From the Great Divergence to South–South Divergence: New comparative horizons in global economic history

Abstract

The Great Divergence debate has been the leading conversation in economic history for the past 25 years. This review article explores new comparative horizons in global economic history. I argue that questions of South–South Divergence form a logical and timely extension to the Great Divergence research agenda. Asia's economic renaissance did not only put an end to a century-spanning process of widening global income disparities, it also set a new process of divergence within the global south in motion. Deeper understandings of the historical nature and origins of this transition are pertinent in light of the increasing demographic and economic weight of the global south. South–south comparisons also offer an opportunity to counter the dominance of western-centered and north–south perspectives and incentivize the development of new approaches and theories that go beyond mainstream concepts designed by development economists and political scientists. I argue that these novel approaches will have to grapple with the opportunities and constraints to “late” development being shaped by the quadruple challenge of vast technology gaps, limited state autonomy, global competition, and rapidly closing land and resource frontiers.

1 INTRODUCTION

The Great Divergence debate has been the field-defining conversation in economic history for the past 25 years. This debate revolves around the key question why the Industrial Revolution originated in Western Europe, and more specifically in Britain, and not in China, India, or Japan (Allen, 2009; Bin Wong, 1997; Mokyr, 2018; Parthasarathi, 2011; Pomeranz, 2000; Vries, 2020). By unifying scholars around an overarching comparative agenda, the debate has done much to globalize the field of economic history and to stimulate the construction of world-spanning databases on historical GDP, real wages, skill premiums, government revenues, terms-of-trade, human capital, land use, and more (cf., Maddison, 2003; van Zanden, 2009; Allen et al., 2011; Williamson, 2011; Morris, 2010; van Zanden et al., 2014; Broadberry et al., 2015). Such data collection and estimation efforts, in turn, have provoked heated discussion on the methodological and theoretical underpinnings of income and welfare measurements (Bin Wong & Rosenthal, 2011; Deng & O'Brien, 2017; Ma & van Zanden, 2017; Hatcher & Stephenson, 2018; Goldstone, 2021), and on the critical importance of reciprocity in comparative economic history (Pomeranz, 2000; Austin, 2007; O'Brien, 2020). The topic has also attracted scholars from adjacent disciplines ranging from economists and political scientists who have applied novel econometric techniques to argue that global inequality is all about institutions (Acemoglu et al., 2001; Acemoglu & Robinson, 2012), to evolutionary biologists and psychologists who have advocated the role of critical disparities in biogeography or the human psyche (Diamond, 1997; Henrich, 2020).¹ Finally, the Great Divergence debate gave inspiration to new concepts such as the European and Asian “Little Divergence” (Sng & Moriguchi, 2014; De Pleijt & van Zanden, 2016).

As goes for all major academic debates, however, at some point in their life cycle, decreasing marginal returns are inevitable. Once original questions fade as new adjacent windows of exploration open. In this review essay, I argue that one of the most promising new research avenues that can emerge out of Great Divergence scholarship focuses on the rapid, and more recent, economic divergence across––what for want of a better term is labeled as––the global South. I will develop this argument by doing five things in the following order: (1) providing a basic definition of the concept of South–South Divergence and a motivation for why this is of interest to economic historians; (2) demonstrating how the phenomenon of South–South Divergence is embedded in preceding processes of global economic divergence; (3) explaining why the notion of reciprocal comparison, one of the fruits of Great Divergence scholarship, can be used to construct meaningful south–south comparisons; (4) outlining three major questions that can guide the South–South Divergence research agenda; and (5) indicating how this new comparative horizon in economic history can be made relevant for anyone (policy-makers, journalists, fellow academics) trying to come to terms with contemporary reconfigurations of the world economy. In doing so, I will emphasize the need to consider, theoretically and empirically, how “late” development has been conditioned by the globalized, post-colonial and closed-frontier world order of the late 20th and early 21st century. I will explain why these conditions matter for more systematic explorations and explanations of South–South Divergence. But before we embark on the path to a new comparative horizon in global economic history, we first ought to step back into the Great Divergence literature to locate a shared point of departure.

2 TWO VERSIONS OF THE GREAT DIVERGENCE

The term Great Divergence was originally coined to refer to the widening gap in per capita incomes, and related disparities in industrial and technological capacity, between Western Europe and Eastern Asia since the British Industrial Revolution (Bin Wong, 1997; Pomeranz, 2000; Vries, 2010; Goldstone, 2021), or in the centuries leading up to it, as others have argued (van Zanden, 2009; Allen et al., 2011; Bassino et al., 2019). And yet, as oft-used concepts easily wither, the term has also been increasingly adopted in references to the yawning income gap between the Northern industrializing “core” and the underdeveloped (or de-industrialized) Southern “periphery,” a divide epitomized in the age of high imperialism during which a handful of European metropoles controlled more than half of the world population (Williamson, 2011; Baldwin, 2016; Henrich, 2020).² For the sake of clarity, I will distinguish the “Eurasian” from the “Global” version of the Great Divergence.

Of course, the theme of global economic inequality had been on the radar long before Pomeranz (2000) adopted the Great Divergence as the title of his famous book, but his work did mark the start of a new wave of research that explicitly sought to understand the British or European Industrial Revolution in a global perspective. Building on Bin Wong (1997) and Frank (1998), Pomeranz’ made a convincing case for replacing Eurocentric perspectives on the timing and causes of the Eurasian Divergence, by more balanced, reciprocal comparisons in which neither region is taken as the norm to mirror the other. Reciprocity amongst others implied a correction of comparative scales. Northwestern Europe, and Britain in particular, was no longer to be compared with China as a whole, but with its most dynamic core, the cloth-producing Yangzi delta.³ Doing so, Pomeranz argued, reveals “a world of surprising resemblances” in Western Europe and China before 1800 (2000, p. 29). These points about how to conduct global economic history have been widely reinforced (O'Brien, 2006; Austin, 2007; Bin Wong & Rosenthal, 2011; Li & van Zanden, 2012) and closed an era in which processes of divergence were articulated in one-sided frames such as The rise of the Western World (North & Thomas, 1973), The European Miracle (Jones, 1981), or The Wealth and Poverty of Nations (Landes, 1998).⁴

The Eurasian Great Divergence debate condensed around the revisionist narrative of the so-called California school (of which Pomeranz is one of the leading proponents) and the older, mainstream narrative of the advocates of European exceptionalism. Pomeranz’ explanation of the Divergence emphasizes British access to “coal” and “colonies,” fortunes of nature, rather than the supposedly unique features of Northwestern European capitalism, underpinned by the “bourgeois virtues” or “cultures of enlightenment,” that are central in explaining the scientific revolution and the ultimate transition to modern economic growth (Landes, 1998; McCloskey, 1998; Mokyr, 2009, 2018).

This paradigmatic divide is not so clear in all scholarship though. For example, Allen's (2009) emphasis on disparate trajectories of factor-biased technological change in Britain and China is compatible with the revisionists’ argument that the accessibility of coal was a sine qua non for British industrialization, while his explanation of Northwestern Europe's “high-wage economy” is rooted in the exceptional features of European marriage patterns, market development, human capital accumulation, and early forms of mercantile and industrious capitalism (De Vries & van der Woude, 1997; van Zanden, 2009; Broadberry et al., 2015). In a similar vein, historians who have emphasized how the twin-forces of military competition and fiscal state centralization gave rise to a distinctly European political economy of state-formation, do not necessarily believe that this difference resulted in exceptional European productivity and welfare levels before 1800 (He, 2013; Vries, 2015; Hoffman, 2015; O'Brien, 2020).

Although the central historical questions underpinning the Eurasian and Global Great Divergence debates are intertwined, they differ in critical respects. One way to appreciate these differences is to consider how India, and other parts of the global South for that matter, have been written into the Great Divergence frame. Contrary to Japan, where the threat of European imperialism has been considered as a catalyst for domestic reform, or China where European and Japanese invasions are primarily seen as a result of the divergence, for India, the agenda has been informed by the question, whether, how and to what extent European imperialism and related forces of globalization have caused the divergence (Broadberry & Gupta, 2006; Studer, 2008; Parthasarathi, 2011; Roy, 2012; De Zwart & Lucassen, 2020). In a similar vein, Reid (1993) has argued for Southeast Asia that the increasing encroachment of militant European trade companies in the region put an end to a remarkably prosperous “Age of Commerce” between 1450 and 1680. European colonialism also interrupted the “parallel integrations” which Lieberman (2010) has described so evocatively. He argued that from c. 800 to 1830 CE, around the perimeter of Eurasia, a dramatic long-term reduction occurred in the number of independent policies, in which Europe, Japan, and mainland Southeast Asia went through comparable processes of territorial and administrative consolidation and fiscal centralization. In Trade and Poverty, Williamson (2011) links 19th-century globalization and European imperialism to the delayed diversification of production structures across the major part of the global south. For large parts of Africa, the trans-Atlantic slaves trades as well as later colonial invasions have also often been mentioned as part of the explanation for economic retardation (Rodney, 1972; Acemoglu et al., 2001; Nunn, 2008; Beckert, 2015). Distinctively different forms of colonial rule and settlement patterns also play a major role in debates regarding the North—South Divergence in the Americas (Sokoloff & Engerman, 2000; North et al., 2000; Elliott, 2006; Fukuyama, 2008; Mahoney, 2010).

This is not to say that European imperialism plays a major role in all accounts of global economic divergence, as is testified by work on the supposed economic predicaments of Islamism, or in studies pointing to diverging cultural, psychological, or genetic evolutionary trajectories (cf., Kuran, 2011; Henrich, 2020; Galor, 2022). And indeed, the case for distinct biogeographic conditions of agricultural and commercial development has been put forward by many scholars, either as the main determinant or as an important confounding factor. So it is probably more accurate to state that theories accounting for the Eurasian Divergence focus on the differences and similarities in the largely autonomous long-run development trajectories of China, Japan, and Western Europe, and treat empire-building as a political option open to both, while theories accounting for the Global Divergence must grapple with the long-term implications of exogenously (read: European) imposed state institutions, population settlements, trade relations, and production structures, and how these interact with local conditions and developments. This distinction is also relevant for theories aiming to explain South–South Divergence, which by definition have to take the interactions of the local with the global—including the imperial—into account, as well as processes of regional (dis)integration.

While the Eurasian and Global divergence in income levels came to an end during the second half of the 20th century, income gaps within the global south began to widen. The turning point is impossible to pin down in time precisely, but Figure 1 offers three possible ways of looking at the question of periodization. The black line shows the ratio of per capita GDP in the UK and China, which indicates that the income gap plateaued in the 1950s to 1970s, and that rapid convergence has set in since the late 1970s. At its peak, average British per capita incomes were 16 times larger than Chinese incomes, to fall back to a ratio just below 3 by 2020. Another approach is to take the unweighted average GDP per capita levels of 40 Northern and 105 Southern economies, using the so-called “Brandt line,” to divide countries in both groups (see also Section 3). Since we lack reliable GDP per capita series for a large share of Southern countries, this series can only be constructed for the post-1950 period. The north–south ratio is shown by the blue line. It reveals a peak in per capita income gaps in the late 1980s, followed by a rapid decline to levels observed in the 1950s. The red line shows the north–south ratio in total GDP and is based on rudimentary estimates of regional GDP in the Maddison Project Database. This measure indicates that around 1820, the share of world income generated in the south still exceeded the share of the north, while the forces unleashed by the Industrial Revolution reverted the picture during the 19th century. At its peak, around 1950, the total GDP in the north was two and a half times as large as in the south. In the past seven decades, this ratio began to decline, however, so that by 2020 the relative shares had come back to the estimated levels of 1820.

FIGURE 1

Ratios of GDP per capita and total GDP in constant 2011 US$, 1820–2020. Source: Maddison Project Database (2020). Notes: Global South excluding the Gulf Countries; Global North excluding former Soviet states. To avoid India and China dominating the picture, the GDP per capita North/South ratio is not weighted by population, total GDP north/south is of course influenced by demographic trends. [Colour figure can be viewed at wileyonlinelibrary.com]

Of course, the heroic assumptions that underpin such relative income estimates must be kept in mind, as part of the Great Divergence scholarship has argued (Deng & O'Brien, 2017). Assigning start or end dates to processes of long-run divergence does not only depend on the type of indicator selected, but also on the quality of the data required to construct it. Yet, these considerations should not bother us too much here. The key point is that all the available income estimates demonstrate that both the Eurasian and Global versions of the Great Divergence came to an end during the second half of the 20th century, and as we will see in the next section, their endings intertwined with rapidly widening income gaps in the global South, a term that now warrants further precision.

3 SOUTH-SOUTH DIVERGENCE, SO WHAT?

The “Brandt-line” (Figure 2) offers the most widely used delineation between the global south and north. According to this line, the south includes the whole of Latin America, Africa, the Middle East, South and Southeast Asia, and the major part of Northeast Asia, except Japan, which is included in the global North. I do not adopt this division here because I find it very helpful to split the world economy of today in these two parts. I adopt it because it epitomizes the historical conception of the Global Divergence as it reached its hightide in the 1950s through to 1980s. The juxtaposition of the north to the south was emblematic for a global political order that had just experienced the crumbling of major European empires, but was still enmeshed in the habit of dividing the world in major blocks.⁵ By implication, South–South Divergence is defined as the growing income divide between countries (and regions) across the global south, a process that put the merit of north–south perspectives itself into question.

FIGURE 2

The Brandt line. Source: Brandt et al. (1980, p. 31–32) and front cover.

In their 1980 report North–South: A Program for Survival, the Independent Commission on International Development Issues (ICIDI) chaired by former German Bundeskanzler Willy Brant, distinguished between a developed, industrialized north and a developing, nonindustrialized south, spliced in two by some sort of economic equator. The report was influential. The notion of a “global South” was increasingly embraced in academic and policy circles during the 1990s and became, according to bibliometric analyses, mainstream in scientific parlance during the past two decades (Pagel et al., 2014). Interestingly, this happened right at a time that the south as an already rather heterogeneous group of “underdeveloped” economies really began to fall apart.

The Brandt report breathed a strong sense of crisis stemming from the global economic ruptures of the 1970s, including the collapse of the Bretton Woods system, two oil crises and mounting debt problems across the developing world. Presented as a “program for survival,” the committee warned that in view of rapid demographic growth and resource depletion, a failure to address the intertwined issues of global inequality and mass poverty could lead to a collapse of social orders worldwide. Their main policy recommendations pointed towards equitable trade agreements, increased development assistance and price stabilization in major commodity markets on which the trade revenues of Southern economies critically depended. These recommendations soon stumbled upon the neo-liberal reform agendas of the Reagan and Thatcher administrations in the early 1980s (Lees, 2021), but some of its bleak, albeit vaguely formulated, forecasts did materialize: large parts of Africa and Latin America were at the verge of their “lost decades,” an exceptionally long depression that nullified much of the mid-century gains in growth and poverty eradication (Bates et al., 2007).

While the process of South–South Divergence was already underway, it accelerated during the 1980s. A growing group of Asian economies embarked on a path of sustained income and welfare growth, which contrasted sharply to a large group of nations that remained trapped into long-term sequences of booms and busts, episodes of growth alternated with significant periods of stagnation or even shrinking (North et al., 2009; Broadberry & Wallis, 2017). As Figure 3 shows, South–South Divergence occurred at least since the end of World War II and the start of formal political decolonization in Asia and Africa.⁶ Meanwhile, income convergence dominated in the global north.⁷ In addition to growing income gaps, South–South Divergence was also evident in other structural factors of economic development: a growing group of Asian economies developed more diversified production structures and export packages; more balanced sectoral distribution of labor productivity growth; higher rates of technology adoption and diffusion; more rapid accumulation of human capital complementing frontier technologies; and more efficient legal, fiscal, and administrative regimes. The seven decades of divergence since 1950 thus deprived the global south of its popular universalistic connotation as the “underdeveloped,” “impoverished,” “dependent,” or “pre-industrial” part of the world. Indeed, it is valid to ask whether the “global south” is still a relevant economic or political category today (Lees, 2021, 93; see also Johnson & Papageorgiou, 2020), an issue I will return to in Section 4. For now, I will only maintain that this category has significant historical relevance.

FIGURE 3

Coefficient of variation of per capita GDP in the Global North and global South, 1950–2018. Source: GDP per capita from Maddison Project Database (2020). Note: The coefficients of variation in GDP per capita are taken from a constant sample of 40 Northern and 105 Southern countries listed in Appendix 1. The global south excludes the oil-rich Gulf countries; the Global North excludes the former Soviet states that gained independence in the 1990s. [Colour figure can be viewed at wileyonlinelibrary.com]

Why should economic historians care about studying the nature and determinants of South–South Divergence more than they have done so far? Let me offer five interrelated arguments. First, as discussed above, the South–South Divergence agenda builds upon and extends the Great Divergence literature. As the contours of the Eurasian, and partly also the Global Great Divergence are fading, they are being replaced by new spatial inequalities which are worth exploring in their own right (cf., Milanovic, 2016). As Figure 3 shows, these new inequalities are to be found within the global South, which is adding a new chapter to world histories of inequality that economic historians may care to introduce their students to. This chapter opens up new questions. Instead of exploring the origins of the Industrial Revolution and the birth of Modern Economic Growth, these questions focus on the uneven spread of it across the global south. Such questions about the nature and drivers of catch-up growth have been addressed by development economists and political scientists, but economic historians have not intensively participated in their discourse. Consequently, leading concepts and theories of comparative underdevelopment remain characterized by short-termism, static concepts of “initial conditions,” shallow understandings of colonial legacies, missed opportunities of diachronic comparison, and a dearth of transnational historical perspectives. Ample scope for economic historians to contribute.

Third, the South–South Divergence agenda will engage economic historians more explicitly with the histories and economic realities of a growing majority of the world population. Whereas the global south today already comprises more than 80% of the world population and nearly 60% of the world GDP, its demographic and economic weight is bound to increase further during the remainder of the 21st century. Table 1 shows the distribution of the world population since 1820 up to the end of the 21st century, using UNDP population projections. Whereas the North's share in world population has steadily risen during the 19^th century to a peak of 36% around 1900, it began to decline soon after the turn of the century. While still one-third in 1950, it fell back to one-fifth in 2020, and by 2100, the North is projected to hold just 12% of the world population. By that time, Asia and Africa together will harbor more than 80% of the world population. Within the global south, a dramatic shift in demographic composition is occurring as well. Asia's share in the world population is expected to decline from 60% in 2000 to 42% in 2100, whereas Africa's share is set to rise from 13 to 39%.⁸ Taking average GDP per capita growth rates of the past two decades in the global south (2.8%, excluding the Gulf) and the global north (1.8%, excluding former Soviet republics) to these population forecasts, the share of world GDP that will accrue to the south is projected to rise from 57% in 2020 to 72% in 2050. The elephant in the room grows bigger.

TABLE 1. Population and income shares per world region, 1820–2100.

	1820	1850	1900	1950	2000	2050	2100
Population
Asia	0.66	0.62	0.53	0.54	0.60	0.53	0.43
Africa	0.06	0.06	0.06	0.09	0.14	0.26	0.39
Europe	0.22	0.24	0.28	0.23	0.12	0.08	0.06
Americas	0.03	0.05	0.10	0.14	0.14	0.13	0.11
Global south	0.74	0.71	0.64	0.67	0.80	0.86	0.88
Global north	0.26	0.29	0.36	0.33	0.20	0.14	0.12
GDP
Global south	0.58	0.46	0.28	0.26	0.43	0.72	…
Global north	0.42	0.54	0.72	0.74	0.57	0.29	…

• Note: Europe includes all former Soviet republics and Central Asian states; Asia includes New Zealand and Australia.
• Source: 1820–1900 from Maddison Project Database (2020); 1950–2100 from UNDP, World Population Prospects, 2022, revision, medium variant.

A fourth reason follows by extension. The reconfiguration of global economic gravity as outlined in Table 1 cannot but have profound implications for international economic and political relations in terms of divisions of labor, food demand, capital flows, trade, investment, and migration patterns. In fact, one of the most important consequences of South–South Divergence has already materialized. Extreme poverty had long been a predominantly Asian phenomenon. Back in 1990, more than four out of five of the world's extreme poor were living in Asia. Just three decades later, in 2020, two out of three of the world's extreme poor lived in Africa (ca., 65%). In Asia, relative and absolute poverty rates have both been declining, while in Africa, the gradual decline in relative poverty rates combines with a rise in absolute numbers.⁹ Finally, fifth, in a field that has long been dominated by Western-centered research agendas and north–south perspectives, more systematic engagement with south–south comparisons can inspire new data collection efforts and help stimulate the development of reciprocal comparative frames that do not take Western economic development as the mirror image.

In arguing that south–south comparisons have received unwarranted little attention by economic historians I do not mean to say that the economic history of the global south as such has been neglected. To the contrary. Especially in the past two decades, there has been a notable shift towards the construction of historical narratives of global economic development in which the south has received due attention (e.g., Austin & Sugihara, 2013; O'Rourke & Williamson, 2017; Riello & Roy, 2019).¹⁰ New networks, organizations, and conferences have emerged to promote the study of African, Asian, or Latin American economic history as a spatially defined substrand of the field.¹¹ I contend though that these new research networks and agendas have remained rather inward-focused, dealing with region-specific questions and topics (e.g., Latin American inequality, African colonial legacies, Asian industrialization, Middle Eastern Islam) and that research approaches are predominantly embedded in prevalent imperial or north–south perspectives. Seldom do these communities venture out to discuss the nature and drivers of cross-regional divergence in long-term economic development. Benchmark estimates of output, living standards, inequality, education rates, or what have you, and the intellectual endeavor to make sense of historical processes of divergence remain primarily constructed in reference to North Atlantic or Western standards. Research that has explicitly addressed questions of South–South Divergence, and in which the mirror is placed in the Indian Ocean, the Southern Atlantic, the Pacific, or the South Chinese Sea has barely scratched the surface (e.g., Andersson & Axelsson, 2016, Frankema and Booth 2020, Otsuka & Sugihara, 2019).

4 WHAT BINDS THE GLOBAL SOUTH AS IT GROWS APART?

To be fair, the Brandt report of 1980 did signal changes in Asia's economic outlook, which up to the 1970s had been assessed in rather pessimistic and particularly Malthusian terms (Wittfogel, 1957; Geertz, 1963; Ehrlich, 1968; Myrdal, 1968). The report alludes to incipient industrialization in Malaysia, South Korea, Taiwan, Singapore, and Hong Kong and the promises of a “Green Revolution” in India and the Philippines (Brandt et al. 1980, p. 52–53). Yet, the authors also cautioned against taking these developments as evidence of sustained poverty eradication. The full recognition that several Asian economies had embarked on a trajectory of catch-up growth came in the early 1990s with the publication of the World Bank's The East Asian Miracle (1993). According to the Bank, the exceptional rates of Asian economic growth that were recorded over the past three decades were not all that surprising in light of the “pragmatic” and “diverse” public policies that had promoted impressive rates of physical and human capital accumulation in a context of macro-economic stability (World Bank, 1993, p. 25–26). More surprising, according to the Bank, was that the eight East Asian Tigers had managed to achieve rapid growth with equity. The World Bank turned its long cherished neo-liberal dogma of free markets, free trade, privatization, and noninterventionist government on its head by emphasizing how public policies, market interventions, and state-led institutions were key in bringing about this miraculous combination (p. 5–6). The Bank insisted, however, that the East Asian example should not be taken as an excuse to abandon market-oriented reforms in many other Southern economies grappling with structural macro-economic weaknesses and unsustainable debt ratios (p. 26).

What, then, binds the south together besides its juxtaposition to the global north? The answer to this question is not obvious. After all, the global south was made up of an extremely diverse mix of older and younger states (cf., China vs. Zaïre); of larger and smaller economies (cf., Indonesia vs. Mauritius); of labor-abundant and labor-scarce areas (cf., India vs. Brazil), of rice cultivators and meat and wheat producers (cf., Vietnam vs. Argentina). Moreover, all these polities had rather distinct “national” histories, including very different relations with the rest of the world (cf., Turkey vs. Haiti). And despite the economic ascendance of Asia, within this vast region, there are also large differences. In contrast to the Tiger economies, countries like Pakistan and the Philippines failed to sustain earlier episodes of growth, while Myanmar or Sri Lanka now cope with deep and prolonged depressions which appear, at first sight, to be more in line with African or Latin American experiences. Overseeing this patchwork of development paths, how much scope is there really to make balanced, reciprocal comparisons of the kind that have been advocated for the Eurasian Divergence? Is the south a useful analytical category for comparative economic history to begin with?

The standard response to this question is that the economies of the global South were, at least during the second half of the 20th century, unanimously confronted with the challenge of “late” development. All Southern economies operate(d) at a considerable distance from the global technology frontier, and thus face(d) a particular challenge of incorporating foreign knowledge and technologies to realize catch-up growth. In view of this challenge, the literature on state-directed development argues that “late” industrialization requires a sufficiently powerful state capable of simultaneously addressing the problems of market failure and underinvestment in essential public goods (e.g., health, education, infrastructure) through institutional reform. The economics of “backwardness,” as Gerschenkron (1962) put it, are ultimately about lifting the very constraints that cause retardation in the first place. The “developmental state,” a term coined by Johnson (1982) in reference to the late industrializing economies of Northeast Asia (Japan, South Korea, Taiwan), is a state capable of lifting these constraints. Developmental states do so by mobilizing productive resources and aligning the interests of politicians, bureaucrats, workers, entrepreneurs, and investors to cooperate in long-term development plans, including the sacrifices that such programs entail for groups who have to forego part of their stake for the common cause. Such cohesive policies may, for instance, be politically justified by a nationalist agenda to keep Western or Chinese imperialism at bay (Johnson, 1982; Amsden, 1989; Woo-Cumings, 1999).

Leading development economists of the 1940s to 1970s, including Nobel laureates Simon Kuznets and Gunnar Myrdal, reasoned about theories of a “big push” and a “hard state” that can override conservative powers and promote social reform to lift the economy (Rosenstein-Rodan, 1943; Myrdal, 1968; Kuznets, 1973). In his analysis of the Russian economy, Gerschenkron (1962) focused on the role of the state in mobilizing resources for the development of transport infrastructure and a financial sector capable of funding industrialization in a context of continuously increasing minimum efficient scales. Johnson's (1982) study of the Japanese “miracle” focused on the role of MITI in industrial policy-making. Johnson considered Japan to be an example of a “plan-rational capitalist” developmental state, which differed from the “plan-irrational” Soviet state in that it combined private ownership with state management of enterprises, while in the latter, the state-controlled both.¹² From these archetypes of state interventionism (communist plan economy, capitalist plan economy, liberal market economy), the literature on state-directed development fanned out to incorporate many varieties, of which South Korea is often hailed as the most prominent success case (Amsden, 1989; Kohli, 2004; Freund, 2019).

And yet, the challenge to catch-up with the technology and productivity leaders of the day was, and is, not an exclusive feature of Southern economies. Among the early industrializing nations in the north only a handful may have been actively pushing segments of the technology frontier outwards, the great majority were technology-adopters following, or borrowing from, their neighbors as well. Concerns of catch-up growth in the wake of fierce economic and military competition were as prominent in the United States, Europe, and Japan as they are in present-day China or Ethiopia. Confronted with British-free trade doctrines and the problem of British manufactures flooding continental European markets, Friedrich List, a founding father of the German historical school, pointed to the need of industrialization under state coordination, calling for protectionist policies and the establishment of a German customs union.¹³ So if the distinction between the north and the south is not primarily about invention versus adoption and implementation, what is it?

On closer inspection, I believe that the defining features of the south consist of a combination of four elements, which I will call the quadruple challenge. First, by the mid-20th century, the distance of Southern economies towards the global productivity and technology frontiers was of unprecedented magnitude. This made the design of growth-promoting policies for late industrializers arguably more complex than they were for earlier industrializers. The size of the gap also justifies the question whether for some the gap has simply become too large to bridge (Austin, 2016; see also Section 7).

Second, as a result of the widening gap in industrial and military capacity, all countries in the south were confronted with at least the threat, if not the praxis of imperial domination. Forms of imperialism varied from colonial occupation and settlement to foreign control of domestic capital markets, or interference with specific parts of domestic or foreign policy in the context of gunboat diplomacy. Imperialism was mostly, albeit not exclusively, exercised by a handful of Western metropoles. In its diverse appearances, imperialism and the lived experiences of bounded autonomy shaped the political economic contexts in which Southern states developed their “national” identities and the economic policies that were to back up national sovereignty.

Third, the intertwined objectives of economic prosperity and political autonomy, as for instance articulated by the nonalignment movement during the Bandung conference in 1955, were to be realized in a progressively globalizing world order, in which domestic markets and producers were no longer shielded from foreign competition through natural isolation, only through deliberate protectionist policies, which also gave way under pressure of the increasingly influential Washington consensus.

And finally, fourth, the challenge of closing the productivity gap took (and increasingly takes) place in a context where vast open land and resource frontiers that had attracted so many imperial investors, have been closing rapidly (Barbier, 2010). These closing land and resource frontiers also mitigated the mobility of labor, which became increasingly state-controlled over the course of the 20th century (McKeown, 2008). There are strong indications that resource depletion (e.g., soil erosion), natural hazards (e.g., climate shocks), and ecological degradation (e.g., biodiversity loss, deforestation) hit the global South harder than the global north, but it remains an open question to what extent the global distribution of long-run growth-impeding effects of climate change, biodiversity loss and other forms of environmental degradation compound the South–South Divergence.

Taken together, the simultaneous experience of imposed external technologies and knowledge, the increasing weight of world markets, the limited state autonomy, and the tangible impact of heightened environmental pressures offer coherence to the global south as an analytical category in its own right. This does not mean that the boundaries of the concept are watertight, they are not. The Japanese experience illustrates how a state threatened by foreign power was pressed to forge internal political coalitions to prevent imperial intrusion, and in the process became a major industrial and imperial power itself. A similar argument can be made for the United States.¹⁴ Neither is there a good reason why Albania or Moldavia would be part of the global north. But when we allow for some hybridity, there is a good case to consider the south as a vast heterogenous space that offers scope for systematic comparative research, as the great majority of societies did and does confront the quadruple challenge in one way or another.

5 WHY IS STATE-LED DEVELOPMENT SO HARD TO ACHIEVE?

If we are to accept these four defining elements, what are then the first-order questions of the South–South Divergence? In what follows I propose three overarching questions that are central to any vision on the origins and nature of South–South Divergence. These questions are not exhaustive, they just serve to start the dialog.

The first question departs from the observation that nearly all post-colonial states in the global south have, at some point in time, formulated policy programs to grow and diversify their economies to enhance welfare levels. So far, only a significant minority can claim true success in achieving this ambition. How do these programs compare, and why did state-led development translate into sustained growth and welfare development in North- and Southeast Asia, while they mostly failed in Africa, Latin America, and large parts of the Middle East? Put differently, what explains the limited replicability of the East Asian developmental state?

The literature on state-led development offers roughly three lines of explanation. First, the Weberian notion of state capacity, that is, the ability of the state to mobilize resources to secure a monopoly on large-scale violence, to gather and administer information for taxation and policy making, and to establish legal structures that protect investors, correct market failures, and secure transparency in financial markets (Tilly, 1990; Besley & Persson, 2009). In the absence of a sufficiently trained, professional bureaucracy or a trustworthy legal system it is hard for governments to navigate the ever-present conflicts of interests among societal stakeholders in pursuing its development agenda.

In his comparative analysis of the emergence of state-sponsored computer industries in Brazil, India, and South Korea, Evans (1995) argues that the critical success factor is a well-trained, competent bureaucracy with a professional morale capable of collecting information, interpreting societal stakes, weighing conflicting interests and explaining policy decisions. Evans’ notion of “embedded autonomy” refers to the paradox inherent to state-led industrial policies. To be effective, the state needs to act as an internally coherent organization with sufficient autonomy vis-a-vis societal interest groups, while at the same it needs to have close links with, and be firmly embedded in, different segments of society to collect information and decide upon the right type of market interventions.

A second line of explanation concerns the political economy of state-directed development. How do governments align the interests of diverse societal bodies or private trust networks, and incentivize them to bring part of their resources (human, financial, intellectual) under the umbrella of the state to pursue long-term development schemes? (cf., Levi, 1988; Tilly, 2004). How do states prevent partisan politics if those who make decisions about resource mobilization themselves have stakes in specific economic activities? What makes that some states manage to successfully mediate the conflicts of interest between capital and labor, between farmers and industrialists, between ethnic or religious factions, between old and young, while others collapse under their weight?

In his comparison of the industrialization trajectories of South Korea, India, Brazil, and Nigeria, Kohli (2004) emphasizes the role of coercive state power. He argues that bureaucrats in India and Brazil were as competent and well-informed as in South Korea, but that the Korean model of a cohesive-capitalist state was most conducive to late industrialization. The uncomfortable lesson from Kohli's State-Directed Development is that major industrial pushes have largely occurred under strong authoritarian regimes, precisely because cohesive power is needed to suppress labor demands and sacrifice human freedoms to achieve long-run welfare growth. Kohli denounces the idea that Asian industrializing economies were successful because of any exceptional, culturally conditioned labor discipline, or because of exceptional administrative professionalism. In his view, state repression was decisive in aligning diverging societal interests behind a state-directed development agenda.

Whether successful long-term economic planning necessarily requires state repression remains an open empirical question to which economic historians can contribute. Historical comparisons are helpful to tease out what forms of state regulation resulted in sustained economic growth. In line with Chang (2002), Kohli argues that state-directed economic protectionism was one of the critical success conditions. Since “sovereign” post-colonial states have often not been in the position to set their trade and monetary policies autonomously, historical trajectories of decolonization are important to understand with respect to questions of autonomy and economic policy.

Third, economic policies will always have to be aligned with existing fiscal and financial systems and the institutions that regulate markets. There are plenty of historical and contemporary examples of well-intended developmental programs that have failed because they overlooked critical societal realities and, unintentionally, deepened market failures rather than relieving them (Scott, 1988). Such failures also include parts of the recommendations on structural adjustment provided by international organizations such as the IMF and World Bank, organizations that do not always accurately foresee the health effects of the medicines they prescribe (Stiglitz, 2002). With regards to the challenge of late industrialization, a trade policy that allows countries to protect and nurture their emerging industries appears to be key, but the optimization problem involved in pairing effective protection with market discipline differs from industry to industry, and time to time.

Economic historians have not been absent from these discussions altogether, but compared to the input of evolutionary economists, development economists, and political scientists their contributions have remained sparse. This is unfortunate because theoretical and conceptual studies of state capacity building and state-directed development suffer from several drawbacks. One of these drawbacks is their strong focus on the design and praxis of national politics and the role of national bureaucracies in economic policy making, which emphasizes inward-looking perspectives on the sources of industrial transformation (Pempel, 1999). Colonial legacies are mostly treated as a set of “initial conditions” of national policies (cf., Kohli, 2004), overlooking the fact that colonial rule itself was subject to continuous change, and has had multifaceted effects on long-run economic development. Instead of molding colonial histories into a set of fixed initial conditions, they are better conceptualized as a co-determinant of long-term change processes. Colonialism did not only bring new elements into local and regional economic systems, the particular intertwining of the local with the colonial also established a new dynamic.

Take, for example, the theory on the origins of state capacity by Besley and Persson (2009, p. 1218) which, in their words, offers “a framework where ʻpolicy choicesʼ in market regulation and taxation are constrained by past investments in legal and fiscal capacity.” Their theory departs from the notion that states evolve through distinct feedback loops between taxation and property rights protection. Inspired by political science literature on European state formation (cf., Tilly, 1990), their theory stipulates how these feedback mechanisms are affected by the threat of external conflict, that is, interstate warfare. However, fiscal capacity building in large parts of the world had little to do with international warfare. The legal and fiscal systems in the majority of present-day nations in the global south were shaped in a colonial context. Instead of the threat of conflict between sovereign powers, it was the ambition to build fiscally autonomous colonial states that would unburden metropolitan tax-payers, which determined legal reforms and market regulations. The same goes for legal capacity building. This was not so much the outcome of the investments that colonial governments were willing to make in legal institutions, but rather a function of the specific integration of local, indigenous systems of law in colonial legal frameworks imposed by the metropole. Theories of the origins or development of state capacity in the global south will have to take these notions of bounded sovereignty, and local-colonial entanglements into account (Frankema & van Waijenburg, 2021).

Another example is the bi-modal concept of open and limited access orders developed by North et al. (2009), which focuses on the political economy of long-run institutional development. North et al. study the problem of how social orders (a broader concept than the state) respond to the problem of arbitrary violence under anarchy. The default solution, which they refer to as a “limited access order,” is one in which access to markets and the political arena are limited by a dominant coalition of elite factions, who redistribute resources via systems of patronage and cooptation. The alternative is an “open access order,” in which control over police and military forces (i.e., the violence monopoly) is placed in hands of parliaments, while access to political power and economic markets is, at least de jure, open to all (adult) members of society. In open-access orders, meritocratic values and a well-developed civil society support creative talent and innovation as drivers of welfare growth. By suppressing individual freedoms and civil society, limited access orders waste part of their human and collective resources. Following their model, North et al. (2009) predict that limited access orders in Asia (e.g., China) will be unable to sustain economic development once the potential of catch-up growth based on the import of foreign technology has been realized: developmental states that seek to sustain growth, require an ingrained capacity to change their political foundations from autocratic to democratic, and from limited to liberal forms of capitalism.

Whether liberal democracy is a conditio sine qua non for sustained economic development presents itself as an important open question for South–South Divergence debates, but the point here is that the social orders framework is based on the assumption of a closed economy and a sovereign polity (North et al., 2009).¹⁵ But do transitions from a limited to an open-access order ever take place in a closed economy? How often do elite factions, or societal interest groups, who oppose one another, operate without external support? For the late industrializers of the global south who seek to catch-up in a hyper-globalized, post-colonial world order, the role of international and transnational forces seems impossible to assume away without losing ample explanatory power. All of this is to say that global economic historians are well-placed to bring transnational perspectives and dynamic conceptions of colonial history into theories of the political economy of late development.

6 WHY IS DEVELOPMENT CLUSTERED?

A second question—which also sits uneasy with closed-economy perspectives—is why growth and stagnation appear to occur in regional clusters? If sustained growth would be more or less randomly distributed across the globe, terms like the “global South,” the “industrial core” or “Asia's economic renaissance” would be pointless. However, clustered development can be observed in time and space. Figure 4 presents one of several ways to illustrate this phenomenon. It plots levels of GDP per capita in 1960 to GDP per capita growth rates for 1960–2018 across 105 countries in the global south. The black lines indicate the averages. The graph shows that Sub-Saharan African, Latin American, and Asian economies are grouped together: Most sub-Saharan economies were poorer in 1960 and grew slower than average. Asian economies were also poorer than average, but grew faster. Latin American economies were clearly richer than the average in 1960, but their growth rates fell short. Only among the economies in the Middle East and North Africa (MENA), a pattern of clustering was lacking. Regional clustering of long-term economic development is a key feature of global processes of economic divergence, and also underpins the widening income gaps across the global south that are shown in Figure 1.

FIGURE 4

GDP per capita in 1960 (x-axis) versus GDP pc growth 1960–2018 (y-axis) in the Global South. Source: Maddison Project Database (2020). [Colour figure can be viewed at wileyonlinelibrary.com]

Clustering is also reflected in a range of structural characteristics of long-term economic development. Asian rates of growth after 1960 were not necessarily higher, but definitely more stable than in Africa or Latin America, which were subject to alternating periods of economic booms and extended episodes of economic depression (Broadberry & Gardner, 2016; Broadberry & Wallis, 2017). Table 2 shows the average number of years of the longest period of economic depression of countries in Southeast Asia, Latin America, and sub-Sahara Africa between 1950 and 2018. The end of the depression is defined as the year in which income levels exceed the previous peak year. The table also shows the three countries within each region with the longest time interval. The inter-regional differences are huge. The Southeast Asian average is one decade of zero net growth, the Latin American average is two decades, and in sub-Sahara Africa, the average is a staggering 29 years. Haiti constitutes one of the most dramatic examples of long-term stagnation, as the per capita income levels of 1950 were never surpassed in seven ensuing decades. In Southeast Asia, the Vietnam war caused 20 years of stagnation. In sub-Saharan Africa, the CAR, Niger, and Senegal have all recorded more than 50 years of zero net growth, heading a group of 22 countries that have experienced stagnation for more than three decades in a row.

TABLE 2. Longest consecutive period without net GDP per capita growth, 1950–2018.

	Years	Era
Southeast Asian average	10
Vietnam	20	1964–1984
Philippines	14	1982–1996
Cambodia	13	1986–1999
Latin American average	20
Haiti	68	1950–2018
Nicaragua	44	1974–2018
Argentina	29	1974–2003
Sub-Sahara African average	29
Central African Rep.	57	1961–2018
Niger	53	1965–2018
Senegal	51	1965–2016

• Source: GDP per capita from Maddison Project Database (2020).

To be sure, Asian economies suffered prolonged depressions before 1960 too, especially in the 1930s and 1940s when the Great Depression was followed by the ravages of the Second World War and, in many cases, followed again by large-scale violence in the wake of political decolonization. But this further underscores the importance of the historical juncture that sustained growth has created in large parts of Asia. Indeed, in a globalizing world economy, booms and busts appear like tidal waves, some of which hit all major world regions simultaneously (e.g., the Great Depression), while others affect just one, or a few regions (e.g., the “lost decades” of 1973–1995). Limited economic diversification is one of the factors associated with boom-bust growth. Undiversified export portfolios put economies at risk of world market price volatility and sudden worsening terms of trade. Indices of export diversification reveal that African, Latin American, and Middle Eastern economies rank at the bottom of the global distribution, while many Southeast Asian economies that started out with comparably undiversified, resource-based export packages, have managed to diversify (Deaton, 1999; Hidalgo & Hausmann, 2009; Bértola, 2016). This is one of the reasons why the debt crisis and related depression hit the economies of Africa and Latin America particularly hard, while this storm largely bypassed Asia.

But why are such structural economic features clustered? To explore this question, the metaphor of family resemblances can be helpful. This metaphor offers an intuitive basis for a theory about the role of regional proximity and identity. Family resemblances can be the result of nature (i.e., shared genes, traits) or nurture (i.e., exposure to intra-familial norms, values, habits, routines as well as competition). To study the drivers of East Asia's economic miracle, one can either emphasize the innate personality traits of the East Asian family such as shared favorable agricultural conditions or a shared predisposition to labor discipline or coercive state power. Alternatively, one can try to understand the miracle in terms of intrafamilial exchange: exposure to, and competition (or threats) from a few states setting the example (e.g., Japan, Singapore), consciously or unconsciously incentivizing other family members to emulate. The concept of family resemblances can also serve as a starting point to explore why so many African economies jointly, and simultaneously, experienced their “lost decades”; or why Latin American states underwent such dramatic political swings between democracy and autocracy in the context of exceptional levels of economic and social inequality (Sokoloff & Engerman, 2000; North et al., 2000). Of course, not every family member necessarily shares all family traits. There is always scope for outliers (cf., Mahoney, 2010). When major international trade blocs (e.g., the EU, ASEAN, Mercosur, or ECOWAS) self-identify as a family of states, they emphasize their commonalities and express difficulties with members who do not abide to the rules, norms, or values of the family (e.g., Mali, Hungary, Myanmar). Members may also withdraw if they feel they are not part of the family (e.g., Brexit), while states that choose isolation (e.g., North Korea) may not share the development patterns of the family at all.

Interestingly, a more fundamental nature-nurture debate has never taken-off in the economic history of developing regions. There is a small literature on Asia that is looking into the role of Japan and China in shaping the emergence of an Asian international economic order (Sugihara, 2005, see also Otsuka & Sugihara, 2019; Latham & Kawakatsu, 1994), inspired by the “flying geese” model formulated by Akamatsu (1962), in which the international division of labor within Eastern Asia is regarded as an outcome of changing structures of comparative advantage: rising labor costs in the more advanced Asian economies incentivized firms to outsource their labor-intensive production processes to less advanced economies with lower labor costs. Cultural, linguistic, and geographic proximity play a central role in their location decisions.

A next step to understand the phenomenon of clustering would be to systematically compare histories of regional (dis)integration. Two regions that lend themselves very well for such an endeavor are Southeast Asia and sub-Saharan Africa. Both regions were historically lightly populated. Both specialized in the production and export of primary commodities, tropical cash-crops in particular. Both were colonized by a handful of Western powers who either imposed, or further deepened existing specialization patterns in primary commodity production. Both experienced mass migrations of labor under colonial rule. Both underwent extremely violent processes of decolonization. Up to 1970, both regions recorded comparably low levels of per capita GDP (although we should not exclude the possibility of structural mismeasurement here), and considerable parts of their populations were living in extreme poverty. Yet, in the past 50 years, these regions rapidly grew apart, a divide that hardly any expert had foreseen in the 1960s.

Standard accounts of the post-1970 divergence point to the successes of the Green Revolution in Southeast Asia and lagging agrarian productivity growth in sub-Saharan Africa (Hayami et al., 1998; Otsuka et al., 2009). Related to this, scholars have pointed to the urban bias in African economic policies versus the rural orientation of Southeast Asian policies (Lipton, 1977; Henley, 2012; Berendsen et al., 2013). In their comparison of two major oil-producing economies, Indonesia and Nigeria, Bevan et al. (1999) suggested that the divergence was primarily due to targeted poverty-reduction policies in the former, and their absence in the latter. They see no evidence that the divergence is rooted deeper in history (Bevan et al., 1999, p. 1–6). More recently, Henley (2015) has re-instated this point, claiming that the Asia-Africa divergence largely boils down to distinctly different post-colonial policy intentions.

But why would Southeast Asian societies be “blessed” with enlightened autocratic regimes, while African societies were suffering from short-run oriented kleptocrats? Is it a structural difference in the “nature” of political culture that explains regional clusters of well- and ill-intended development policy? Deeper historical roots of the divergence may surface when comparing the evolution of intraregional grain or textile trades, labor migration, and concomitant capital flows and business networks, that is, when looking at the wider regional contexts and dynamics in which national and colonial states were embedded. One of the central hypotheses that warrant deeper investigation is the specific interplay between colonial policies and local economic development, and how this has led to a disruption and transformation of intraregional exchange networks in sub-Saharan Africa, whereas it may have strengthened existing exchange networks in Southeast Asia, networks that were not completely dismantled by the violence and instability of the early post-colonial era. Is sub-Saharan Africa still recovering from, and more frequently thrown back into, episodes of large-scale instability that are more likely to emerge in a disintegrated region? Such questions about colonial legacies and processes of regional (dis)integration can be fruitfully addressed in south–south comparisons.

7 CAN THE WHOLE WORLD BE DEVELOPED?

In 1981, Richard Easterlin published an article in the Journal of Economic History with the captivating title: Why isn't the whole world developed? The question seems to suggest that a fully “developed” world is a realistic possibility. But is it? The third question I propose builds on the two preceding ones but is probably the most fundamental: can the whole world be developed? Let us, for the sake of the argument, define “developed” here as the sustained eradication of all forms of poverty.

An influential segment of the Great Divergence literature argues that the great enrichment of the West has at least partly taken place at the expense of the growth opportunities of the global South, and especially through the way the “periphery” has been integrated into the emerging world system (Wallerstein, 1974). The idea that global economic development was a zero-sum game, and that mounting inequality was an inextricable feature of the modern world system has at least three dimensions. First, the British Industrial Revolution would not have been feasible without the feedback loops that came from American plantation slavery and the structural repression of African labor associated with it (e.g., cotton, sugar, investment capital) (Williams, 2022 [1944], Inikori, 2002; Beckert, 2015; Berg & Hudson, 2023). Second, accelerated globalization in the 19th and 20th centuries has led to de-industrialization in the global South (Inikori, 2009; Williamson, 2011; Parthasarathi, 2011). This idea builds on neo-classical theories of comparative advantage and dependista arguments about a structural decline and high volatility of the terms-of-trade of primary commodity exporters in the periphery (Prebish, 1950; Deaton, 1999; Williamson, 2012). Third, a global “reversal of fortune” has taken place which has been deepened by colonial extractive institutions with long-lasting effects on the global income distribution (Rodney, 1972; Acemoglu et al., 2001; 2002).

The impressive catch-up of Asian economies, including many former colonies, defies overly deterministic accounts of zero-sum game globalization and long-run colonial legacies. It shows that world economic orders and related power relations are not set in stone. But it remains an open question to what extent, and how, the rapid economic ascent of one world region pre-empts the rise of another in a hyper-globalized world order? Following the South–South Divergence frame, another way of phrasing the question is: does the rise of Asia open or close windows of opportunity for Africa, or middle-income countries in Latin America and the Middle East? Let me confine the discussion to three strains of thought.

First, Asia's expanding share in global trade stemmed from cost advantages in the production of agricultural commodities, industrial manufactures, and supply of international services. Japan, China, and other Asian NIEs have found ways to shift these advantages from labor-intensive to more capital-intensive products (Austin & Sugihara, 2013) while keeping costs and prices below those of vanishing manufacturing industries in Europe and North America. In the massive relocation of production sites towards Eastern and Southern Asia, comparative labor costs play a key role. Will this shift repeat itself, for instance towards Africa, or was the Eastward shift a once-and-for all “race to the bottom”?

In the article Africa Rising? I have argued with van Waijenburg (2018) that a new shift out of Asia is unlikely to happen soon because the wage gaps between Asian NIEs and the poorest African economies are too small to overcome existing productivity gaps. The wage gaps between Western Europe and Japan around 1900, when Japanese textiles began to make an inroad on a global scale, were of unprecedented magnitude (2018, p. 561–562). In fact, wages in most of Africa's low-income countries are much higher than one would expect on the basis of per capita GDP estimates, being pushed up by relatively high prices of food and urban housing. Gelb & Diofasi (2015, p. 18) estimate that sub-Saharan African countries on average have a more than 15% higher PPP exchange rate than warranted based on their comparative income levels, which rises to 32% for IDA eligible countries. Allen (2017) also shows that urban food prices in West Africa are significantly higher than expected given current GDP per capita levels.¹⁶ As long as the price and wage gaps with Asian NIEs remain small, the opportunities for African economies to exploit a labor cost advantage in global manufacturing or (mobile) service markets will be limited.

A second consideration is that in a world of rapidly dwindling or degenerating natural resources, the political incentives to seek rents from mining monopolies will become even stronger than they already are. Figure 5 shows real composite indices of world market prices for manufactures, minerals (including oil), and tropical export crops (see for the commodities included the note below the figure). These series show that since the beginning of the 20th century, and especially since the 1970s, the terms of trade of mineral exporters have developed favorably. These trends partly explain why so many African countries have switched to becoming major mineral exporters. Tropical cash-crops, of which many are produced by smallholder farmers, have seen a much more pronounced and prolonged drop in prices, also compared to manufactures (a sector in which labor productivity gains were much larger). Countries that have specialized in mineral exports have grown faster, but this specialization pattern has compromised export diversification and entrenched boom-bust growth due to high price volatility and other features of the resource “curse” (Auty, 1993; van der Ploeg, 2011; Williamson, 2012).

FIGURE 5

Real composite world market price indices of tropical cash-crops, minerals, and manufactures, (1900 = 100). Source: author's own. Notes: The tropical cash-crops index is an unweighted composite of prices of palm oil, cotton, coffee, and sugar; The minerals index is a composite of prices of gold, copper, and petroleum; The manufacturing price index is a composite taken from Grilli and Yang (1988) for 1900–1960, and from the World Bank Commodity Price Database (The Pink Sheet; version March 02, 2022) for 1960–2021. [Colour figure can be viewed at wileyonlinelibrary.com]

Closing resource and land frontiers also affect labor migration, which can be an alternative way to exploit labor cost advantages. In the long period of European industrialization, there were ample opportunities for low-skilled workers to cross the Atlantic in search of work and better living conditions (O'Rourke & Williamson, 1999). China and India also were major net-emigrant countries in times that their agrarian economies were barely capable of sustaining all. Millions of migrants moved into the lower populated areas of Southeast and Northeast Asia (McKeown, 2008). However, similar population valves do not exist for Africans, where the population is projected to grow from ca 1.5 to 2.5 billion in the next three decades, or for Latin Americans who seek work further North. Territorial border protection and exhausted natural resource frontiers are setting this age apart from the long 19th century in which the early industrializing economies began to eradicate extreme poverty. Add to this the uncertainties and inequality-enhancing effects of climate change, especially on agricultural production and employment possibilities, and the question whether the whole world can be developed becomes even more pressing.

Third, the digital revolution has created a worldwide infrastructure for the rapid dissemination and sharing of knowledge, but it has thus far not resulted in more egalitarian distributions of human capital. Even though an instant leap towards the global technology frontier is not a precondition for the eradication of hunger and poverty, the use of some high-end technologies seems crucial for sustained development and export diversification. The production systems of the future will require a specific bundling of human expertise and AI. It is thus valid to ask whether the least developed economies will ever be able to close that gap (Austin, 2016), and whether economies would need a minimum scale to do so (Chaudhary et al., 2012). Tacit monopolies in specific technology-skill complexes and economic size will thus have to be factored into historical theories of “late” development.

This brings me to a final methodological point: diachronic comparative frames can help to explore changing opportunities and constraints to catch-up growth. It is surprising how little use is made of this analytical lens. One of the rare but better-known examples of a diachronic south–south comparison in economic history is the study by Bates et al. (2007) that contrasts the early post-colonial “lost decades” in 19th century Latin America with the “lost decades” of Africa in the late 20th century.¹⁷ The authors argue that a long phase of economic stagnation in Latin America associated with the political instability and violence caused by processes of decolonization, was eventually concluded by a golden age of commodity-export-led growth. They ask whether such a pattern of change could apply to Africa as well. While the similarities in the process of decolonization may not be so evident as Bates et al. suggest (cf., Prados de la Escosura, 2009; Travieso, 2020), the diachronic comparison helps to sharpen questions about shifts towards new political economic equilibria, and how such shifts occur in specific global economic and political contexts. For this reason alone, this thought experiment would already deserve a follow-up. Moreover, many Asian countries also experienced eruptions of large-scale violence, hyper-inflation, and catastrophic famine, especially during the early post-colonial era (e.g., Indonesia, Malaysia, Cambodia, Vietnam, Burma). In the pursuit of a South–South Divergence agenda, systematic diachronic comparisons of such episodes of crisis and recovery have great potential to think more deeply about the conditions required for countries to achieve stable middle-income levels without poverty.

8 NEW COMPARATIVE HORIZONS IN GLOBAL ECONOMIC HISTORY

The Great Divergence debate has been an extremely fruitful conversation for the economic history community. It has given our discipline new impetus through the construction of major global databases, by the forging of new cooperations between Western and Asian scholars, by attracting students to exciting differences in opinion, and by rethinking the very practice of comparative historical research through the lens of reciprocity.

In this review article, I have argued that the South–South Divergence research agenda offers a promising extension to Great Divergence scholarship by opening up a set of important new questions and approaches that are tailored to studying the highly diverse development trajectories of “late” industrializers. Chief among these approaches are the adoption of long-run time horizons, transnational perspectives, and diachronic comparisons. These approaches are necessary to complement the obsessive quest for causal identification in economics and economic history, a trend evidenced by the increasing use of applied micro-economic techniques. These techniques can help break down parts of the puzzle, but they are of little avail in linking the pieces together in order to reconstruct the bigger picture. In these bigger pictures, “endogeneity” cannot be considered as a technical or methodological problem, but has to be taken as a central tenet of long-run processes of divergence and convergence.

I have proposed three leading questions to structure the South–South Divergence research agenda. These can be summarized as interrogating (1) the limited spread of the developmental state; (2) the remarkable spatial and temporal clustering of growth and development, and (3) the implications that the economic ascendance of one region may have for others aspiring the same. These questions are not exhaustive. To pursue them will require a rethinking of central theoretical concepts that have been designed to explain the economic development of early industrializers or state formation processes in sovereign (Western) countries. I have argued that such theories will have to speak to four conditions that apply to the great majority of present-day societies in the global south, namely that (1) distance to the technology frontier today means something different than a century ago; (2) post-colonial societies operate with varied degrees of economic and political autonomy; (3) “late” industrializers have to catch-up in a hyper-globalized world economy; and (4) “late” industrializers are confronted with tighter frontiers, especially with regards to the exploitation of natural resources, but also considering institutional constraints to human mobility. To this, we may add the increasing uncertainty of agricultural development as a result of accelerated climate change and biodiversity loss.

Finally, the South–South Divergence research agenda opens up two other opportunities to rejuvenate the field of economic history. First, in building on the tradition of Great Divergence scholarship, experts of the region are invited to leave their comfort zone and integrate their work in the larger frame of South–South comparisons. This will lead to new cooperations and places more emphasis on the critical importance of economic history capacity building across the global South. Second, there are ample opportunities to make economic history relevant for policy-makers and media interested in the rapidly changing world economic order of the 21st century. Topics abound, as they range from a newly unfolding debt crisis in Africa, to a historically-informed interpretation of the Chinese growth slowdown; from comparative processes of regional (dis)integration, to new ideas about degrowth and green growth. This essay is a plea to integrate the global south in the historiography of global economic development on its own terms, and to rethink, not for the first time, how new comparative horizons can be exploited to preserve economic history as a craft in its own right.

ACKNOWLEDGMENTS

For comments on earlier drafts of this paper, I thank Pim de Zwart, Sandra de Pleijt, Marlous van Waijenburg, and Tom Westland, as well as seminar participants at the University of Siena and Wageningen University. I am grateful for the financial support from The Netherlands Organisation for Scientific Research for the project “South-South Divergence: Comparative Histories of Regional Integration in Southeast Asia and Sub-Saharan Africa since 1850” (NWO VICI Grant Number VI.C.201.062).

ENDNOTES

• 1 Some have even gone as far as to suggest that the origins of global economic inequality are rooted in more or less favorable degrees of genetic diversity (Galor, 2022).
• 2 To be sure, the terms north–south and core-periphery both preceded the term “Great Divergence” (Wallerstein, 1974; Brandt et al., 1980).
• 3 This correction is less important for reciprocal comparisons between Britain and Japan, which shared their island geography and differed much less in population numbers (He, 2013; Francks, 2016; Vries, 2020).
• 4 The meta-question why Europe heralded the industrial revolution instead of China, which after all had been the world's longstanding technology leader, has been touched upon by many key thinkers of the 19th and 20th century. Hegel, Marx, Weber, and Wittfogel all formulated their own explanations for the rise of capitalism in Europe and the perceived long-term economic stagnation of Asia. Especially in Marxist circles, the idea that Oriental despotism had locked countries such as India and China into a persistent state of feudalism remained popular for long.
• 5 North–south conceptions of the global economic order were, in turn, inspired by earlier tripartite divisions in the First, second, and third world, which were largely synchronous with Wallerstein's (1974) concepts of the core, semi-periphery, and periphery. The Brandt report justified this reductionist perspective amongst others by repeated calls upon the leaders of the second world (i.e., the Soviet Union) to acknowledge and act upon their responsibilities in addressing the world's poverty problem, as being part of the more developed Northern hemisphere (Brandt et al., 1980, p. 10).
• 6 This is not to imply that there were no earlier phases of divergence across the global South.
• 7 While the cross-country variation in average per capita income levels in the global North declined from 0.59 in 1950 to 0.46 in 2018, the variation in the global south rose from 0.77 to 1.08. The temporary rise in inequality in the global north during the 1980s and early 1990s can be attributed to the economic stagnation in the socialist states of Eastern Europe and the recession that followed the fall of the Iron Curtain and subsequent shock therapy. The trend reversed again in the mid-1990s, also because of continued convergence between Northwestern Europe and the poorer parts of Southern Europe (Eichengreen, 2007).
• 8 For further details on the UNDP projections of African population growth see Paice (2021).
• 9 I sidestep the debate on measurement and classification of poverty by the World Bank, except for one remark: the total number of extreme poor defined as living below the $1.90 threshold have grown modestly, but poverty rates with thresholds at $3.20 or $5.10 have increased (also in the precovid age). World Bank, World Development Indicators, 2020.
• 10 A quick glance at the geographic distribution of submissions in the top-field Journal of Economic History confirms the trend: in the early 2000s (2001–2005) submissions on Africa, Asia, Latin America, and the Middle East comprised 18% of total submissions, in the past 4 years (2018–2022), these regions were good for 29% of all submissions. Data from JEH Editor's Report for 2004/5 and 2022, respectively, Tables 2 and 5.
• 11 Examples are the foundation of the Asian Historical Economics Society in 2010 and the African Economic History Network in 2011. Research on major subregions such as North Africa, West and Central Asia, the Indo-Pacific, Central America, and the Caribbean remains underrepresented in these larger networks and conferences.
• 12 In neoliberal market economies, property and supervision are both left in private hands.
• 13 List's Natural System of Political Economy was in turn inspired by ideas of the American politician, economist, and independence fighter Alexander Hamilton (Chang 1999, p. 182-3), whose visions on the role of the state and the union in safeguarding the economic interests of an independent America were anchored in the US constitution.
• 14 That said, these societies did not share the features of economic “backwardness” common across the South by the mid-20th century.
• 15 Another critique on this conceptual framework is its undisguised Euro-centric perspective on the relationship between liberal democracy and income growth potential.
• 16 Party of this effect may be explained by the ICP overweighting globally comparable goods at the expense of cheaper local goods (Gelb & Diofasi, 2015, p. 18) .
• 17 To be precise, this is an example of diachronic interregional comparison.

APPENDIX 1: COUNTRIES INCLUDED IN THE GLOBAL NORTH AND SOUTH

	Global North			Global South
1	Albania	ALB	1	Afghanistan	AFG	41	Indonesia	IDN	81	Sudan (Former)	SDN
2	Australia	AUS	2	Angola	AGO	42	India	IND	82	Senegal	SEN
3	Austria	AUT	3	Argentina	ARG	43	Iran	IRN	83	Singapore	SGP
4	Belgium	BEL	4	Burundi	BDI	44	Iraq	IRQ	84	Sierra Leone	SLE
5	Bulgaria	BGR	5	Benin	BEN	45	Jamaica	JAM	85	El Salvador	SLV
6	Bosnia & Herzegovina	BIH	6	Burkina Faso	BFA	46	Jordan	JOR	86	Sao Tome and Principe	STP
7	Canada	CAN	7	Bangladesh	BGD	47	Kenya	KEN	87	Swaziland	SWZ
8	Switzerland	CHE	8	Bolivia	BOL	48	Cambodia	KHM	88	Seychelles	SYC
9	Czechoslovakia	CSK	9	Brazil	BRA	49	Republic of Korea	KOR	89	Syrian Arab Republic	SYR
10	Cyprus	CYP	10	Barbados	BRB	50	Lao People's DR	LAO	90	Chad	TCD
11	Germany	DEU	11	Botswana	BWA	51	Lebanon	LBN	91	Togo	TGO
12	Denmark	DNK	12	Central African Republic	CAF	52	Liberia	LBR	92	Thailand	THA
13	Spain	ESP	13	Chile	CHL	53	Libya	LBY	93	Trinidad and Tobago	TTO
14	Finland	FIN	14	China	CHN	54	Saint Lucia	LCA	94	Tunisia	TUN
15	France	FRA	15	Côte d'Ivoire	CIV	55	Sri Lanka	LKA	95	Turkey	TUR
16	United Kingdom	GBR	16	Cameroon	CMR	56	Lesotho	LSO	96	Taiwan, Province of China	TWN
17	Greece	GRC	17	D.R. of the Congo	COD	57	Morocco	MAR	97	U.R. of Tanzania: Mainland	TZA
18	Croatia	HRV	18	Congo	COG	58	Madagascar	MDG	98	Uganda	UGA
19	Hungary	HUN	19	Colombia	COL	59	Mexico	MEX	99	Uruguay	URY
20	Ireland	IRL	20	Comoros	COM	60	Mali	MLI	100	Venezuela	VEN
21	Iceland	ISL	21	Cabo Verde	CPV	61	Myanmar	MMR	101	Viet Nam	VNM
22	Israel	ISR	22	Costa Rica	CRI	62	Mongolia	MNG	102	Yemen	YEM
23	Italy	ITA	23	Cuba	CUB	63	Mozambique	MOZ	103	South Africa	ZAF
24	Japan	JPN	24	Djibouti	DJI	64	Mauritania	MRT	104	Zambia	ZMB
25	Luxembourg	LUX	25	Dominica	DMA	65	Mauritius	MUS	105	Zimbabwe	ZWE
26	Malta	MLT	26	Dominican Republic	DOM	66	Malawi	MWI
27	Macedonia	MKD	27	Algeria	DZA	67	Malaysia	MYS
28	Montenegro	MNE	28	Ecuador	ECU	68	Namibia	NAM
29	Netherlands	NLD	29	Egypt	EGY	69	Niger	NER
30	Norway	NOR	30	Ethiopia	ETH	70	Nigeria	NGA
31	New Zealand	NZL	31	Gabon	GAB	71	Nicaragua	NIC
32	Poland	POL	32	Ghana	GHA	72	Nepal	NPL
33	Portugal	PRT	33	Guinea	GIN	73	Pakistan	PAK
34	Romania	ROU	34	Gambia	GMB	74	Panama	PAN
35	Serbia	SRB	35	Guinea-Bissau	GNB	75	Peru	PER
36	Former USSR	SUN	36	Equatorial Guinea	GNQ	76	Philippines	PHL
37	Slovenia	SVN	37	Guatemala	GTM	77	Puerto Rico	PRI
38	Sweden	SWE	38	China, Hong Kong SAR	HKG	78	Paraguay	PRY
39	United States	USA	39	Honduras	HND	79	State of Palestine	PSE
40	Former Yugoslavia	YUG	40	Haiti	HTI	80	Rwanda	RWA

Open Research

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are openly available in large databanks by the Groningen Growth and Development Centre (GGDC), the United Nations, and the World Bank. My own elaborations of GDP and population data can be traced back to a set of excel files which I will be happy to share with anyone interested upon reasonable request.