2.1 Political Economy Approaches to Economic Development

The neo-institutionalist literature in economics has revived the debate about the causes of (under)development (North, Reference North, Harris, Hunter and Lewis1995; Greif, Reference Greif2006; Acemoglu & Robinson, Reference Acemoglu and Robinson2012). This approach contends that strong institutions tend to achieve higher rates of economic development because these institutions provide the necessary incentives for investment, innovation, human capital accumulation, and economic exchange. By now, there is an established consensus that “institutions matter” for long-term growth (Rodrik et al., Reference Rodrik, Subramanian and Trebbi2004; Acemoglu et al., Reference Acemoglu, Johnson, Robinson, Aghion and Durlauf2005). Though disagreement persists over what the best institutional design looks like, the dominant view has it that liberal political and economic institutions trump all other arrangements. Daron Acemoglu and James Robinson have advanced this perspective in their account on “why nations fail” (Acemoglu & Robinson, Reference Acemoglu and Robinson2012). According to them, “inclusive” political and economic institutions associated with free markets, property rights, the rule of law, and pluralist democracy are prerequisites for economic growth. “Extractive” institutions divert resources away from their most efficient and socially optimal use by powerful elites. In a nutshell, Acemoglu and Robinson’s approach arguably proposes one universally applicable model for development: The liberal market economy, which is enforced and sustained by liberal democracy, thereby disavowing or even denying indigenous institutional constellations (Madra et al., Reference Madra, Bengi and Adaman2025, 7–8).

The neo-institutionalist approach has encountered severe criticism, however. To begin with, the dichotomy of inclusive and exclusive institutions is too simple to capture the real-world variety of institutional arrangements, their historical origins, and embeddedness in a hierarchically structured, globalized economy (Chang, Reference Chang2011; Ince, Reference Ince2024). Furthermore, this perspective, focusing on long-run growth rates, fails to explain the volatile growth episodes so prevalent in emerging economies (Pritchett et al., Reference Pritchett, Sen and Werker2017, 12). Finally, such a perspective cannot explain why many countries industrialized under authoritarian rule with heavy state intervention – the most obvious contemporary examples being East Asian “developmental states” and, more recently, China (Sen et al., Reference Sen, Pritchett, Kar and Raihan2018; ten Brink, Reference ten Brink2019). At the very least, an endorsement of these facts calls for more nuanced, context-sensitive institutional analyses devoted to mid-level theorizing (Merton, Reference Merton1968), rejecting universally valid one-size-fits-all accounts.

The developmental state literature mentioned earlier is one major example of a deep engagement with country case studies and comparative historical analysis (Johnson, Reference Johnson1982; Amsden, Reference Amsden1989; Wade, Reference Wade1990; Evans, Reference Evans1995). Primarily occupied with explaining the growth miracles of first- and second-tier East Asian late industrializers, this literature has informed a broad and lively debate about the role of the state in economic development (Rodrik, Reference Rodrik2009; Stiglitz & Lin, Reference Stiglitz and Lin2013). After the 2008 financial crisis, this discussion has gained considerable momentum even in OECD countries (Mazzucato, Reference Mazzucato2013; Breznitz & Gingrich, Reference Breznitz and Gingrich2025). However, given the peculiarities of the East Asian developmental states and the specific external conditions permissive of their export-led growth models (Haggard, Reference Holland2019, 54–9), Johnson (Reference Johnson1982, 307) warned early on: “the dangers of institutional abstraction are as great as the potential advantages” of adopting Japan’s growth strategy in other contexts. Hence, the “quest for capacious and strategically oriented state institutions has ultimately proved elusive,” as only a handful of countries have achieved such a high degree of state capacity (Naseemullah, Reference Naseemullah2023, 7). Over time, the notion of developmental state has become a rather abstract buzzword detached from its origins and main conceptual elements (Fine, Reference Fine, Fine, Saraswati and Tavasci2013). Accordingly, the debate has shifted toward new directions, such as the question of whether a capacious state bureaucracy is possible at all in times of financialization and neoliberalism (Evans & Heller, Reference Evans, Heller, Leibfried, Huber, Lange, Levy and Stephens2015; Fine & Pollen, Reference Fine, Pollen, Fagan and Munck2018).

Hence, both neo-institutionalist and developmental state approaches tend to imply a (normative) linear model of development, which countries would only need to apply fully in order to achieve high levels of economic growth. Furthermore, development is treated as a purely domestic process, which, in our view, underestimates the structural conditions of the global political economy. Both approaches usually highlight the flaws of development in different countries, as the record of successful adopters of either model is quite small. In particular, Korea and Taiwan are seen as role models for development, successfully transitioning from a low- to high-income status in the twentieth century. However, these cases are economies with particular geopolitical and historical-institutional legacies, making it hard to generalize their developmental state model globally (Naseemullah, Reference Naseemullah2023). In addition, these countries recorded robust export growth throughout a large part of their development trajectory, highlighting the particular global economic conditions of the second half of the twentieth century in which they could grow through such a model. These structural conditions of the global political economy, so we contend, have changed in such a fundamental way that it is impossible to directly compare the growth trajectories of the 1960s and 1970s with those of the 2000s and onward.

This argument has also been highlighted by approaches that explain growth and development exogenously, locating the sources of and conditions for (under)development at the level of the global political economy. Drawing on dependency theory, world systems, postcolonial, and other critical approaches (e.g., Cardoso & Faletto, Reference Cardoso and Faletto1979; Wallerstein, Reference Wallerstein1976), this important strand of research investigates the hierarchical structure of the global economy and its constraining impact on economic development in peripheral countries. Contemporary versions of this argument underline three aspects of today’s global capitalism: global commodity markets, financial flow cycles, and global production networks. Studies on global commodity markets highlight the dependence of many developing and emerging economies on exporting commodities. These dependencies are profound, given the cyclical nature of commodity price movements over time (Akyüz, Reference Akyüz2022; Schedelik et al., Reference Schedelik, Nölke, May and Gomes2023). Similarly, scholars working on “dependent” or “subordinate” financialization highlight the severe repercussions of short-term financial flows on these economies (Bonizzi et al., Reference Bonizzi, Kaltenbrunner and Powell2022; Lapavitsas & Soydan, Reference Lapavitsas and Soydan2022). Finally, the literature on global production networks argues that developing countries often are integrated in a subordinate position in these networks, being responsible for low-tech production, low-skills employment, harsh working conditions, and profound pollution (Wang et al., Reference Wang, He and Song2021). In all of these regards, the fate of peripheral economies is perceived as more or less determined by the overwhelming forces of the global economy, leaving little room for economic agency (Petry & Nölke, Reference Petry, Nölke, Petry and Nölke2025; Schedelik & Nölke, Reference Schedelik and Nölke2025; on this see also Avlijaš & Gartzou-Katsouyanni, Reference Avlijaš and Gartzou-Katsouyanni2024, Reference Avlijaš, Gartzou-Katsouyanni and Regini2025).

All exogenous–focused approaches arguably suffer from an underestimation of domestic politics beyond elites and state bureaucracies. This shortcoming has been addressed by the political settlements analysis approach (Khan, Reference Khan, Khan and Jomo2000, Reference Khan2010). It emphasizes how the distribution of power in society, including formal institutions but also informal networks and patron-client relations, can support or thwart developmental goals. Particularly in developing economies, where the economic structure is dominated by primary commodities, there is a need to promote economic diversification and technological upgrading. Contrary to the dominant neo-institutionalist view, certain types of rents can contribute positively to development. In this context, rents are broadly understood as “excess incomes, which do not exist in perfect markets” (Khan, Reference Khan, Khan and Jomo2000, 21). “Rents for learning” and “Schumpeterian rents,” which are necessary for the promotion of economic diversification, indigenous technological capabilities, and innovation, do not necessarily equate with waste and inefficiencies. As long as the distribution of power allows those rents to be used productively, state capture can be avoided. Traditionally, the political settlements literature lends itself to detailed country case studies, but recent endeavors have also tried to build a more quantitative approach (Kelsall et al., Reference Kelsall, Schulz and Ferguson2022; for a recent review, see Behuria, Reference Behuria2025).

By employing the growth model approach for developing and emerging economies, we aim to offer an alternative route mediating between the grand theories of neo-institutionalism, developmental state, and dependency. The growth model perspective is both more pragmatic and realist: It is pragmatic because it accounts for the incremental improvements of growth-enabling institutions and policies, knowing that hardly any country can “imitate” the trajectories of Korea, Taiwan, or – nowadays – China. It is realist in the sense that it treats development conditions as they are, not as they should be. Crucially, this entails that the task of development is first and foremost to minimize vulnerability, which is the major factor for stagnant or negative growth in the first place.

To summarize, existing political economy approaches to economic development have considerable strengths, but also major shortcomings. Regarding the latter, a preference for universally valid one-size-fits-all accounts seems to be the most prominent one. This is not only a problem for large parts of the neo-institutionalist literature, but also for part of the developmental state discussion, which tends to abstract from the particular circumstances of the East Asian cases during the mid twentieth century. Moreover, the recent rejuvenation of dependency approaches thus far tends to underplay the economic agency of emerging economies. Against this backdrop, typological approaches in CPE have the potential to complement existing perspectives on development by highlighting opportunities for macroeconomic agency in a context-sensitive way.

2.2 Approaches in Comparative Political Economy: Varieties of Capitalism and Growth Models

In contrast to the neo-institutionalist focus on a narrow set of institutions (e.g., property rights and the rule of law), CPE approaches build on a much richer set of socioeconomic institutions that account for the multidimensionality of growth trajectories. Along with regulation theory and the national business systems approach, particularly the VoC paradigm has been a key reference point in the CPE literature. This scholarship generally assumes that country-specific institutions matter for economic performance and that different institutional configurations are conducive to achieving economic and social objectives. Institutions are understood in a pragmatic way as solutions for coordination problems of firms. More specifically, it argues that key institutional spheres (such as corporate governance, financial system, industrial relations, or education and training), the identification of cross-cutting coordination mechanisms (inter-firm networks and associations versus competitive markets and formal contracts), as well as the notion of (positive) complementarities among institutions, allow for a parsimonious analysis of capitalist systems.

Crucially, VoC scholarship introduced ideal types of capitalist economies as analytical tools. Drawing on a long-established research tradition (Shonfield, Reference Shonfield1965; Streeck, Reference Streeck2010, 6–15), Peter Hall and David Soskice made the now-classical distinction between liberal market economies and coordinated market economies (Hall & Soskice, Reference Hall and Soskice2001, 8). In the former, firms coordinate themselves primarily through competitive markets and formal contracts, whereas in the latter, they rely principally on inter-firm networks and associations. In contrast to the convergence thesis predominant in the 1990s, claiming that all capitalist economies would adapt themselves to the “superior” Anglo-Saxon model (Streeck Reference Streeck2010, 15–7), VoC argues that both types can be economically successful and internationally competitive, even in the long run. This idea rests on Hall and Soskice’s notion of “comparative institutional advantage,” which states that “the institutional structure of a particular political economy provides firms with advantages for engaging in specific types of activities there” (Hall & Soskice, Reference Hall and Soskice2001, 37). As it does not focus on absolute, but comparative institutional advantages of nations, it is able to account for cross-national variation in product specialization.

Recent research has extended the VoC perspective, identifying four main types of emerging market capitalism (Schedelik et al., Reference Schedelik, Nölke, Mertens and May2021; Friel, Reference Friel2024): (1) a dependent market economy, originally envisaged for Central and Eastern Europe (Nölke & Vliegenthart, Reference Nölke and Vliegenthart2009); (2) a hierarchical market economy, modelled after Latin American economies (Schneider, Reference Schneider2013), but also identified in Turkey (Kiran, Reference Kiran2018); (3) a state-permeated market economy, found in East Asia (Carney, Reference Carney2016), China (ten Brink, Reference ten Brink2019), India and temporarily in Brazil (Nölke et al., Reference Nölke, ten Brink, May and Claar2020); and (4) a patrimonial market economy (Becker, Reference Becker2013), as depicted for the Arab world (Schlumberger, Reference Schlumberger2008) as well as for Russia and other former Soviet countries (Becker & Vasileva, Reference Becker and Vasileva2017). Focusing on supply-side institutions and firm characteristics, the VoC research program, however, has largely tended to ignore three major points: demand-side macroeconomics, (distributive) politics, and a country’s insertion in the global economy.Footnote ³ These points relate to the source and composition of economic growth (both domestic and external) as well as its stability through (domestic) political support and international factors such as financial flows or commodity price movements (Nölke et al., Reference Nölke, May, Mertens and Schedelik2022; Schedelik et al., Reference Schedelik, Nölke, May and Gomes2023). Since growth trajectories in emerging economies are more vulnerable and volatile than in advanced ones, these points are key for the purpose of this Element.

The GM perspective provided a response to the shortcomings of the VoC approach. It argues that national economies grow based on distinct patterns of demand (Baccaro & Pontusson Reference Baccaro and Pontusson2016; Baccaro et al., Reference Baccaro, Blyth and Pontusson2022). Instead of focusing on institutions and firm-based production systems, this perspective emphasizes how economies generate and sustain demand to fuel growth. Growth models in this perspective are defined descriptively based on the decomposition of GDP growth and the respective leading demand components, differentiating between consumption-led (e.g., the UK) and export-led models (e.g., Germany), for instance (Baccaro et al., Reference Baccaro, Blyth and Pontusson2022). Crucially, the GM approach argues that such growth models are – to some degree – politically constructed and maintained. Governments shape particular models through their political choices and the respective macroeconomic policies. Business groups, unions, and voter segments form coalitions to support or contest a given model. Some GM analyses have used the Gramscian-inspired concept of a dominant social bloc to denote the political constellations of an entrenched bipartisan consensus on a growth strategy and crucial economic policies (Akcay & Jungmann, Reference Jungmann2023; Apaydin, Reference Apaydin2025; on the social bloc concept, see Amable & Palombarini, Reference Amable and Palombarini2024). The social bloc concept points to the occasional ideological framing of growth models as being in the “national interest,” downplaying the interests of marginalized groups and the inherent contradictions of a particular model (May et al., Reference May, Nölke and Schedelik2024). At the same time, growth models are inherently unstable and can be disrupted by external shocks (like financial crises), internal contradictions (like rising inequality), or political shifts, leading to pressures for adjustment or transformation. To summarize, the GM perspective makes crisis and change central to understanding capitalist diversity. Comparative political economists using this lens study how and why countries adopt certain growth models, how they cope with crises, and how institutions, coalitions, and ideas help sustain or shift these models over time (Baccaro et al., Reference Baccaro, Blyth and Pontusson2022).Footnote ⁴

2.3 Conceptual Framework: Peripheral Growth Models in Emerging Capitalist Economies

By looking at the components and drivers of economic growth, its distributional implications and political underpinnings, the GM research program has shifted scholarly attention from the institutional structures of capitalist economies and their supply side effects on growth – the focus of the VoC perspective – back to macroeconomics and aggregate demand (Schwartz & Tranøy, Reference Schwartz and Tranøy2019). The analytical shift from supply to demand-side factors of the economy coincided with an evolving debate about the adequate macroeconomic theoretical foundations of growth model-inspired analyses (Hope & Soskice, Reference Hope and Soskice2016; Baccaro & Pontusson, Reference Baccaro and Pontusson2018; Stockhammer, Reference Stockhammer2022). This discussion has led to substantial engagement of CPE scholars with PKE and its focus on “demand” or “growth regimes” (see Stockhammer, Reference Stockhammer2022: 164–170 for overviews).Footnote ⁵

Although most growth model research to date has focused on advanced capitalist economies, first inroads have been made with regard to emerging economies (Schedelik et al., Reference Schedelik, Nölke, Mertens and May2021; Wood & Schnyder, Reference Wood and Schnyder2021; Akcay et al., Reference Akcay, Hein and Jungmann2022; Stockhammer, Reference Stockhammer2022; Jungmann, Reference Jungmann2023; Baccaro & Hadziabdic, Reference Baccaro and Hadziabdic2024; Campana et al., Reference Campana, Vaz, Hein and Jungmann2024; Bulfone et al., Reference Bulfone, Madariaga and Tassinari2025). Country case studies and paired comparisons have used the framework to analyze stability and change in emerging capitalist economies (e.g., Morgan et al., Reference Morgan, Doering and Gomes2021; Nölke et al., Reference Nölke, May, Mertens and Schedelik2022; Apaydin, Reference Apaydin2025; Güngen & Akcay, Reference Güngen and Akcay2024; Kalanta, Reference Kalanta2024) while others have leveraged the economies of Central and Eastern Europe to explore the political economy of FDI-led growth (Avlijaš et al., Reference Avlijaš, Hassel, Palier, Hassel and Palier2021; Ban & Adascalitei, Reference Ban, Adascalitei, Blyth, Pontusson and Baccaro2022; Bohle & Regan, Reference Bohle and Regan2021). Akcay et al. (Reference Akcay, Hein and Jungmann2022), Campana et al. (Reference Campana, Vaz, Hein and Jungmann2024), and Baccaro and Hadziabdic (Reference Baccaro and Hadziabdic2024) mapped the growth models of several large emerging economies and revealed not only variation between the countries but also a temporal shift in their growth models after the global financial crisis of 2008 and the end of the commodity boom in 2011 onward. Furthermore, recent scholarship has begun to conceptualize the notion of peripheral growth models (Stockhammer, Reference Stockhammer2023; Bulfone et al., Reference Bulfone, Madariaga and Tassinari2025; Schedelik & Nölke, Reference Schedelik and Nölke2025), which refers to the “structural constraints imposed by global economic hierarchies” on growth trajectories in developing and emerging economies (Bulfone et al., Reference Bulfone, Madariaga and Tassinari2025, 296).

These and other studies on the political economy of emerging capitalist economies have suggested several factors that fundamentally enhance our understanding of growth, stagnation, and crisis in these countries. These factors, we argue, need to be more systematically integrated into the empirical GM research program. This Element aims to do so with respect to what we think are the three most important factors that have to be reflected in the study of peripheral growth models in emerging economies:

1. (1) A coherent typology of peripheral growth models needs to be developed. Such a typology must go beyond OECD-centric views and needs to take into account the stronger role for investments in emerging economies, which results from the need to expand less developed infrastructure and productive capacities. In particular, an investment-led model needs to be added to the prevalent juxtaposition of consumption-led and export-led growth models.

2. (2) The international context is important in a different way for most ECEs than for advanced economies. Growth in ECEs depends more strongly on how the country is integrated into global economic hierarchies and on systemic dynamics that come in the form of commodity cycles (which undergird many export-led growth models), a dependency on foreign capital (e.g., via FDI), and through processes of financial subordination magnified by global financial cycles. These international interdependencies create profound vulnerabilities vis-à-vis external shocks, which are typical for peripheral growth models.

3. (3) Growth and stagnation in ECEs depend differently on the embeddedness of economic actors in the political sphere than in advanced economies. While political coalitions and state-business relations are likely to define growth models across all types of political economies, the mechanisms through which politics shape particular growth trajectories in ECEs will usually differ from relatively stable Western systems of liberal democracy.

The selection of these factors is based on the literature on the political economy of development during the last decades. Based on this research, we are suggesting a more comprehensive research program for the GM perspective in emerging capitalist economies that not only studies macro-economic growth contributions in terms of aggregate demand components, but also systematically reflects on their international and macro-political embeddedness.Footnote ⁶ More specifically, we argue that, depending on the drivers of economic growth, a peripheral growth model’s vulnerability vis-à-vis external shocks varies considerably. But as growth models do not emerge automatically, their nature and degree of vulnerability are subject to various political factors.

2.4 Case Selection and Methods

This Element investigates the growth trajectories of emerging capitalist economies, that is, mainly middle-income countries as defined by the World Bank. Our sample comprises thirteen ECEs that belong to the largest thirty-three countries with the biggest proportion of global GDP, according to the IMF (Table 1). This is where a large part of global growth has been attributed to, based on purchasing power parities. Given the GM perspective’s (narrow) focus on analyzing growth trajectories, we use this metric for our case selection instead of GDP per capita. In addition, we also do not aim to account for catch-up processes and the transition from middle- to high-income status. This would require an ordering of countries according to certain income thresholds, usually defined in per capita terms. Finally, the countries in our sample are a very diverse set of economies with different political-economic regimes and historical-institutional legacies, covering all the major world regions. This, we contend, should assist us in developing our theoretical perspective of peripheral growth models as a useful and adaptive framework for the analysis of growth trajectories in emerging economies.

Methodologically, we are using the following approaches for tackling the identification of growth models in ECEs and macro-political factors. For the detection of peripheral growth models, we employ a simple demand-side approach for growth accounting, whereas our study of political factors sustaining these models will be based on country vignettes. Vignettes, which have found their way from social psychology into other social sciences, are commonly understood to present case histories for the purpose of illustrating important traits of a situation or constellation. Importantly, they lend themselves particularly well to the development of a research program, in comparison to other traditions of case study research that require thick descriptions (Hughes, Reference Hughes1998, 381; Welch et al., Reference Welch, Piekkari, Plakoyiannaki and Paavilainen-Mäntymäki2011, 751–2).

Our vignettes are short case studies on the political factors supporting specific growth models. They do not claim to provide comprehensive evidence, but rather to validate a conceptual point. Correspondingly, the selection of country cases is based on their representative character for the conceptual argument at hand. Our empirical vignettes draw primarily on paired comparisons of “most similar” countries: Brazil and Indonesia, Turkey and South Africa, as well as Thailand and Vietnam. These pairs are structured similarly: They follow the same overall growth model but differ in their institutional and political approaches toward addressing the resulting structural vulnerabilities. While three of these countries were able to uphold growth overall in the time period under scrutiny (Indonesia, Turkey, and Vietnam), the other three experienced a substantial decline in their growth rates and stagnation or outright economic crisis (Brazil, South Africa, and Thailand; see Table 3). A notable finding in Section 5 relates to our argument that growth models depend on certain political-institutional constellations. We show that the less vulnerable economies in our sample, Indonesia, Turkey, and Vietnam, have associations or similar (informal) institutional arrangements that are a crucial element in the accommodation between various social groups, as well as between the state and business. These play little or no role in the more vulnerable cases of Brazil, South Africa, and Thailand. This shows that, at least within the limited time period studied in this Element, the political stabilization of a growth model rests on broad societal support and functional mechanisms for state-business relations.

3.1 Adapting Growth Model Typologies to Emerging Capitalist Economies

Being aware of the fallacies of typological theorizing (Hay, Reference Hay2020), we acknowledge the limits of this framework right from the start. As ideal types are “limiting concept[s] with which the real situation or action is compared” (Weber, Reference Weber and Finch1947 [1922], 93, original emphasis), our typology is primarily meant to be a heuristic device for comparative analysis (Swedberg, Reference Swedberg2018, 189–90). It aims to reduce the complexity of contemporary capitalist development in emerging economies and to provide the conceptual basis for hypothesis formulation and empirical research. Of course, it does not claim to cover all growth experiences in the global South. Nor should the aim be to group all countries into ideal-typical categories. In reality, many cases display features of several of the ideal types mentioned below, thereby creating tensions with the temporal and spatial specificity of growth trajectories. Needless to say, not every country exhibits a well-defined growth model at any point in time, as there is ample evidence of crisis and stagnation in peripheral economies.

The identification of growth models starts with a basic distinction between their main components of aggregate demand: private consumption, government consumption, investment, and (net) exports. Next to these broad types of growth models, we introduce several subtypes, which, on the one hand, capture the main sources of growth experiences of emerging economies identified in the literature and, on the other hand, differ with regard to their degree of economic vulnerability (Table 2).

Table 2Growth models in emerging capitalist economies

Source: Own elaboration.

Table 2Long description

The table further differentiates between several subtypes of each model, based on the respective growth driver: manufacturing-based export-led growth, commodities-based export-led growth, wage-based consumption-led growth, debt-based consumption-led growth, domestically-based investment-led growth, and F D I-based investment-led growth.

Due to their less developed productive structure and lower capital stock, investments play a much more important role in emerging economies than in advanced economies. The recent experience of India or China exemplifies this point (Ahuja & Nabar, Reference Ahuja and Nabar2012; ten Brink, Reference ten Brink2019).Footnote ⁷ Therefore, we propose a third type of growth model – investment-led growth – which has so far not been identified in the core countries (see, e.g., Picot, Reference Picot, Hassel and Palier2021; Baccaro & Hadziabdic, Reference Baccaro and Hadziabdic2024). Moreover, due to their position in the world economy, emerging economies are generally characterized by higher vulnerability – even though their vulnerability may vary with their economic profile and regime type. This elevates the possible negative impacts of variables linked to their international interdependencies (Bonizzi et al., Reference Bonizzi, Kaltenbrunner, Powell, Mader, Mertens and van der Zwan2020), such as massive capital outflows as witnessed during the 2008 global financial crisis and the beginning of the COVID-19 pandemic. In order to account for this salient feature of growth in peripheral economies, we introduce the separate dimension of vulnerability, ranging from low to high, depending on the respective growth driver. The latter are factors, which are “not themselves part of aggregate income but influence the growth of its components” (Kohler & Stockhammer, Reference Stockhammer2022, 1319).

Against this backdrop, we distinguish the following peripheral growth model types and their respective subtypes:

1. (i) Export-led growth models: These growth models are defined as sustained periods of growth that are primarily driven by exports. Here, it matters substantially which sectors of the economy are the main exporters. Economic complexity, value added, and the structure of global markets vary widely and have important repercussions on growth trajectories. Arguably, the main export sectors in emerging economies are either manufacturing or commodities. Services such as tourism are a third “export” sector that accounts for a sizable share of GDP growth, albeit mainly for smaller developing countries such as Bhutan and Jamaica.

   1. (1) Manufacturing-based export-led growth refers to episodes when the main export sectors are in manufacturing. Countries can thrive based on higher price or non-price competitiveness, depending on the degree of price elasticity of demand of different manufacturing sectors and the structure of global markets. But typically, with poor technological capabilities, prices have a binding constraint on the manufacturing-based growth models of most low- and middle-income countries. Consequently, there is a tendency to repress real wage growth and, as a result, domestic consumption in economies with a large export sector is generally lower. This has negative repercussions on domestic welfare and inequality in the short term. However, in a mid- to long-term perspective, a manufacturing-based export-led growth model potentially leads to more sustainable and inclusive development, especially if a country manages to transition from price to non-price competitiveness (product differentiation, quality, and innovation), as exemplified by Korea.Footnote ⁸

   2. (2) Commodities-based export-led growth refers to episodes when the main export sectors are natural goods. This holds true for the vast majority of peripheral economies. Lacking short-term alternative options, many countries, by necessity, adopt such a development strategy. However, this type of growth model has potentially destabilizing effects. First, commodity prices are highly volatile and often lead to pronounced boom-bust scenarios. Second, rising commodity prices generally trigger capital inflows from abroad, resulting in a loss of competitiveness for manufacturing and thereby gradually leading to deindustrialization (the Dutch Disease phenomenon). Further negative effects relate to political instability and armed conflicts arising from distributional conflicts over the exploitation of commodity rents (the so-called resource curse).

2. (ii) Consumption-led growth models: These growth models are defined as sustained periods of growth primarily driven by private consumption. Household consumption can be financed by two main sources: wages or debt. Hence, we further distinguish between two main subtypes here, wage-based growth and debt-based growth.

   1. (1) Wage-based consumption-led growth refers to episodes when sustained real wage growth leads to higher incomes, which in turn fuel household consumption expenditure. Increased demand may then spur further investment by companies, resulting in a virtuous cycle of capital accumulation. Politically, such a growth model can be induced by several policies, such as minimum wage increases, public sector pay increases, or the strengthening of collective bargaining institutions and trade unions. We witnessed such a growth model in many countries in the two decades after World War II. However, the model has an inherent tendency to fuel inflation through continuous wage pushes and to squeeze the profits of capital (Baccaro et al., Reference Baccaro, Blyth and Pontusson2022, 506).

   2. (2) Debt-based consumption-led growth refers to episodes when credit expansion leads to higher disposable incomes of households, fueling consumption expenditure. This is particularly relevant for capital goods industries and real estate. One of the main drivers of debt-based models is capital inflows from abroad, which can drive down long-term interest rates and push up local asset prices. This may contribute to a credit boom, spurring domestic economic activity. However, relying too heavily on the inflow of foreign capital in turn creates vulnerability, as the danger of sudden stops and boom-bust cycles looms large.Footnote ⁹

3. (iii) Investment-led growth models: These growth models are defined by sustained periods of growth that are primarily driven by investment. Investment is based on two main sources: domestic or foreign capital.

   1. (1) Domestically-based investment-led growth refers to episodes when increased investments in infrastructure, housing, plants, or machinery fuel growth. Fiscal policy is arguably one of the main drivers of such a growth model, albeit there are numerous other possible drivers of (private) investment (such as new technologies, for instance). Beyond boosting demand in the short run, investments build long-term economic capacity, enhance productivity, and create additional employment. If investment predominately takes place in (residential) housing, there may be overlaps with the debt-based growth model discussed earlier. Instability in this model may result mainly from overproduction and overindebtedness of the business sector.Footnote ¹⁰

   2. (2) FDI-based investment-led growth refers to episodes when investments by multinational corporations (MNCs) in local production plants drive the growth model. This model, by definition, heavily relies on a continuous inflow of FDI. Interrupting these flows, for example, through the relocation of production plants by MNCs, represents the main source of vulnerability for this type of growth model. In a mid-term perspective, this translates into a structural challenge for an FDI-led development strategy, as MNCs often are not willing to share advanced technology with host countries (Gertler, Reference Gertler, Peck and Yeung2003; Liu, Reference Liu2008). Consequently, the latter eventually might be stuck in a so-called middle-income trap, which refers to a loss of competitiveness in labor-intensive manufacturing, while still not having reached the capacities for advanced production and R&D related activities.

We restrict ourselves to these types of growth models, which cover the main options for pursuing growth in emerging economies. Other models, which are, for example, based on remittances, foreign aid, or tourism, are mainly relevant for smaller developing countries with specific economic endowments and are hardly comparable to other economies.

3.2 Identifying Peripheral Growth Models

To develop an empirical approximation to growth models in ECEs, we use the relative contributions of aggregate demand components to GDP growth. Following Hein et al. (Reference Hein, Paternesi Meloni and Tridico2021), we calculate the relative contributions to GDP growth by dividing the change in one aggregate demand component (e.g., C for private consumption) by the change in GDP (Y: dC/dY). We are aware of the limitations of this method, which is less suited to specify the role of exports, the vulnerability of growth models, and their main drivers (Kohler & Stockhammer, Reference Stockhammer2022, 1318–19; Baccaro & Hadziabdic, Reference Baccaro and Hadziabdic2024, 1330–31). Traditionally relying on net exports as one component of aggregate demand, this method tends to either over- or underemphasize the actual role of exports for a country’s growth model. Many emerging economies have a less sophisticated productive structure and thus have high volumes of consumption or capital goods imports. In addition, their major export items are commodities, that is, raw materials and agricultural goods. Due to the high share of imports that are consumed by the private and/or the public sector of these economies, their massive gross exports often do not show up in growth contributions, where net exports are generally low or even negative. Therefore, we focus on gross instead of net exports.Footnote ¹¹ In order to identify the vulnerabilities and drivers of each growth model, we employ qualitative case vignettes in the subsequent sections of this Element.

The period of observation depends on the availability of harmonized data (mostly 2000–2022), which we further split into two subperiods (2001–2011 and 2012–2022). The rationale for this division relates to the main growth drivers of peripheral growth models: the years 2011 and 2012 mark the end of the commodity super cycle, a peak of capital flows to ECEs, and a plateauing of FDI inflows to ECEs (Miranda-Agrippino & Rey, Reference Miranda-Agrippino, Rey, Gopinath, Helpman and Rogoff2022; Jungmann, Reference Jungmann2023; see Section 4). Furthermore, prior studies have identified growth model shifts in several ECEs around the year 2011 as well (Campana et al., Reference Campana, Vaz, Hein and Jungmann2024).

Table 3 displays the results for the fourteen selected ECEs. While ECE annual growth rates are, on average, much higher than in the United States or European countries in that period (2.15 percent for the US and 1.56 percent for the EU respectively, see World Bank, 2025), data for some ECEs, notably Argentina, Brazil, Russia, and South Africa, also show a steep decline from the 2000s to the 2010s. These countries seem to follow those vulnerable trajectories of boom-bust growth as laid out earlier. Thailand is another case that experiences a substantial drop in its GDP growth between the two periods, although less pronounced than the other four countries. The rest of the countries in our sample, notably China, Indonesia, and Vietnam, also show relatively declining growth rates overall, but appear to have been more successful in sustaining growth. Egypt and Turkey are the only countries in the sample whose growth rates increase in the later period.

Table 3Relative contributions to GDP growth in emerging capitalist economies, 2001–2022

Note: We have calculated the compound annual growth rate (CAGR) for each aggregate demand component for each period. CAGR represents the annual growth rate required to move from the starting GDP to the final GDP over a specified period, assuming constant growth at this rate. By smoothing out yearly fluctuations, it eliminates the effects of volatility and is well-suited for long-term comparisons and trend analysis. For private consumption, we use household consumption expenditure; for investments, we use gross capital formation; for government consumption, we use general government final consumption expenditure; for net exports, we calculate separately the growth rates of exports and imports of goods and services. The absolute contribution for each aggregate demand component was obtained by multiplying their CAGR by their average weight shares in each period. Finally, we normalized them to obtain their relative contributions. The sum of private and government consumption, investment, and net exports equals to one. As the table displays exports rather than net exports, the sum will typically exceed unity.

Source: United Nations National Accounts, constant 2015 prices, GDP expenditure approach, extracted on January 12, 2025; own calculations.

Table 3Long description

The table shows the relative contributions of each aggregate demand component to G D P growth for Argentina, Brazil, China, Egypt, India, Indonesia, Mexico, Nigeria, Russia, South Africa, Thailand, Turkey, and Vietnam in the period from 2001 until 2022.

According to this data, there is not one single growth model for ECEs (Table 3). In general, however, private consumption plays an important role in many countries in our sample. Only China and Nigeria exhibit a slightly lower but still significant contribution of private consumption to GDP growth. South Africa appears to align with the consumption-led model, although government consumption plays an important role as well. The same holds true for Turkey, albeit with less pronounced government consumption. As we will demonstrate in the next section, household indebtedness has been driving this trend in both economies. While credit expansion in South Africa was especially salient in the period before the global financial crisis, it became more important in the subsequent period in Turkey. In both cases, capital inflows have been drivers of debt-based growth (Section 4). Wage-based growth models have been rare since the 1980s – with Israel between 2008 and 2018 as an exception (Bondy & Maggor, Reference Bondy and Maggor2024). Consequently, we do not find it in our sample.

Overall, government consumption is less significant than the other demand components in our sample countries (Figure 3). Only Argentina, and to some extent Mexico and Nigeria, exhibit a relatively high degree of this component. In the former, this is mainly due to a sustained period of expansionary fiscal and social policies under the Kirchner governments (2003–2015) (Ianni, Reference Ianni2024), with public administration as well as educational and health services being major drivers of GDP growth (Baccaro & Hadziabdic, Reference Baccaro and Hadziabdic2025).

Figure 3 Relative contributions to growth by demand component, 2001–2022

Source: see Table 3.

Investment, by contrast, seems to play a major role in many countries in our sample. Notably, China and India display a strong investment-led profile similar to that of Turkey and South Africa, though this tendency is declining sharply in the latter two cases. In all the cases, the construction sector plays an important role, contributing up to 34 percent to overall GDP growth in China in the 2010s, for instance (Baccaro & Hadziabdic, Reference Baccaro and Hadziabdic2025). Thailand and Vietnam also feature a significant share of investment in their GDP growth, but mainly in the manufacturing sector, dominated by foreign MNCs. However, this share is declining substantially in Thailand, while Vietnam seems to uphold it in the period 2012–2022 as well. As we will show in the next section, investment in both countries is heavily driven by FDI (Section 4).

Finally, exports are very volatile but still significant in countries such as Brazil, Indonesia, Nigeria, Thailand, and Vietnam. While Thailand and Vietnam are highly integrated into global production networks, there is no country in the sample that displays a proper manufacturing-based export-led growth model, relying on indigenous manufacturing capacities. Given their heavy GVC participation, both Thailand and Vietnam only feature very high gross exports, while their net export contributions to growth are negligible or even negative. There is a striking similarity to the FDI-led models of East Central Europe, which also exhibit high volumes of gross exports, while foreign investment is the main driver of GDP growth (Avlijaš et al., Reference Avlijaš, Hassel, Palier, Hassel and Palier2021; Bohle & Regan, Reference Bohle and Regan2021). Brazil and Indonesia, by contrast, are part of a larger group of peripheral economies featuring commodities-based export-led growth, where commodity revenues at least temporarily serve as a basis for redistributive policies supporting domestic consumption (Nölke et al., Reference Nölke, ten Brink, May and Claar2020, 137–9, Passos & Morlin, Reference Passos and Morlin2022; Schedelik et al., Reference Schedelik, Nölke, May and Gomes2023). As we demonstrate in the subsequent section, both countries managed the vulnerabilities associated with this type of growth model very differently.

In sum, investment and private consumption are the key contributors to growth in the countries in our sample, while government consumption and exports are less significant overall. This questions the popular diagnosis of emerging economy growth as being mainly export-led, as discussed in the experience of East Asian “tiger states” in the 1980s (World Bank, 1993) or the rise of the BRICS countries, especially China, in the 2000s (Feenstra & Wei, Reference Feenstra and Wei2010). Further juxtaposing peripheral growth models with the consumption- and export-led models in advanced capitalist economies, we find investments to be heavily important for some ECEs, supporting our intuition for the addition of an investment-led growth model to the typology of an extended GM perspective. For a theory of capitalist diversity this should not be a surprising point to make since investments should be at the core of any theory about growth – and in fact has been for political economy scholarship of the so-called “golden age of capitalism” (Marglin & Schor, Reference Haggard, Kim and Moon1991; Schwartz & Tranøy, Reference Schwartz and Tranøy2019) and earlier development theories, such as those associated with the model of import-substitution industrialization in the global South (Irwin, Reference Irwin2021). It stands that only investments, besides representing an increase in aggregate demand, also expand productive capacity. Consumption and export growth boost demand and thus the utilization of the capital stock, yet only investments increase the capital stock. This point is particularly salient for ECEs since they usually have a lower capital stock than advanced economies and need to invest in order to catch up. At the same time, since higher investments yield additional capacity, there is also a need to find demand for this increased productive capacity.

Systematically introducing the possibility of an investment-led growth model helps us to understand the various political-economic trajectories of high-growth economies. We can also learn from studies on former transition economies in Central and Eastern Europe, as well as Ireland, which pointed to an FDI-based model (e.g., Bohle & Regan, Reference Bohle and Regan2021). For instance, the macro-political and institutional factors that shore up and constantly challenge such an investment-led growth model vary with the latter’s integration into the world economy and the type of investments within that model. For this reason, we later turn to a more qualitative vignette approach that is able to carve out potential causal effects supporting and sustaining an investment-led growth model by studying the growth trajectories of Thailand and Vietnam.

After all, and it is worth repeating, none of these models is dominant for ECEs as a group, and all types require differentiation. Table 4 proposes such a differentiation with a view toward exemplary cases. However, in contrast to the clear-cut growth models which could be found in Europe (only) before the global financial crisis (Baccaro & Pontusson, Reference Baccaro and Pontusson2016; Hein et al., Reference Hein, Paternesi Meloni and Tridico2021), we should be careful with the neat categorization of countries into singular types. We find a much higher degree of hybridity for the growth models in ECEs. Moreover, traditional growth model decomposition may lead to misleading impressions, given the high export shares for FDI-based models and the (partial) utilization of commodity export revenues for the stimulation of domestic consumption in commodity-led models discussed earlier. Correspondingly, our simplified classification should be taken as an invitation for further research on how to understand the nature of peripheral growth models.

Table 4Types of growth models and country cases

Note: ** Not found in our sample.

Source: Own elaboration.

Table 4Long description

The table identifies country cases for several growth model types, for instance Brazil and Indonesia as commodities-based export-led growth, Turkey and South Africa as debt-based consumption-led growth, Vietnam and Thailand as F D I-based investment-led growth, as well as China as domestically-based investment-led growth.

Moreover, we posit that the considerable heterogeneity within the category of ECEs demands an institutionally sensitive process when classifying growth models. In order to substantiate this claim, we move to the level of international interdependencies of peripheral growth models (Section 4) and their political underpinnings (Section 5). Through the study of commodities-based export-led growth (Brazil and Indonesia), debt-based consumption-led growth (South Africa and Turkey), and FDI-based investment-led growth (Thailand and Vietnam), we show how the international insertion of a country’s growth model shapes its vulnerabilities toward exogenous shocks. Still, each paired comparison demonstrates that despite structural dependencies, governments in ECEs can successfully mitigate the vulnerabilities arising from similar peripheral growth models. This capacity does not only rest on appropriate macroeconomic policies, but also on certain macro-political conditions. Subsequently, we illustrate the importance of these various factors, which represent constellations neglected in the study of growth models in the global North.

4.1 Commodity Price Cycles and Export-Led Growth Models

Recently, GM scholarship has begun to study export-led growth models, especially the commodity-based subtype. Arguably, this is one of the most prevalent peripheral growth models globally (Schedelik et al., Reference Schedelik, Nölke, May and Gomes2023). This model heavily relies on the development of commodity prices, which are set on global markets with little or no pricing power for producer countries. In addition, commodity prices are characterized by large swings, dubbed commodity super cycles, where prices rise sharply for several years before plummeting (Erten & Ocampo, Reference Erten and Ocampo2013). Commodity price fluctuations overproportionally affect developing and emerging economies, which, on average, are less diversified and more commodity-dependent than advanced economies. These price cycles not only have a direct impact on export volumes and earnings, but also on macroeconomic stability, as economic activity in commodity-dependent ECEs closely mirrors the price movements of their major export items (IMF, 2012, 125). Here, we can draw on decades of development studies research, which has pointed toward the significant macroeconomic effects of an overreliance on natural resource extraction (see the literature on the “resource curse,” cf. Ross, Reference Ross2015).

The most important mechanism associated with fluctuations in commodity prices is the so-called “Dutch Disease” – termed after the economic side-effects of natural gas discoveries in the Netherlands in the 1950s (Corden, Reference Corden1984). It refers to the real appreciation of the currency due to huge capital inflows into investment projects in the booming commodity sectors and/or large increases in commodity revenues during boom periods (Frankel, Reference Frankel2010, 20). This results in increased domestic income and spending by the private and especially public sector, leading to higher prices and output in the nontradables sectors and consequently to higher wages across the economy (the so-called “spending effect”). At the same time, capital and labor move from other parts of the economy to the booming commodity sectors, which results in rising prices of nontradables vis-à-vis other tradables (the so-called “resource movement effect”). Both effects lead to a real exchange rate appreciation and, therefore, to a loss of international competitiveness and declining output in the manufacturing sectors and, hence, to deindustrialization (Brahmbhatt et al., Reference Brahmbhatt, Canuto and Vostroknutova2010). The predicament of the “Dutch Disease” is much less likely to occur in advanced economies that have developed a strong reliance on exports (such as Germany). These economies never fully tilt toward structural dependency – not least because the market structures for manufactured goods and commodities are fundamentally different: “what you export matters” (Hausmann et al., Reference Hausmann, Hwang and Rodrik2007). More specifically, avenues for diversification and technological upgrading are limited for most commodity-dependent developing and emerging economies, and depend on sustained political efforts, which are negatively affected by commodity price swings (UNCTAD, 2021).

The reliance on commodity exports is a recurring and prevalent feature of most developing and emerging economies. In the 2018–2019 period, 101 countries out of 195 UNCTAD member states, that is, 53 percent, were commodity-dependent, that is, with a commodity export product share of more than 60 percent (UNCTAD, 2021, 5). Beyond that, we can identify a correlation between the degree of commodity dependence, that is, commodity exports as a share of merchandise exports, and the contribution of commodity exports to a country’s gross domestic product (Figure 4). A higher share of commodity export dependence is associated with a higher share of commodity exports in a country’s GDP. Commodity dependence, thereby can be as high as 99.8 percent in the case of Iraq and even 100 percent in the case of South Sudan. Furthermore, several countries rely on only one or a few commodities. Countries such as Iraq (93.5 percent), Angola (88 percent), Chad (79.6 percent), and Guinea-Bissau (88.4 percent) depend heavily on the export of one single product (UNCTAD, 2019).

Figure 4 Commodity dependence of developing and emerging economies

Source: Own elaboration based on UNCTAD 2021

In the recent past, China’s rapid industrialization and the related output growth in the 2000s generated immense demand for primary products. From 1997 to 2017, China’s share of global energy consumption more than doubled, from 11 to 23 percent, while its share of global metals consumption even increased fivefold, from 10 to 50 percent (World Bank, 2018, 11). In comparison, advanced economies’ share of global energy and metals consumption declined from 50 to 40 percent and 70 to 30 percent, respectively, in the same period. This demand shock fueled a sustained commodity price boom, which lasted until the early 2010s. The commodity super cycle came to a halt during the GFC and particularly after 2011, when commodity prices contracted sharply (see Figure 5).

Figure 5 Commodity price cycles

Source: Own elaboration based on IMF, Primary Commodity Price System

We can see the effects of these fluctuations on growth models by comparing Brazil and Indonesia, which are both commodity exporters and of a similar size. In both countries, the change in the growth trajectory at the end of the 2000s is striking, also in comparison with other major emerging economies (Section 3). Whereas exports were growth drivers for the two countries in the early 2000s, they were both hit hard by the strong reduction in export revenues after the end of the commodity boom that had been driven by China. In other words, the investment-led growth model in China created the demand for resources provided by extractivist suppliers with their commodities-based export-led growth model, only to increase pressure for adjustment when prices fell.

The growth models of the two economies, however, reacted differently to the end of the commodity cycle. Growth in Brazil collapsed, whereas it remained stable in Indonesia (Table 3). While Indonesia was able to retain a high contribution of investments alongside private consumption when exports dwindled, Brazil was unable to stimulate along similar lines. Furthermore, private consumption – together with government consumption – had to carry the whole burden of growth stimulation in Brazil in the medium-term, in line with the notion of a wage-based recovery but increasingly also rising consumer debt. In other words, while Brazil shortly moved from a commodity-based export-led model to a consumption-led model, Indonesia came to a more sustainable model based on investments and private consumption. Whereas Indonesia in this way managed to rebalance its growth model and diversify its export profile away from raw materials, Brazil’s heavy reliance on commodity exports even intensified and peaked at almost 60 percent of total exports in 2021 (Figure 6).

Figure 6 Export product share by stage of processing of Indonesia (a) and Brazil (b)

Source: Own elaboration based on WTO, World Integrated Trade Solution database

Openness toward the global economy and pressures from financial subordination seem to be core factors here. Brazil is much more open toward foreign finance and trade, whereas Indonesia has followed a more protectionist path since the global financial crisis. Beginning during the second term of President Susilo Bambang Yudhoyono (2009–2014), Indonesia turned away from its previously rather orthodox neoliberal model (Jepson, Reference Jepson2020). In this case, protectionism refers both to financial flows – a legacy of the ill-fated opening of the financial sector after the 1998 Asian financial crisis (Pepinsky, Reference Pepinsky2013) – and trade, where the risk of deindustrialization via open borders for Chinese imports has been taken much more seriously than in Brazil. This means, in turn, that Brazil was not able to minimize its structural dependency on foreign demand and finance in the same way as Indonesia.

The governments led by the Brazilian Workers’ Party (PT) embarked on a program for stimulating domestic demand via private consumption when the PT took power in 2003. Measures to alleviate poverty, raising the minimum wage, and the promotion of credit access for private households (partly through social transfers) were successful in stimulating domestic demand and reducing inequality (Passos & Morlin, Reference Passos and Morlin2022; Schedelik, Reference Schedelik2023, 64–70). Yet, a large part of this additional demand was met by foreign companies, most notably from China. Furthermore, massive capital inflows led to a substantial overvaluation of the Real, proving a juggernaut for the domestic manufacturing sector and contributing to a sustained process of deindustrialization (Schedelik, Reference Schedelik2023, 70–80). When the Brazilian economy recovered briefly after the GFC due to massive public investments in infrastructure and the expansion of subsidized credit, financial inflows surged to record levels (Gallagher & Prates, Reference Gallagher, Prates and Schneider2016). In this context, Brazil temporarily implemented capital controls and interventions in the foreign exchange market, but powerful transnationally oriented financial capital prevented the introduction of really effective controls (Dierckx, Reference Dierckx2015, 154–5).

The broader macroeconomic trends, mainly rising Chinese demand for commodities, affected investment flows, which mostly targeted the booming commodity sectors rather than manufacturing, counteracting several industrial policies of the time (Mesquita Moreira et al., Reference Mesquita Moreira, Rodríguez Chatruc, Lage de Sousa and Merchán2020, 2). These industrial policies included huge tax exemptions for the domestic industry under the so-called “Greater Brazil Plan” launched in 2011, which led to a slowdown in budgetary revenue growth. Together with rising spending on pensions, social benefits, and an array of subsidies to uphold domestic consumption, this contributed to a rapidly deteriorating fiscal situation (Holland, Reference Holland2019; Ricz & Schedelik, Reference Schedelik2023). When tax revenues shrank with the end of the commodity super cycle (and it became clear that expectations about the oil rents to be extracted from new hydrocarbon discoveries were overconfident), further stimulation of private domestic demand became politically highly contested, and the consumption-led growth model turned sour (Nölke et al., Reference Nölke, May, Mertens and Schedelik2022; Passos & Morlin, Reference Passos and Morlin2022).

Indonesia, in turn, has been more restrictive with regard to imports and financial flows, also as a reaction to previous destabilization: “Since the 1997–98 Asian financial crisis, Indonesian economic policy has consistently prioritized stability over riskier pathways to economic growth” (Rajah, Reference Rajah2018, 2). President Joko Widodo (2014–2024) further strengthened the move toward neo-developmentalism, which began during the second term of his predecessor, Susilo Bambang Yudhoyono (Warburton, Reference Warburton2016). Correspondingly, the government compensated for the end of the commodity boom with massive public investment, predominantly in construction and infrastructure projects, but also in “downstreaming,” where raw materials are processed domestically rather than exported (Camba et al., Reference Camba, Lim and Kevin2022). For example, in the electric vehicles value chain, Indonesia sought to restrict the export of unprocessed mineral ores in order to encourage the localization of a domestic mineral beneficiation industry. This is a strategy that can be seen as a reaction to China’s increasing imports of cheap mineral ores (such as nickel) to produce electric vehicle batteries (Schröder & Iwasaki, Reference Schröder and Iwasaki2023). Similarly, the government has put into place a number of measures to limit imports, particularly of consumer goods. Together with expanding government social programs, the jobs created by these activities have contributed to the strong growth of domestic consumption (OECD, 2018a, 131–3). Similar observations can be made with regard to financial flows. Again, the Indonesian government has consciously put a premium on stability. Not only is the economy less integrated into GVCs than other countries in the region (OECD, 2018a, 133), but the Indonesian government also puts a strong emphasis on limited vulnerability with regard to global financial markets. This includes capping external borrowing as well as the current account deficit, combined with the accumulation of considerable foreign exchange reserves (Rajah, Reference Rajah2018, 3). The restrictive Indonesian strategy was supported by its political system with a high concentration of power in the presidency, whereas the more fragmented Brazilian system allowed for more influential lobbying by the financial sector (Nölke et al., Reference Nölke, ten Brink, May and Claar2020, 120–2).

To sum up, both Brazil and Indonesia followed export-led growth paths during the commodity boom period, partly driven by demand from China. When the super cycle came to an end, Indonesia was less pressured by capital outflows than Brazil, enabling its government to expand fiscally and stimulate investment and consumption to boost growth. Brazil, in contrast, had – also fiscally – bolstered the expansion of domestic markets and consumption, but with rising private indebtedness and strengthened political pressure from a coalition between financial capital, the traditional middle class, and increasingly from industrialists (Marques, Reference Marques2025), ran into crisis and stagnation. By pursuing a more protectionist approach, Indonesia has been able to rebalance its growth model toward investment and consumption and diversify its export structure away from raw materials. Brazil, by contrast, has even deepened its heavy reliance on commodity exports in the recent past. Needless to say, smaller developing and emerging countries, that is the vast majority, are even more prone to the pitfalls of commodities-based growth. Still, the case of Indonesia demonstrates that mid-sized emerging economies are not necessarily helpless vis-à-vis global commodity price cycles.

4.2 Financial Cycles and Consumption-Led Growth Models

Further important international interdependencies for developing and emerging economies stem from their integration into the global financial system. It has been shown that the global financial crisis has been a watershed for the trajectory of growth models in advanced economies (Hein et al., Reference Hein, Paternesi Meloni and Tridico2021; Kohler & Stockhammer, Reference Stockhammer2022). Arguably, the GFC played a similar role in some ECEs, where it fostered a change of growth models (Akcay et al., Reference Akcay, Hein and Jungmann2022). More specifically, global financial markets shape political economies in the global South by means of subordinated financialization, potentially destabilizing domestic growth models via large-scale speculative financial flows and the dynamics of global financial cycles under US dollar hegemony (Bonizzi et al., Reference Bonizzi, Kaltenbrunner and Powell2022). Financial subordination thereby refers to many developing and emerging economies’ strong reliance on external borrowing and the resulting vulnerability to sudden capital flows and exchange rate swings. Due to the still prominent role of foreign currency-denominated debt (famously called the “original sin,” cf. Eichengreen & Hausmann, Reference Eichengreen and Hausmann2005) and foreign investors in local currency bond markets (“original sin redux,” cf. Carstens & Shin, Reference Carstens and Shin2019), a large portion of these countries’ debt induces currency mismatches in the event of exchange rate depreciation which increases the risk of local financial crises. Positive inflows can drive down long-term interest rates and push up local asset prices. This may assist domestic economic activity, contributing to a credit boom. Large negative portfolio flows, however, very often trigger financial crises with severe domestic macroeconomic consequences (Kohler, Reference Kohler2022).

Since the early 2000s, capital flows to developing and emerging economies have risen markedly, as international investors increasingly discovered “emerging market” assets as a promising investment (Figure 7). In particular, portfolio flows have recently reached unprecedented levels, outpacing FDI (Molina & Viani, Reference Molina and Viani2019). A recent empirical literature, therefore, has investigated the drivers of these flows (see Koepke, Reference Koepke2019 for an overview). Global risk aversion and a strong US dollar are negatively correlated with capital inflows to developing and emerging economies, whereas output growth in the recipient countries is positively correlated (Molina & Viani, Reference Molina and Viani2019, 4). In addition, commodity prices are another important determinant of capital inflows for commodity-dependent countries, even since the nineteenth century (Reinhart et al., Reference Reinhart, Reinhart and Trebesch2016). Although many ECEs have sought to protect their economies against financial volatilities by imposing capital controls and accumulating foreign reserves, they still tend to be affected significantly (Bauerle Danzmann et al., Reference Bauerle Danzmann, Wincoff and Oatley2017). As capital flows toward developing and emerging economies are predominately pro-cyclical, they tend to exaggerate boom-bust cycles and seriously impinge on their domestic macroeconomic stability (Akyüz, Reference Akyüz2022).

Figure 7 Portfolio capital inflows to developing and emerging economies

Source: Own elaboration based on IMF, Balance of Payments database; World Bank, International Debt Statistics

We can observe the impact of global capital cycles on peripheral growth models by comparing South Africa and Turkey, which are both relatively open toward international financial markets and of a similar size. Both countries experienced debt-based growth trajectories recently, culminating in major crises. Whereas Turkey’s growth model can be classified as regular domestic demand-led before the global financial crisis, it shifted toward a debt-based model in the post-GFC period (Akcay et al., Reference Akcay, Hein and Jungmann2022). South Africa, by contrast, sustained a debt-based growth model throughout the 2000s and 2010s, despite a major recession induced by the GFC (Akcay et al., Reference Akcay, Hein and Jungmann2022). Both countries exhibit different types of debt booms, however. Whereas South Africa experienced a classical housing bubble, similar to countries such as Ireland or Spain in the European (semi-)periphery or the UK and the US before the GFC (Karwowski, Reference Karwowski2018, 418), Turkey’s growth model of the 2010s was based on a corporate debt bubble, which resulted in a classical currency-and-debt crisis (Akcay & Güngen, Reference Akcay and Güngen2022). In both cases, large inflows of foreign capital played a crucial role in fueling the debt booms (Figure 8).

Figure 8 Private debt as a percentage of GDP in South Africa and Turkey

Source: Own elaboration based on IMF, Global Debt database

Both South Africa and Turkey have been in a similar situation with regard to the potentially devastating effects of global capital cycles. Still, Turkey has managed the vagaries of these cycles better than South Africa, not only with regard to the attraction of capital inflows, but also in terms of GDP growth (Ay et al., Reference Ay, Akar and Akar2016; Figure 9). After the GFC in particular, Turkey’s GDP growth far exceeds that of South Africa, which entered into a phase of economic stagnation, characterized by high levels of unemployment, low growth, and extreme inequality (Nölke et al., Reference Nölke, ten Brink, May and Claar2020, 145). Arguably, we can explain this divergence by the Turkish turn toward a stronger focus on industrial development (Kutlay, Reference Kutlay2020, 692–8) and several “unconventional” economic and monetary policies after the crisis (Zayim, Reference Zayim2022).

Figure 9 Annual growth of gross domestic product (right axis) and consumer price inflation (left axis) in Turkey and South Africa compared

Source: Own elaboration based on World Development Indicators

While some emerging economies reacted to the massive global liquidity during the quantitative easing (QE) phase in Europe and the United States with capital controls, with varying success (Section 4.1), the two debt-led countries analyzed here pursued different paths. South Africa, for its part, stuck largely to the neoliberal orthodoxy and even relaxed capital outflow restrictions in the wake of the crisis (Zayim, Reference Zayim2022, 544). Turkey, by contrast, implemented a series of unconventional policies such as re-regulating reserve requirements and adopting an asymmetric interest rate corridor (Apaydin & Çoban, Reference Apaydin and Çoban2023, 1056). Although the Turkish central bank modified the policy after the US Fed’s taper tantrum in 2013, the Turkish economy navigated the turbulent post-GFC period better than many of its emerging market peers (Ay et al., Reference Ay, Akar and Akar2016). As an example, the Turkish currency experienced far less volatility than other ECE currencies, and domestic credit growth was slowed (Kara, Reference Kara2012, 18–9). Subsequently, the government implemented several industrial policies to shift the growth model toward more manufacturing exports, including neomercantilist policies for accessing new non-Western markets (Akcay & Jungmann, Reference Jungmann2023, 547–9; Arnold & Naseemullah, Reference Arnold and Adnan2024; Güngen & Akcay, Reference Güngen and Akcay2024, 162–3). However, state interventionism had serious political consequences as well, consolidating President Erdoğan’s increasingly authoritarian regime (Apaydin & Çoban, Reference Apaydin and Çoban2023; Güngen & Akcay, Reference Güngen and Akcay2024, 163). Since 2018 at the latest, several regulatory agencies and the central bank became subject to political capture, forcing interest rates down to fuel credit expansion despite rising inflation (Çoban & Apaydin, Reference Apaydin2025, 136–7). Consequently, the Turkish economy experienced a major currency crisis in 2018 and skyrocketing inflation of more than 70 percent in 2022 (Figure 9).

To sum up, both South Africa and Turkey pursued a debt-based growth model in the first two decades of the new millennium, in part driven by global excess liquidity before the GFC and during the QE episode. When the global financial cycle contracted during the GFC and the 2013 taper tantrum, Turkey managed the threat of massive capital outflows better than South Africa by adopting a set of unconventional monetary policies. This enabled its government to push for more industrial development and manufacturing exports. South Africa, by contrast, continuously relied on financialization and the debt-based model in spite of signs of stagnation and endemically low growth (Bonizzi & Karwowski, Reference Bonizzi and Karwowski2024, 309). By pursuing a more heterodox economic approach, Turkey has been able to rebalance its growth model somewhat toward manufacturing and exports. South Africa, following a strictly orthodox macroeconomic framework, has even deepened its heavy reliance on capital inflows until recently. Hence, the case of Turkey shows that emerging economies dispose of a set of economic policy measures to mitigate the vagaries of global financial cycles. However, Turkey’s more recent political-economic trajectory warns of the serious medium to long-term negative side effects of certain debt-based growth models on the economy more generally and democratic institutions in particular. This has been amply demonstrated by economic history as well (Funke et al., Reference Funke, Schularick and Trebesch2016).

4.3 Foreign Direct Investment Flows and Investment-Led Growth Models

Growth model research on emerging economies has identified FDI-based growth models in several countries of the European periphery, such as Ireland and East Central Europe (Bohle & Regan, Reference Bohle and Regan2021; Reference Bohle, Regan, Blyth, Pontusson and Baccaro2022; Ban & Adascalitei, Reference Ban, Adascalitei, Blyth, Pontusson and Baccaro2022). In addition, earlier research in CPE has already shown how complete regions of the global economy, such as Central and Eastern Europe or Central America, have become structurally highly dependent on foreign investment (Nölke & Vliegenthart, Reference Nölke and Vliegenthart2009; Schneider, Reference Schneider2009). Since FDI comes from multinational firms, it does not merely reflect macroeconomic conditions, but also business strategies and global power relations (Strange, Reference Strange1996). Hence, we need to take global corporations and their influence on national growth models seriously when studying the growth trajectories of developing and emerging countries (Kaczmarczyk, Reference Kaczmarczyk2020; Bohle & Regan, Reference Bohle and Regan2021). Such an approach complements research on GVCs, which mainly operates at the sectoral level (Gereffi, Reference Gereffi2018).

Foreign direct investment flows to developing and emerging economies have increased manifold since the 1990s, representing 70 percent of the world’s total (Figure 10). This period has been marked by the information and communication technology revolution and the second phase of globalization, called the “second unbundling,” as factories became separated internationally due to rapidly enhanced ways of communication and, hence, coordinating complex activities over long distances (Baldwin, Reference Baldwin2016). Consequently, we have witnessed the rapid evolution of GVCs. More and more firms have organized production processes globally, offshoring parts or services to producers elsewhere. Therefore, GVC trade has increased significantly over time, accounting for about one-half of world trade recently (Antràs, Reference Antràs2020, 6). The rise of GVCs had important implications for developing and emerging economies, as it lowered the capabilities needed to enter such chains (Antràs, Reference Antràs2020, 11). However, not every type of GVC integration has proven equally supportive of sustainable growth and development. Commodity-based GVCs or lower value-added segments of production in general tend to exhibit lower economic returns for host countries. In addition, opportunities to upgrade from these lower production stages to higher-value-added stages, associated with more sophisticated technological know-how and more qualified employment, considerably depend on concerted industrial policies (Pipkin & Fuentes, Reference Pipkin and Fuentes2017). Empirically, FDI and GVC integration are significantly correlated and mutually reinforce each other (Qiang et al., Reference Qiang, Liu and Steenbergen2021, 6).

Figure 10 Foreign direct investment flows to developing and emerging economies

Source: Own elaboration based on UNCTAD FDI statistics

While FDI flows can lead to a transfer of technology toward developing and emerging economies, they also carry substantial disadvantages. Whereas the lead firms in global production chains are located in the high-income economies of the North, companies in the South often integrate in a subordinate position, being responsible for low-tech production, low-skills employment, and shouldering pollution-intensive activities (Wang et al., Reference Wang, He and Song2021). Furthermore, deep integration into GVCs negatively affects a country’s national sovereignty with regard to labor or environmental regulation, as multinational corporations can sue host states via investor-state dispute settlement mechanisms in international investment agreements and threaten to disrupt GVC integration of the respective country (Moehlecke et al., Reference Moehlecke, Thrall and Wellhausen2023). The political choice for an FDI-based growth model, therefore, includes a certain “retreat of the state” (Strange, Reference Strange1996).

We can observe the effects of FDI flows on growth models in emerging economies by studying Thailand and Vietnam, two countries with a very high share of FDI due to their proximity to East Asian manufacturing nations such as China, Korea, and Japan. So far, Southeast Asia is relatively underexplored in the GM literature. However, the region has been among the leading FDI destinations in the early 2020s, largely withstanding the global downward trend. In 2023 alone, the region received 226 billion US$ in FDI inflows with 1,568 greenfield announcements, a 42 percent increase from the previous year (UNCTAD, 2024, 7, 13–4). Comparing these countries to other FDI-dominated regions, such as Eastern Europe and Central America, may lead to promising insights into the intricacies of FDI-based growth. Thailand and Vietnam, in particular, have received vast amounts of direct investment since the 1990s, way above the ECE average. More recently, however, FDI inflows as a percentage of GDP have continued to increase strongly in Vietnam, while having declined in Thailand (Figure 11).

Figure 11 Foreign direct investment inflows as a percentage of GDP in Thailand and Vietnam

Source: Own elaboration based on UNCTAD FDI statistics

Competition among Southeast Asian countries has been fierce, as other neighboring countries such as Malaysia, Indonesia, and the Philippines are also attracting large volumes of FDI, and high-income economies have proclaimed a geopolitical “pivot to Asia.” Key factors for MNCs’ investment project decisions are labor costs and the overall political and economic risk, that is, a favorable “business climate” (Pandya, Reference Pandya2016, 462). Having attracted large sums of FDI early on, since the mid-1980s, and having become a relatively advanced economy compared to other countries in the region, Thailand’s cost advantage vis-à-vis its competitors has been shrinking over time (OECD, 2021, 96). Consequently, MNCs have gradually relocated operations to neighboring countries such as Vietnam or Cambodia. Adding to this, Thailand’s recent history since the mid-2000s is characterized by a series of political crises, social unrest, and two coup d’états in 2006 and 2014, all of which have undermined its investment appeal (Lorch, Reference Lorch2021, 87–9). Dwindling FDI inflows have gradually diminished Thailand’s export capacities and competitiveness (Nidhiprabha, Reference Nidhiprabha2017). In contrast to its troubled neighbor, Vietnam is considered one of the more politically stable countries in Southeast Asia. In addition to lower production costs, Vietnam has gradually opened its investment regime and has become one of the leading FDI destinations in the region (OECD, 2018b).

Hence, the case of Thailand shows the main problems associated with the FDI-based growth model. First, such a growth model is genuinely dependent on external factors, that is, the investment decisions of multinational corporations. If domestic factors, such as political conflicts or partisan politics, negatively affect the country’s status as an investment location, the mid-term sustainability of the model is threatened. Second, one of the key pillars of the model is low production costs, which are eroded by wage growth over time. At some point, MNCs tend to relocate their low-cost operations to other countries. Third, we observe differences in GVC participation across countries, generally captured by the distinction between backward and forward GVC participation. “Backward” participation indicates the degree of foreign inputs in a country’s exports, that is, the foreign value-added content of exports. “Forward” participation indicates the degree of a country’s domestic value added contained in exports. Whereas forward GVC participation is usually prevalent in commodities and commodity-related sectors, backward GVC participation is highest in simple or advanced manufacturing (Rodrigue, Reference Rodrigue2024). In contrast to Vietnam, we observe a divergence between backward and forward GVC participation in Thailand in the mid-2000s, with backward participation – involving a potentially higher technology transfer – developing less well (Figure 12).

Figure 12 GVC participation of Thailand (a) and Vietnam (b)

Source: Own elaboration based on OECD, Trade in Value-Added database

At the background of the more positive development in Vietnam is the conscious strategy of state authorities to maintain some degree of control over FDI and to extract a substantial transfer of technology in favor of local industry (Lim, Reference Lim2021). The government has been able to make sure that some value of FDI is retained in the country, in contrast to the more liberal approach in Thailand. Again, we are observing an industrial strategy, which is able to tackle some of the challenges of a dependent integration into GVCs.

In sum, this section has illustrated three types of growth models in the global periphery – commodity-based export-led, debt-based consumption-led, and FDI-based investment-led – that differ fundamentally in terms of the drivers underpinning countries’ economic growth and their associated vulnerabilities. However, the paired comparisons within each subtype have revealed substantial heterogeneity. These highlight the potential role of macroeconomic agency in the political management of growth drivers and vulnerabilities, which leads us to examine the importance of the political underpinnings of peripheral growth models.

5.1 The Politics of Peripheral Growth Models

The political support for growth models in ECEs is generally not established via electoral democracy (May et al., Reference May, Nölke and Schedelik2024). By contrast, the prevalence of the state in many high-growth economies has led to the impression that economic growth in ECEs is “state-directed” (Kohli, Reference Kohli2004).Footnote ¹² When we see the state not only as a regulator and mediator of political interests, as in the prevalent neo-institutionalist perspective, but also as a source of essential economic resources, its relations to the economic and social environment become important. In fact, there are varying degrees of how strongly the state apparatus is embedded in society (Migdal, Reference Migdal2001). This includes totalitarian or predatory regimes, where the state dictates social relations, and cases of state capture, where the state tends to be an instrument of powerful social groups. The concentration of fiscal resources within the state makes capturing a promising strategy, which leads to higher economic vulnerability (Masi et al., Reference Masi, Savoia and Sen2024). Therefore, it is crucial that the state displays a certain level of autonomy vis-à-vis demands by social groups, as the developmental state literature has convincingly argued (Evans, Reference Evans1995).

In what follows, we argue that two factors are crucial to achieve such a “balance of power” between the state and social groups in order to decrease the vulnerability of certain peripheral growth models. First, socioeconomic institutions, such as corporatist arrangements or patron-client networks, serve to include key social interest groups (or stakeholders) to benefit from a particular growth model and compensate excluded groups. Thereby, they help to resolve conflicts among social groups in peaceful ways and promote common over sectional interests (Besley & Persson, Reference Besley and Persson2011). These institutions support, second, the emergence of broad, bipartisan political coalitions which prioritize long-term investments and institution building over short-term political concerns (Doner & Schneider, Reference Doner and Schneider2016). They do this by helping to solve manifold collective action and coordination problems between state actors, businesses, and other social groups such as labor or local power groups (Waldner, Reference Waldner1999).

This contrasts with the widespread view that only impartial institutions, particularly legal ones, are able to bring sustained support to a growth model (see Section 2.1). In contrast, and in line with the political settlements approach, we take a realist stance toward political mediation: patron-client networks can also provide means for political stability and social cohesion. Under certain conditions, they can even increase government effectiveness by resolving principal-agent problems through enhanced monitoring and mutual trust between bureaucrats and politicians (Jiang, Reference Jiang2018; Toral, Reference Toral2023). Generally, we have to assume that “family,” kinship, or the “nation” can also be strong sources of cohesion. In most emerging economies, actors are likely to rely on trust, reciprocity, and loyalty in their economic and political relationships (Maxfield & Schneider, Reference Maxfield, Schneider, Schneider and Maxfield1997; Amsden, Reference Amsden2001; Witt & Redding, Reference Witt and Redding2013; Nölke et al. Reference Nölke, ten Brink, May and Claar2020; Taylor, Reference Taylor2020). For the analysis of growth models in these contexts, it is therefore important to note that political support does not (necessarily) mean democratic support. Still, relevant actors in the political economy will only support a growth model if they benefit from it.

Consequently, apart from the economic sources of growth (or “growth drivers”) covered in Section 3 and the international interdependencies discussed in Section 4, political factors also account for the degree of growth model vulnerability. In fact, they are partly interdependent. Current or historical growth trajectories are likely to shape the productive structure of an economy and the power resources of social groups benefiting from it. As an example, many emerging capitalist economies are characterized by deep structural cleavages between agrarian and financial elites and manufacturing sectors or between domestic business groups and foreign MNCs (Schneider, Reference Schneider2013). Furthermore, religious, ethnic, or spatial cleavages can play out politically and have an effect on growth model stability. These cleavages undermine efforts toward coalition building and can lead to sustained social fragmentation (Doner & Scheider, Reference Doner and Schneider2016, 611). Fragmentation, in turn, increases coordination problems between political and economic actors. In particular, in the event of external shocks, such as those emanating from capital outflows or plummeting commodity prices (see Section 4), lacking coordinating capacities becomes a binding constraint for effective policy solutions confronting an unfolding economic crisis. Therefore, we argue that growth models with a broader political and social support, relying on a relatively high degree of elite cohesion, tend to be more stable and resilient toward external shocks.

Our empirical analysis in the subsequent sections calls for a realist view on the politics of growth. Thereby, we aim to move beyond some traditional fault lines in the development discourse. Neither liberal approaches highlighting markets and legal systems, nor statist approaches focusing on bureaucracies and state capacities, are appropriate conceptual frames on their own for analyzing the political and economic realities in the majority of emerging capitalist countries. The political stabilization of a growth model, especially in times of crises, requires the broad political support of crucial social actors whose interests need to be accommodated. In this regard, formal peak associations or informal patron-client networks can intermediate state-business relations successfully and lend support to a given growth model or to a change in growth strategy. Although business associations in emerging economies vary substantially in terms of organizational capacities, they tend to facilitate coordination between their members, other stakeholders, and the state (Doner & Schneider, Reference Doner and Schneider2000). Well-organized associations may even contribute to competitiveness and economic growth (Nugent, Reference Nugent2023). As stated, these functions can also potentially be performed by informal state-business relations within elite networks (Nölke et al., Reference Nölke, May, Mertens and Schedelik2022). However, the latter are obviously more prone to decay into harmful versions of rent-seeking and corruption.

5.2 The Politics of Export-Led Growth Models

How does the politics of growth models now play out in the different types we presented earlier? To start with, an export-led growth model is in danger of being captured by groups that benefit most from an exclusive access to the sources of export revenues. This is mainly the case when exports are dominated by commodities rather than manufacturing, as there are generally less producer groups involved in raw materials exports than in manufacturing. Consequently, for commodities-based export-led growth models, we observe a high incentive for dominant groups to pursue rent-oriented politics. In order to maintain exclusive access to rents, powerful groups often resort to authoritarian means for the stabilization of the resource-based growth model. This is a well-known feature of many commodities-exporting economies and has been extensively discussed as the “resource curse” (Section 4.1; Ross, Reference Ross2015). When individual interest groups effectively monopolize the income/rents from growth, strategies for diversifying the growth model and making it more resilient toward exogenous shocks become politically less salient. Therefore, such an export orientation has to be balanced with other sources of growth in order not to endanger the stability of the model.

In contrast to commodities-based models, growth models based on manufacturing exports are more likely to sustain higher growth rates for a longer period of time and are supported by different political settlements. Generally, the South Korean case provides some lessons about the political resilience of manufacturing export-led growth models, despite different historical circumstances: As South Korea’s model drew (in the absence of notable commodities) on manufacturing exports, it managed to prevent the capture by groups that would focus on only a few products (Haggard et al., Reference Haggard, Kim and Moon1991). Old agrarian elites in Korea had lost political power with previous agrarian reforms, and the industrial business groups that emerged in the post-War period were politically weak, at least initially. The state thus was a powerful actor vis-à-vis industrial rent-recipients, making them work in line with industrial upgrading and economic diversification goals (Amsden, Reference Amsden1989; Khan, Reference Khan2010).Footnote ¹³ The crucial question is how to follow a growth model based on commodity exports that promotes economic diversification, industrial upgrading, and innovation. We argue that politics plays an important role here. Drawing on our case vignettes, we now show how Indonesia has managed to mitigate some vulnerabilities stemming from the commodities-based model politically, while the Brazilian political economy proved to be more vulnerable to political pressure from a coalition between financial capital, agribusiness elites, and the traditional middle class, and therefore deepened its commodity dependence.

From the late 2000s onward, the Indonesian government actively sought to put the commodities sector back under national control, predominantly in the oil, copper, and palm oil sectors (Winanti & Diprose, Reference Winanti and Diprose2020; Warburton, Reference Warburton2017; Reference Warburton2023). This had two effects: First, the partial exclusion of foreign firms from these sectors increased the power of both the state and domestic businesses; second, natural resources have been actively framed as a case of national sovereignty, particularly in the 2019 election campaign (Winanti & Diprose, Reference Winanti and Diprose2020, 8). In 2020, the Indonesian commodities sector was more or less in domestic hands again. In this way, the commodities sector plays a crucial role in the common project of Indonesian developmentalism (Warburton, Reference Warburton2023, 181–4). As a national project, it provides a political frame that lends itself to include the majority of relevant stakeholders. Renationalization changed the distribution of power among relevant groups, which became represented in a political settlement that stretches across state and business actors. Since natural resources have been framed as a national treasure, it has become hard for particularistic groups to claim exclusive ownership.

Through renationalization, the Indonesian state was able to offer domestic businesses the profit opportunities that hitherto had been exclusively exploited by foreign MNCs (Winanti & Diprose, Reference Winanti and Diprose2020, 2). In return, the government insisted that domestic business takes part in the resource nationalism program and invest in the domestic economy. At the same time, renationalization changed the contours of the domestic political economy. When commodities had been predominately marketed by foreign MNCs, Indonesia benefited merely in the form of foreign exchange and tax revenues. However, once the earnings of commodities exports flowed into domestic businesses, they would reinvest profits and thus engage in capital accumulation and building relations with up- and downstream businesses, thereby reducing the share of natural resource rents in GDP (Figure 13). Capital accumulation, in turn, also builds on institutions that businesses themselves cannot provide, such as education, credit, and labor legislation. Thus, once commodities became a national endeavor, the number of relevant stakeholders increased. Resource nationalism lent legitimacy to the growth strategy in the wider public and led to interest alignment between key domestic business actors, which proved to benefit from renationalization (Winanti & Diprose, Reference Winanti and Diprose2020). Although oligarchs and patronage networks are widely accounted for in Indonesia (Fukuoka, Reference Fukuoka2012; Hadiz & Robison, Reference Hadiz and Robison2013), these have not led to state capture but to increased competition both within state and private capitalist networks (Pepinsky, Reference Pepinsky, Ford and Pepinsky2014; Warburton, Reference Warburton2023, 180–81).

Figure 13 Natural resource rents as a percentage of GDP in Brazil and Indonesia

Source: World Bank Development Indicators

The differences in the Brazilian case are striking. First, the developmentalist project in Brazil halted after 2014 and was mainly framed as a classic case of catch-up industrialization with a higher degree of social justice. Domestic firms were not obliged to invest in forward or backward linkages inside the country, but rather encouraged to internationalize (Ricz & Schedelik, Reference Schedelik2023). Consequently, the resource boom did not materialize in a higher degree of capital formation of domestic firms, nor economic diversification. High revenues from commodities exports mainly served to accumulate record levels of international reserves, easing the risk of exchange rate crises (common in the 1990s). Lacking a national project, the usual cleavages in the Brazilian political economy persisted: the strong position of the landowning classes vis-à-vis the poor; Brazilian MNCs enjoying state support for the penetration of world markets vis-à-vis domestic firms aiming for the domestic market; powerful conglomerates vis-à-vis an overall weakly organized business sector (Taylor, Reference Taylor2020; Morgan et al., Reference Morgan, Doering and Gomes2021; Schedelik, Reference Schedelik2023, 129–31).

Weak business organization affected the clientelistic favoritism of single firms that sought to further monopolize their influence (May, Reference May, Mikler and Ronit2020). For these reasons, they would push for further export specialization rather than diversification. The tight connection between the state and particular firms, especially in commodity-related sectors, led to a situation in which the ruling coalition became ever more dependent on firm support via campaign donations or bribes and kickbacks, leading to ever-increasing rent payments (Schedelik, Reference Schedelik2023, 108). Importantly, firms involved in this dynamic typically did not belong to high-value-added manufacturing. Whereas the state in Indonesia was able to urge domestic firms to support a resource nationalist strategy, the political mediation in Brazil became less diversified and more dependent on the state-owned oil company Petrobrás and on large corporations in the construction sector. This Petrobrás-construction nexus became a steadily growing allocator of rents, with ever more groups enjoying rents fueled by windfall profits (Taylor, Reference Taylor2020).

In the absence of competitive pressures from other firms or a stronger transmission of revenues into capital formation and technological upgrading, the incentive to further push for extractivism became ever stronger, minimizing state autonomy vis-à-vis firms and making the commodity-based growth model ultimately unsustainable. The coalition behind extractivism has been too small to actually stabilize the whole growth model politically (Morgan et al., Reference Morgan, Doering and Gomes2021). Brazilian developmentalism has been short-lived and never became a collective project, neither among business nor in the wider public. Widespread skepticism about self-interested elites intensified particularly after the “lava jato” corruption scandals (Nölke et al., Reference Nölke, May, Mertens and Schedelik2022, 324). When the lava jato investigations uncovered the vast corruption schemes involving the PT and many of its allies, the parts of the growth coalition that relied heavily on the Petrobrás-construction nexus virtually imploded. Ultimately, the fragmentation of the coalition, alongside global turbulence, undermined the stability of an existing growth path, giving way to the subsequent right-wing governments of Temer and Bolsonaro, who ran on a neoliberal platform (Morgan et al., Reference Morgan, Doering and Gomes2021, 547).

Clientelism reigns both in Indonesia and Brazil (Epstein, Reference Epstein2009; Sugiyama et al., Reference Sugiyama and Hunter2013; Aspinall, Reference Aspinall2014; Aspinall & Berenschot, Reference Aspinall and Berenschot2019; Power & Rodrigues-Silveira, Reference Power, Rodrigues-Silveira and Ames2019). Both countries are by no means poster children for a politically efficient export-led model, as we know from the state autonomy literature (Evans, Reference Evans1995; Kohli, Reference Kohli2004). In contrast to earlier prevalent theories of economic development (Section 2.1), however, the growth model approach also accounts for the less-than-perfect cases where the rule of law, property rights, and impartial bureaucracy are not necessarily granted. In this realist perspective, patronage can equally serve as a means to establish cohesion, especially when the circles of relevant stakeholders become larger and closely connected to the production of higher value-added tradable goods. While important groups have managed to form a political settlement in Indonesia, they have further intensified confrontation in Brazil.

Indonesia, thus, is a case that defies the notion of a resource curse or patrimonial capitalism in some Arab countries (see also Gellert, Reference Gellert2019; Hilmawan & Clark, Reference Hilmawan and Clark2019; on patrimonial capitalism, see Schlumberger, Reference Schlumberger2008). The Indonesian growth model is not about resource rents all the way down. This means that even within a commodity-based export-led model, there is room for maneuver for a strategy that is not excessively dependent on export revenues, which are vulnerable to fluctuations. It also means that such a growth model is not doomed to get stuck in wasteful rent-seeking in the commodity sector, where export revenues are fueling a clientelist regime, making it politically vulnerable because crucial groups are naturally excluded from those rents. Moreover, the Indonesian case highlights how a political economy with both state- and privately owned firms in the commodities sector is able to support a developmentalist, that is, long-term program. It also shows that states, even if they do not have high levels of state capacities, such as in China, are able to limit the influence of foreign capital. Such a constraining policy toward FDI can have mobilizing effects for domestic capital, which solidifies an otherwise highly volatile sector (for the Chinese case, see Gomes & ten Brink, Reference Gomes and ten Brink2023).

5.3 The Politics of Consumption-Led Growth Models

A consumption-led growth model is potentially threatened by groups that aim to expand (external) debt, thereby increasing the exposure to volatile capital flows (Section 4.2). Through resorting to external debt, it is possible to fuel consumption-led growth without actually having to promote sustainable gains in workers’ income through wage increases. Special interest groups, mainly from the finance and real estate sectors, can push such a model, without incorporating other important social groups, such as workers and the poor. Such a strategy requires political resources, but is often easier than organizing broad social compromises, which also involve sharing the revenues of the model among a broader social base. Hence, if a consumption-led model tilts too heavily toward external debt in foreign currency, it becomes vulnerable to fluctuations in international currency markets. Subsequently, we draw on the Turkish and South African case vignettes to show how political dynamics in the former helped to stabilize and diversify a faltering growth model. By contrast, entrenched financial and mining interests locked-in a failed growth model in the latter, leading to crisis and stagnation.

For similar reasons that made Indonesia more stable than Brazil, the political support for balancing Turkey’s growth model has been stronger than in South Africa. Again, both countries are far from perfect examples of sustainable economic development, but this is precisely why one can learn from their experience. Like Indonesia, Turkey learned its lessons from severe financial crises and sought to limit the external financial vulnerabilities, for example, by crafting national economic projects. “Anatolian capital” or “Islamic capitalism” have been only two economic “identities” that unified a range of political and economic actors throughout the 2000s (Demir et al., Reference Demir, Acar and Toprak2004; Demiralp, Reference Demiralp2009; Balikci, Reference Balikci2015). Organized capital, labor, and broad segments of voters from multiethnic and various socioeconomic backgrounds supported the Justice and Development Party (AKP), following its first election in 2002 (Apaydin, Reference Apaydin2025, 360).

More substantial, however, has been the structure of the Turkish business sector. The Turkish economy is characterized by small and medium enterprises rather than a few powerful national giants (Topal, Reference Topal, Yalman, Marois and Güngen2019, 221). As we saw previously, the dominance of a few large companies tends to increase instability because they (mostly domestic MNCs) hardly represent a broad range of groups or sectors and usually are able to pull exit options if necessary. The concentration of bargaining power in such large MNCs leaves too many groups unrepresented in a growth model, particularly when states follow pro-business politics, heavily catering toward a few well-connected firms.

In Turkey, we do not observe such a dominance but rather a relatively effective business organization through associations. Not only do they represent important groups along economic and political cleavages, but they are also particularly powerful in their capacity to organize resources. The Union of Chambers and Commodity Exchanges of Turkey (TOBB), for instance, is representative for a large part of the SMEFootnote ¹⁴ sector but is, at the same time, a peak association that provides crucial resources to both state and business. As Özel (Reference Özel2015, 138) summarizes, it

Hence, it is difficult for the state to ignore its preferences while at the same time, it urges larger firms to join in the domestic economic project. Other associations, such as TÜSÍAD, represent big business in a secular orientation, while MÜSÍAD would call for a stronger Islamic foundation for the Turkish economy (Yavuz, Reference Yavuz2010; Buğra & Savaskan, Reference Buğra and Savaşkan2014, 125). Thus, business associations function not only as powerful organizations, but also as key political actors, representing both businesses and wider political orientations. These and other major business organizations across sectoral lines lent substantial support to the government and its growth strategy (Apaydin, Reference Apaydin2025, 361–2). However, in the course of the 2010s, this broad supporting coalition declined substantially, especially in light of Erdogan’s authoritarian turn from 2013 onward. Following the attempted coup of 2016, authoritarianism was further consolidated. Still, key business groups in the manufacturing, construction, and energy sectors, as well as influential segments of the Islamist middle classes and workers, endorsed the government’s shift toward a more diversified growth model, based on an industrial strategy and export promotion (Akcay & Jungmann, Reference Jungmann2023, 553–4; Apaydin, Reference Apaydin2025, 363–7).

Still, we observe a diversified representation of economic and political interests that amounted to a stable balance of power between state and business for the better part of the 2000s and early 2010s in Turkey. A large part of domestic consumption took place in the construction sector (Erol, Reference Erol, Yalman, Marois and Güngen2019), which in turn involved crucial stakeholders: the state, large domestic construction businesses, and domestic banks. In this form, construction remained a Turkish business because neither foreign firms have been significantly involved in construction nor is retail investment significantly connected to foreign capital markets. It comes as no surprise that the construction sector, where multiple interests can be found, is a major site of clientelist networks (Sayarı, Reference Sayarı2014, 11). However, these networks can be interpreted as means of coordination and compliance rather than purely rent-seeking. Similar to Indonesia, we see the co-existence of patronage and stable political settlements. As neither the state nor major businesses is able to monopolize power, patronage and clientelist networks have been able to include a growing number of relevant stakeholders, thereby mitigating the withdrawal of support by some social groups (Sayarı, Reference Sayarı2014, 11; Apaydin, Reference Apaydin2025).

In South Africa, in turn, the political foundation for a sustainable debt-based consumption-led growth model has been far more unstable. At first, it has been far more transnationally integrated than the Turkish one, leading to a much higher degree of vulnerability from volatile financial markets (Isaacs & Kaltenbrunner, Reference Isaacs and Kaltenbrunner2018; Karwowski, Reference Karwowski2018; Ansari, Reference Ansari2021). Second, the South African state is not powerful enough to garner widespread support for a more diversified growth model or inhibit crucial actors to draw exit-options: Foreign firms and banks are players in the political bargains over the growth model (which they are not in Turkey or Indonesia), at the same time, large domestic businesses have exit-options since they have long established transnational corporate relations (Karwowski, Reference Karwowski2015). The large South African mining MNCs are neither dependent on nor interested in submitting to a national growth project but follow particularistic agendas (Chipkin & Swilling, Reference Chipkin and Swilling2018; Claar, Reference Claar2018; Bowman, Reference Bowman2020). In addition, the powerful financial sector, with its strong ties to the Treasury and the Central Bank, has been highly successful in blocking alternative growth strategies, even in the wake of the global financial crisis. Those alternative industrial strategies were promoted by leftist factions of the African National Congress and the major trade unions but failed to alter the prevailing debt-based model (Naqvi, Reference Naqvi2023, 301–7).

Third, racial cleavages still represent a deep rift in the South African political economy, one of the most unequal countries globally. As an example, the promotion of Black Economic Empowerment seeks for completely different plans on how to organize the economy as a whole, contesting highly unequal and racialized ownership structures of farm land in particular. Instead of a diversification of stakeholders, South Africa is rather characterized by a high degree of fragmentation, where capitalist factions cannot settle on the common support for growth models (Bowman, Reference Bowman2019). Of course, clientelism and patronage are also widespread in South Africa, but here it takes the form of monopolistic rent-seeking, which typically prevents economic diversification and industrial upgrading. Just like Petrobrás in Brazil, the national electricity giant ESKOM served as the main source and distributor for rents (Ansari, Reference Ansari2021, 172). When ESKOM got into trouble, the flow of rents diminished, and so did the compliance of beneficiaries. Political coordination as a whole, therefore, is weak, amplified by the persistence of racial hierarchy and state violence in the resolution of labor disputes (such as in the Marikana massacre during a miners’ strike).

To sum up, the higher vulnerability of the South African debt-based growth model toward global financial cycles in comparison to the Turkish one (see Section 4.2) is also based on its much more fragmented political support base. Neither state nor business is powerful enough to establish a common and coherent alternative growth strategy. In Turkey, in contrast, the AKP-led government had been able to assemble a broad support base for its growth strategy throughout the 2000s and early 2010s, also comprising of small enterprises and a large share of households. Although this broad support was gradually weakened in light of rising authoritarianism by the end of the decade, it helped to stabilize and diversify a faltering debt-based model.

5.4 The Politics of Investment-Led Growth Models

Finally, an investment-led model has distributional implications as well, which highlights the importance of the political support coalitions underpinning this model. Investments can come from external as well as from internal sources. An excessive orientation on foreign investments, however, raises the vulnerability of the growth model because foreign owners are more likely to divest funds when conditions allow.

It is difficult to speak of capture properly in this case, but this kind of danger also applies to investment-led growth models. Some social groups and political elites have a high interest in attracting FDI: FDI is usually indicative of a promising manufacturing industry, it eases the balance of payment constraint, and comes with no strings attached in a political sense, as no social group suffers from a relocation of investments, at least initially (Owen, Reference Owen2018). Hence, from a political perspective, FDI can be seen as a form of rent (Pinto & Zhu, Reference Pinto and Zhu2022; Brazys et al., Reference Brazys, de Soysa and Vadlamannati2023), which will increasingly define the whole growth model and, eventually, the entire political economy can become heavily dependent on FDI (Bohle & Regan, Reference Bohle and Regan2021). Indeed, a large share of FDI in some sectors can lead to a crowding out of investments by domestic groups, typically less resourceful than large MNCs. Moreover, even considering the positive effects of MNCs’ productive investments – those investments that most relevant groups in an economy welcome – over time, the share of portfolio investments may grow as a strategy to keep the balance of payments in check.

At this point, such a foreign-dominated investment-led model becomes vulnerable for at least three reasons: First, in the absence of capital controls, capital that flows in easily often flows out easily as well, especially when we are dealing with large volumes of portfolio capital that is highly mobile. Second, the different domestic economic groups in an FDI-dependent economy are not essentially part of the growth model but merely pacified. As research on the Eastern European economies has shown, foreign chambers of commerce and large foreign investors play an outsized role in shaping and sustaining this type of growth model (Scheiring, Reference Scheiring2020; Bohle & Regan, Reference Bohle and Regan2021). Hence, broader political support for an FDI-led growth model often relies on the promised spillover effects of FDI inflows for growth, innovation, jobs, and tax revenues, which usually remain on a rather abstract scale and are yet to materialize. This means that the political support for an FDI-led model is rather fragile because the costs for most actors to exit the model are relatively low. Thirdly, FDI dependency usually goes together with policies to attract FDI. These policies involve direct (e.g., subsidies) and indirect transfers (e.g., tax exemptions) to foreign corporations – which open the gate for distributional conflicts.

Again, the two cases discussed in the context of international interdependencies are instructive in this regard (Section 4.3). Thailand has a recent history of political instability, thereby discouraging MNC investment. At least as important is the loss of labor cost advantages on the side of Thailand, vis-à-vis neighboring countries such as Vietnam. Finally, Thailand relies on the voluntary transfer of innovations by MNCs, thereby increasing the probability of remaining stuck in a middle-income trap. Vietnam, in contrast, has been developing institutions with the objective of ensuring that investments by MNCs involve technology transfer toward the domestic industry (Klingler-Vidra & Wade, Reference Klingler-Vidra and Wade2020). In fact, Thailand has followed an open inward investment policy since the 1980s already. As mentioned earlier, in contrast to portfolio investments, greenfield investments usually come with technology, management innovation, and integration into international production networks. Yet, any productive effect of FDI requires appropriate “absorptive capacities” (Cohen & Levinthal, Reference Cohen and Levinthal1990) in the host country, which are mainly missing in Thailand. As Thailand never had a substantial industrial or technology policy, foreign investment could not increase the capacities of domestic firms, which has led to hardly any developmental impact (Kanchoochat et al., Reference Kanchoochat, Aiyara and Ngamarunchot2021, 16).

Vietnam, in contrast, represents a gradual process of market opening (Ngo & Tarko, Reference Ngo and Tarko2018). The introduction of market principles from the 1980s onward did not follow a prefigured plan set by the communist party (Gainsborough, Reference Gainsborough2017, 136). Therefore, Vietnam is far from the ideal-typical developmental state that is able to both make grand development plans and implement them through an effective bureaucracy. The evolving state-business relation can rather be understood as a form of “hybrid corporatism ” (Nguyen, Reference Nguyen2014, 338). This has worked because business has some degree of organizational power and therefore could establish peak associations. Unlike the dirigiste model of coordination as it is found in traditional developmental state regimes, associations are potentially helpful in managing the diverse interests that arise in processes of societal and economic fragmentation. Local businesses, political cleavages, and a fragmented state do not call for a political model of a strong state (Gainsborough, Reference Gainsborough2017; Pincus, Reference Pincus and London2023). Both the ambitions and strategies of the state and business are changing constantly, and such an intermediation is able to accommodate the changing alliances that go along with it.

Informal mechanisms of coordination play an important role as well. However, these are not mere instruments for the appropriation of monopolistic rents but help to include a range of actors in a common political project. Although less salient today, a blend of socialism and economic nationalism still serves as an important ideological framework that large parts of Vietnamese society can adhere to (London, Reference London and London2022, 36–7). From this perspective, the “Vietnam example … illustrates that a crony capitalist growth path could enable growth and development, thanks to the interplay between politics, institutions, and markets, despite the lack of universal rule of law as well as contract and property protection” (Ngo & Tarko, Reference Ngo and Tarko2018, 15). Similar to the case of Turkey, we observe a productive kind of clientelism.

Many of the elements that helped to stabilize the Vietnamese growth model politically have been absent in Thailand. There has been no common growth or development project that could serve as a political umbrella. However, what is even more important are stable procedures of building societal coalitions, reliable mechanisms of coordination or accommodation, and non-monopolistic relations of power. In these respects, Thailand is much more unstable and imbalanced. Unlike Vietnam or Turkey, we neither find effective associations that would act as corporate actors mediating between state and business nor other significant institutionalized forms of state-business relations (Doner & Schneider, Reference Doner and Schneider2016, 633). Like in Brazil or South Africa, large corporations have substantial political influence, following particularistic interests.

Under such circumstances, including an unstable polity, the risk of monopolistic rent absorption is high. The Thai economy is dominated by a small number of families and their conglomerates. Through their preferred access to the state (e.g., via receiving licenses), these groups could grow exponentially compared to the average growth in the country (Kanchoochat et al., Reference Kanchoochat, Aiyara and Ngamarunchot2021, 17). We find a broad stratification, where elites compete for power and resources at the top, while poverty increases at the bottom. Politically, Thailand is highly fragile: Instead of common growth projects, we have observed two coups d’état within ten years (2006 and 2014). Added to the “usual” competing elites of state and business, military elites – who pursue their own economic interests – complicate the situation even further. Deep cleavages and open conflicts between supporters of former Prime Minister Thaksin Shinawatra (2001–2006) and his adversaries have marked the Thai polity for two decades, leading to a highly vulnerable authoritarian coalition ruling the country since 2014 (Sen & Tyce, Reference Sen and Tyce2019, 131–2). As one observer put it, “[a] lack of consensus around basic ‘rules of the game’ among elites and civil societal groups renders the country highly volatile and unstable” (Kongkirati, Reference Kongkirati2016, 17). In this context, it should come as no surprise that stable support for a particular growth model is hard to achieve.

In conclusion, associations are a central counterpart for the state as well as a representative for their constituency. Their establishment goes hand in hand with organizational professionalization because, as main negotiators, they have to move beyond particularistic interests. As we know from established corporatist systems, associations represent a crucial element in the coordination of different interests, therefore contributing to the process of accommodation between businesses as well as between state and business. Whether they use formal or informal channels of coordination does not really matter here. What matters is that the political stabilization of a growth model rests on a broad support where crucial actors cannot defect easily. If we look at our case vignettes, we can see that the less vulnerable economies (Indonesia, Turkey, and Vietnam) have peak associations while they play a very little, if any, role in Brazil, South Africa, and Thailand. Although we would not go as far as to say that associations are key to the stabilization of growth models, their existence seems to have a positive effect. They indicate the existence of some level of intermediation between state and business, which helps to foster elite cohesion and to solve manifold coordination problems, in particular in times of crises.

More generally, we have shown that political support is a crucial factor for the stability of a growth model and the mitigation of its external vulnerabilities. Through our analysis of a small number of cases, we are starting to be able to specify which political conditions have to be met and which aspects are analytically less important. Broad support for a growth model adds to its stability, especially under conditions of high vulnerability: When state-business relations are not monopolized by a few large firms, it reduces the risk that these actors exit a given growth coalition and opt for a different strategy. Similarly, it is harder for these actors to lock-in an unstable growth model, thereby increasing the vulnerabilities that arise when economies become too dependent on commodity exports, external debt, and FDI inflows.