidence of structural transformation. As with other cases mentioned earlier, the limited support for domestic capitalists in acquiring technological capabilities presents a significant vulnerability for Rwanda’s future.

Explaining Rwanda’s Rise

Since the end of the 1994 genocide against the Tutsi, RPF rule has divided opinion in the popular press and within academia. This has resulted in Rwanda being depicted in contrasting ways. Rwanda is often portrayed as a ‘donor darling’ (Marysse et al. Reference Marysse, Ansoms and Cassimon2007). Rwanda’s growth has been lauded by a global support base, including former American and European heads-of-state such as Tony Blair, Bill Clinton and Bill Gates, and business leaders at the World Economic Forum (Grant Reference Grant2010; Sundaram Reference Sundaram2014). A wide range of scholars have applauded the RPF government’s socio-economic achievements and ‘best practice’ initiatives (Chinsinga et al. Reference Chinsinga, Weldeghebrael, Kelsall, Schulz and Williams2022; Collier Reference Collier2010; Hayman Reference Hayman2009; Hickey Reference Hickey2023). Rwanda also made significant progress in achieving Millennium Development Goals (MDGs), particularly in health (Abbott et al. Reference Abbott, Sapsford and Binagwaho2017). The government invests in supporting such positive portrayals, dedicating its resources to improving performances in governance indices. The Rwanda Governance Board (RGB) is tasked with ensuring Rwanda continues to progress in relevant governance indices globally, including the Worldwide Governance Indicators. The Rwanda Development Board (RDB), a government institution operating as a nodal agency to promote foreign investment and economic development, had a specialised Doing Business Unit, which was responsible for improving the country’s position in the World Bank’s Ease of Doing Business Index (DBI).Footnote ¹⁴

Rwanda’s development achievements have contributed to the impression of the country becoming a symbol of ‘best practice’ development initiatives. The government’s perceived ‘effectiveness’ has made it a popular ‘lab’ for development initiatives.

Some heterodox scholars also promote positive images of a rapidly ‘developing’ Rwanda. However, heterodox scholars differ from mainstream observers in their interpretations of the sources of Rwanda’s economic progress. Under heterodox interpretations, Rwanda is likened to East Asian developmental states of the past. It is termed as an ‘incipiently developmental state’ (Booth and Golooba-Mutebi Reference Booth and Golooba-Mutebi2014a), a neopatrimonial developer (Kelsall Reference Kelsall2012, Reference Kelsall2013) or a ‘neo-developmental state’ (Goodfellow Reference Goodfellow2017). Prominent heterodox scholars (Booth & Golooba-Mutebi Reference Booth and Golooba-Mutebi2011) argued that the RPF’s use of party- and military-owned enterprises to steer Rwanda’s growth is similar to the way East Asian governments used closely affiliated firms (e.g. family-owned conglomerates or the chaebol in South Korea) to steer their development strategies.

Negative depictions of Rwanda are as prominent as positive ones, with the Rwandan government’s nation-branding and diplomatic apparatus often in a constant battle to manage how Rwanda is perceived externally. Negative depictions are advanced by prominent scholars and transnational advocacy networks, including Human Rights Watch (HRW) and the Enough Project. Kagame’s government is portrayed as ‘an army with a state, rather than a state with an army’ (Reyntjens Reference Reyntjens2011, p. 2). Scott’s (Reference Scott1998) principles of ‘high modernism’ have been used by some scholars (Dye Reference Dye2016; Huggins Reference Huggins2017; Newbury Reference Newbury, Straus and Waldorf2011) to emphasise the dangers of the RPF’s consolidation of political and economic power.Footnote ¹⁶ These critics argue that a small group of elites concentrate benefits accrued from government policies.Footnote ¹⁷ They also highlight the RPF’s exclusionary policies against the population, which have likely resulted in increasing horizontal inequalities between two ethnic groups: the ruling, minority Tutsis and the majority Hutu population. ‘Horizontal inequalities’ refers to ‘inequalities in economic, social or political dimensions or cultural status between culturally defined groups’ (Stewart Reference Stewart and Stewart2008, 3). Though there is no official data regarding present-day horizontal inequalities in Rwanda, scholarship suggests that it is likely to have increased (McDoom Reference McDoom2011). Scholars (who are critical of the RPF) have argued that the government will eventually be a victim of ‘high modernist’ social engineering or that inequalities between ethnic groups will contribute to a return to civil war (Lemarchand Reference Lemarchand2007; Reyntjens Reference Reyntjens2013).

This book contributes to more recent literature that takes a more nuanced analysis of Rwanda’s development strategy (Ansoms et al. Reference Ansoms, Aoun, Chemouni, Niyonkuru and Williams2022; Chemouni & Dye Reference Chemouni and Dye2024; Heinen Reference Heinen2022; Williams 2017). This is a welcome trend within the academic literature that does not tally with the polarised debate about Rwanda’s development record in the international press and social media. This book highlights that capitalist accumulation is always violent, unequal and characterised by disorder and contestation. Such approaches have previously been adopted to unite what may be seen as contradictory depictions of Rwanda into a consistent explanation (Behuria Reference Behuria2015a; Harrison Reference Harrison2017). The book takes the position that ‘capitalist development is everywhere, and always has been, a messy, non-linear and often brutal process’ (Cramer et al. Reference Cramer, Sender and Oqubay2020, p. 8). Thus, unlike the polarised debate in the international press about Rwanda’s development record, the book’s approach suggests that supposedly contradictory aspects of Rwanda’s development may coexist, as this has generally been the case with late development experiences historically.

Capitalist accumulation is rooted in unequal outcomes associated with processes of primitive accumulation. Primitive accumulation refers to a phase of the appropriation of wealth (or surpluses) through largely non-economic or coercive means during which others are displaced from ownership and access. Elites compete to claim assets during this process. Such processes occur through the direct exploitation of the peasantry. Land and other assets are taken away from individuals, and many are forced into wage labour. As Marx (Reference Marx1976, p. 874) writes, primitive accumulation ‘is a process, which operates two transformations whereby the social means of subsistence and production are turned into capital, and the immediate producers are turned into wage labourers’. Marx recognises primitive accumulation as ‘anything but idyllic’. Even in successful experiences of structural transformation, labour was exploited, and states used coercive methods to encourage capitalist accumulation. East Asian states did not achieve equal gains for elites and the people, and neither did earlier industrialised countries in Europe. Their governments were violent, corrupt (Khan Reference Khan1998) and acted viciously against labour unions (Chang Reference Chang2009). Political order was maintained through a mixture of coercion and co-optation to manage the inevitable instability associated with late development.

Late development under contemporary globalisation has new challenges, as countries navigate a very different global economy compared to the 1950s/1960s. But as mentioned already, many countries have also experienced services-driven growth, which is more externally reliant than manufacturing-based growth. Late development in the twenty-first century is thus a significantly more transnational challenge than before. It is subject to a wider constellation of external actors shaping national development processes ideologically, as well as materially.

The Research Questions and Arguments

There have been several influential books that examined RPF rule since 1994 (e.g. Kimonyo Reference Kimonyo2019; Reyntjens Reference Reyntjens2013; Straus & Waldorf Reference Straus and Waldorf2011). However, there are very few systematic analyses of Rwanda’s economic rise. Rwanda’s experience can help us understand more about the contemporary late-development experience because the RPF government has prioritised a services-first strategy (to become a regional and continental hub) explicitly to transform its broader national economy. Most African countries have experienced a similar economic trajectory to Rwanda, characterised by transformation from low-value agriculture to low-value services albeit to varying degrees (McMillan et al. Reference Mcmillan, Rodrik and Verduzco-Gallo2014).

This leads to the central research question: How does Rwanda’s apparent economic development success contribute to our understanding of contemporary late development? The book answers this question by inductively studying what the Rwandan experience can add to our understandings of the constraints facing contemporary late development. In doing so, it takes up Mkandawire’s (Reference Mkandawire2001) call to study African development processes on their own terms. Or as Mamdani (Reference Mamdani1996a) puts it: to study Africa as a process rather than analogy.

The Rwandan government has chosen to address its geographical and structural vulnerabilities by embarking on a services-first trajectory. Service sectors rely much more on foreign investment and maintaining a positive reputation externally. This makes services-based development significantly more reliant on foreign actors and subject to transnational contestation. In the process, Rwanda has experienced rapid growth and significant export diversification. Yet the RPF government has failed to support domestic firms to acquire technological capabilities and, as such, Rwanda has not achieved structural transformation.

This leads us to the book’s related ‘two-way’ question, which helps descriptively show why services-based strategies have taken shape and what challenges they have led to: How have vulnerabilities shaped Rwanda’s choice of development strategy and what vulnerabilities have characterised that strategy? Literature on elite vulnerability has highlighted how external and internal threats have been important in building elite cohesion (Slater Reference Slater2010; Vu Reference Vu2010). While elite vulnerability within Rwanda has built commitment to a paradigmatic goal of achieving self-reliance through export diversification and services growth, it has also contributed to marginalising the role of domestic capitalists.

Financing Rwanda’s development strategy heavily depends on external capital, as well as maintaining a positive global reputation. The most significant threat to the RPF’s rule is from transnational coalitions of dissident Rwandans, which challenge the RPF’s external legitimacy and aim to mobilise marginalised segments of the domestic population, both elite and non-elite. Unlike critics of Rwanda, this book does not see increasing horizontal inequalities or distributional challenges as enough of a threat on their own. Instead, it follows literature (Cramer Reference Cramer2006; Tilly Reference Tilly2003) that highlights political mobilisation against incumbents is most successful when rival elites can mobilise popular grievances into collective action. In Rwanda, though grievances remain latent in society, the RPF government has successfully thwarted political mobilisation. Disgruntled elites remain divided. Marginalised rural populations are unable to mobilise against RPF rule. Financial support to mobilise rival coalitions is consistently thwarted because of the RPF’s centralised control of the economy. Elite cohesion is sustained through appeals to historical collective memories, development performance, as well as state coercion.

The most significant threats to elite cohesion are transnational. Crucially, the most significant opponents of government now reside abroad, though many of these opponents have failed to form a coherent opposition. Neighbouring governments in countries like Uganda and the Democratic Republic of the Congo (DRC) have sometimes been accused by the RPF government of supporting dissident groups abroad. The RPF’s rivals, as well as critics, consistently resort to strategies of discrediting the RPF’s external portrayal of its own achievements, as well as its supposedly progressive attributes. This is strategically crucial because it threatens the financing of the Rwandan economy. The RPF prioritises the promotion of the portrayal of Rwanda’s development achievements abroad, as well as its progressive attributes, since it is directly linked to financing the political settlement and the broader development project.

This book is the first systematic analysis of Rwanda’s services-oriented development strategy, highlighting how key sectors have been transformed in line with broader goals of becoming a ‘hub’ of regional economic activity and of upgrading and diversifying the economy to achieve self-reliance. In all sectors, the book shows that the most significant diversification successes have been a result of government-affiliated investments. The major challenge for the RPF’s economic strategy is its failure to develop partnerships with domestic ‘private’ businesspeople, with nearly all former prominent businesspeople either no longer alive or opponents of the government abroad. In this way, Rwanda has fallen prey to a similar failing as most African countries in being unable to develop effective state–business relationships with domestic capitalists (Whitfield et al. Reference Whitfield, Therkildsen, Buur and Kjaer2015).

Methodology

Research for this book began in October 2011 during my PhD. Fieldwork in Rwanda was conducted for a total of sixteen months between October 2011 and December 2019. This included ten months of research between 2011 and 2012, followed by one-month fieldwork stints in 2013, 2015, 2016, 2017, 2018 and 2019. Since 2019, interviews have also been conducted remotely both with Rwandan government officials and donors, as well as with other stakeholders working in Rwanda. Most interviews were conducted in Rwanda’s capital city, Kigali, where I was based for most of my fieldwork. However, research was also conducted outside Kigali, and interviews were conducted in factories, mining sites, tea factories, coffee farms and secondary cities, including Gisenyi and Butare. Interviews were also conducted with embassy officials and the RPF’s political opponents abroad. This book relies on qualitative methods, including structured and semi-structured interviews; informal exchanges outside the official interview setting, and archival work in ministry offices; statistical data from government sources and public databases such as the World Bank database. Where possible, data has been triangulated with several sources, including interviews and newspaper articles.

Over 580 interviews have been conducted. In the very early stages of fieldwork, interviews were structured and questions were set prior to the interview. However, interviews later became less structured and more fluid. Most interviews took place within formal office environments. Semi-structured interviews are commonly used across the social sciences to achieve such purposes (Sarantakos Reference Sarantakos2005), primarily through creating a ‘conversational space’ (Elliott Reference Elliott2005). Some interviews were undertaken in ‘informal’ environments: for example, some military officials preferred to meet in bars and coffee shops rather than in offices. Building rapport with informants contributed to creating ‘conversational spaces’ and trust. Some respondents were interviewed several times and over several years.

Research on Rwanda’s political settlements did not follow a deductive strategy. For example, no initial assumptions were made based on using Khan’s (Reference Khan2010) 2 × 2 matrix on political settlements or any other similar matrices that have been later developed (Kelsall et al. Reference Kelsall, Schulz, Ferguson, Vom Hau, Hickey and Levy2022; Levy Reference Levy2014). Instead, I adopted a strategy of historically analysing how power relations evolved across the macroeconomy and across several sectors. A more detailed analysis of the methodology to study PSA was discussed in Behuria et al. (Reference Behuria, Buur and Gray2017). The study of Rwanda’s PSA (and broader development trajectory) has been an iterative process, constantly reflecting on how the specifics of Rwanda’s development experience did not neatly reflect what has been published in existing literature about past development experiences elsewhere.

The challenges of undertaking fieldwork in Rwanda have been the subject of consistent academic discussion. Several prominent academics have been asked to leave Rwanda because they did not have the required permissions or because government officials claimed they did not follow required procedures (Jessee Reference Jessee2012). Over the last decade, most academic attention has been dedicated to the study of the genocide, reconciliation (and the gacaca courts), ethnicity, different aspects of governance, health, urban transformation or education. There is very little academic research dedicated to the study of macro-political economy or examining the determinants of Rwanda’s growth trajectory from a qualitative perspective. As a result, I was able to gain significant access primarily because very few foreign researchers undertook this kind of research in Rwanda. Also, this seemed to be a story that the government wanted to tell. As a senior official in the Office of the President said:

In my initial years researching Rwanda, access (though not always easy) was not a major impediment. This book, as you will see, does have an appreciation of what Rwanda has achieved economically since the genocide. It is primarily concerned with analysing how the Rwandan experience contributes to our collective understanding of pursuing late development under contemporary globalisation. However, as the years have proceeded, certain Rwandan government officials and some ministries have denied me access for interviews. This is understandable given the pressures government bureaucrats are under to perform and the dangers of being confronted with criticism. Gradually, as with many researchers studying Rwanda, my access was curtailed (though it is not entirely closed).

The book describes how the Rwandan government prioritises maintaining certain external representations in the foreign media. Since this is the area of most significant contestation and vulnerability for the Rwandan government, it is perfectly understandable that access to research may not be as open as it is in other countries. Elsewhere, I have discussed my research experiences in more detail (Behuria Reference Behuria2015a).

As with all countries, there are concerns over the data presented regarding Rwanda’s economic trajectory. It is crucial to recognise that some of the economic data I present in the book is contested (particularly with regard to agricultural production and economic growth). I have tried to highlight the data limitations throughout the book. These limitations, particularly in relation to specific economic indicators, have been discussed widely within the literature (Ansoms & Rostagno Reference Ansoms and Rostagno2012; Desiere et al. Reference Desiere, Staelens and D’Haese2016; Heinen Reference Heinen2022). However, the book does not try to make claims about the scale of Rwanda’s development achievements, so such discussions are not central to the contributions it intends to make.

Structure of the Book

This book has ten chapters, including this one. Chapter 2 introduces PSA in detail. The political settlements framework employed here is an adapted form of Mushtaq Khan’s work, elaborating the structuralist and historical materialist roots of the approach. Chapter 3 applies the PSA elaborated in Chapter 2 to the Rwandan case while also highlighting Rwanda’s key development achievements. This chapter provides a snapshot of Rwanda’s political settlement, applying key concepts relating to the determinants of holding power (economic structure, rents, ideas and ideology, violence and conflict) to the Rwandan case. The chapter highlights both why services-based development was pursued (partially because of elite vulnerability) and why this strategy obstructs structural transformation from being achieved.

Chapter 4 presents a historical political settlement analysis of Rwanda, highlighting how Rwanda’s future development trajectory is contested transnationally, as the RPF’s legitimacy is heavily based on maintaining a positive image of its nation brand. Rwanda’s political stability is maintained by ensuring popular resistance is contained and domestic elites remain disincentivised to challenge the ruling coalition. Chapter 5 examines the politics of financing the RPF’s development strategy, highlighting how the RPF innovated to mobilise domestic resources to fund strategic projects. However, elite vulnerability, as well as the need to present itself as abiding by donor suggestions and best-practice financial sector reforms, has constrained the RPF’s capacity to do this.

Chapter 6 describes the evolution of policies in Rwanda’s agriculture sector. Based on existing literature, it highlights how significant rural inequality persists. Resistance has been contained by significant state control and the threat of violence. It highlights how the RPF’s transformation of the agriculture sector has focused on intensifying agricultural production through land consolidation (with limited success), exporting higher-value traditional exports (such as coffee and tea) and exporting new agricultural products. In line with the political settlement, Rwanda’s most ambitious diversification has been achieved through supporting its own state-affiliated firms, sometimes in partnership with philanthropists. Evidence is presented of reduced reliance on prominent domestic businesspeople across sectors.

Chapter 7 describes the double-edged dependence in Rwanda’s mining sector, highlighting both how the import of minerals from the DRC has sustained the Rwandan economy but also how reliance on Congolese minerals is a source of elite vulnerability for the RPF’s ruling coalition. Dependence on the DRC’s trading networks represents a significant vulnerability for the RPF. Prominent RPF and RDF officials gain power and loyalty through their positions in these networks. The accumulation of power by such officials represents a threat to the Kagame-led RPF ruling coalition’s centralised control over political and economic decision-making.

Chapter 8 describes Rwanda’s services-based strategy, highlighting how external reliance has become the key vulnerability of Rwanda’s political settlement. The Rwandan government’s use of nation-branding is a central aspect of its strategy. To reduce over-reliance on specific foreign actors, the government has developed relationships with a broader range of foreign countries. However, the RPF’s strategy has become detached from two key needs domestically: integrating domestic capital and meeting domestic employment needs. The RPF’s services-oriented ‘hub’ strategy is the government’s remedy to its geographical vulnerabilities, hoping to achieve self-reliance through having the most advanced governance in the region and thereby convincing external actors of the potential of investing in Kigali to make it a regional hub.

Chapter 9 describes the relative neglect of industrial sectors and the effects this has had on sustaining economic transformation in the long term. The chapter highlights how the RPF has neglected existing domestic industrial capitalists increasingly, with industrial growth led by either state-affiliated firms or foreign investors. High-profile industrial investments by foreign investors have paid little attention to domestic linkages, and there are few signs that domestic firms may emerge to benefit from a transfer of technology.

Chapter 10 concludes by highlighting the central contributions of the book, highlighting how contemporary late development, focused on services, has made development a more transnational challenge. It shows how services-based development lends itself to marginalising domestic capitalists and increasing reliance on foreign capital, which impedes possibilities for structural transformation. The book argues for the importance of more historically informed political economy analysis of single-country cases in Africa to highlight the diversity of experiences on the continent under contemporary globalisation. It ends with thoughts on Rwanda’s development strategy going forward, including the potential future trajectory of RPF-led Rwanda.