This essay first explores how the defining features of comparative-historical analysis (CHA) endow this approach with comparative advantages for substantive research on macro development. It identifies three major traits of CHA that make this tradition an especially powerful approach for the study of development: cased-based research; a focus on context and configurations; and a concern with mechanisms and temporal flow. It then considers what concretely we have learned about the sources of macro development from CHA studies. While some findings are tentative or open to debate, the list of empirical contributions from CHA is quite substantial. Finally, the essay considers some of the frontiers of CHA work on development. Included at this frontier is work on the concept of development itself, displacing the centrality of property institutions with a focus on identity institutions, and moving the focus of analysis from the national level to subnational and supranational levels.
Comparative-historical analysis (CHA) is featured centrally among the best known studies of macro development, including the work of the sociological founders. Marx, Durkheim, Tocqueville, and Weber were all comparative-historical analysts who focused on questions of development. Yet these sociological founders oriented their work mainly to questions of development in Europe. Today's CHA work on development fully incorporates the non-European world (without excluding Europe). This new CHA work features influential studies on the role of states and markets in development (Johnson 1982; Migdal 1988; Amsden 1989; Wade 1990; Evans 1995; and Kohli 2004) as well as on diverse topics such as colonialism and development (Steinmetz 2007; Lange 2009; Mahoney 2010; Go 2011), social movements and development (Heller 1999; Kay 2011; Tarrow 2013; Agarwala 2013), and gender and development (e.g., Charrad 2001; O'Connor, Orloff, and Shaver 1999; Adams 2005). Although the field of development is still often associated with the quantitative work of economists, many of its most important contributions have been made by comparative-historical sociologists.
In this essay, I first explore how the defining features of CHA endow this approach with comparative advantages for substantive research on macro development. I identify three major traits of CHA that make this tradition an especially powerful approach for the study of development: case-based research; a focus on context and configurations; and a concern with mechanisms and temporal flow. In discussing these traits, I intend not to diminish or dismiss the potential of other analytic approaches, but to pinpoint what we gain from a CHA approach when applied to questions of development.
Second, I consider what concretely we have learned about the sources of macro development from CHA studies. While some findings are tentative or open to debate, the list of empirical contributions from CHA is quite substantial. CHA findings are often derived from the analysis of a small number of cases, but they frequently can be cumulated so that they add up to fairly broad generalizations. Accordingly, in this second part, I spell out the midrange generalizations we have gained from CHA studies of development.
Finally, I consider some of the frontiers of CHA work on development. The concept of development itself is one of these frontiers: it has changed substantially over time and will continue to evolve, partly in tandem with the changing experiences of people outside of the Western world. Beyond that, CHA work on development is shifting from a focus on property institutions to a focus on identity institutions and from analyses at the national level to analyses at the subnational and supranational levels.
THE COMPARATIVE ADVANTAGES OF COMPARATIVE-HISTORICAL RESEARCH
Certain traits inherent to comparative-historical research make this approach especially valuable for the study of macro development. These traits explain why CHA has for many decades made theoretical contributions and help to set the substantive agendas in the field (see also Thelen and Mahoney forthcoming).
CHA work is focused on the close analysis of specific cases, usually a small number of countries or other macro units. Research in this tradition places great emphasis on getting the case or cases right: with CHA, a good explanation must make sense in light of fine-grained details from the case. CHA researchers place a premium on showing how their explanation applies to historical specifics and not simply stylized hypothetical cases. In fact, it can be hard to separate CHA works from the specific cases analyzed. For example, works such as Evans's Embedded Autonomy (1995) and Kohli's State-Based Development (2004) are closely associated with the analysis of specific sets of countries (development sociologists can often name these specific countries).
One advantage of CHA's case-based research is that it allows for the inductive generation of new insights and new theory. CHA scholars of development are never simply “testing” preexisting hypotheses. Rather, they are always assessing evidence in ways that allow for the creation of new ideas and new hypotheses. In many cases, novel concepts—such as double movement, embedded autonomy, and mercantilist colonialism—lie at the heart of these theoretical contributions. In other instances, the central contribution is new core positions about development—for example, the hypothesis that state autonomy is required for rapid economic growth or that colonialism sets relative levels of development that prove remarkably enduring. Either way, the concepts or propositions are more general theoretical constructs that can be applied beyond the original formulating case/s.
Case-based research on development yields two kinds of contributions. On the one hand, this research tells us something about the specific case/s under study. For example, we learn many nuanced facts about Korean political economy from Alice Amsden's Asia's Next Giant (1989). On the other hand, the insights generated in explaining the specific case/s under analysis can stimulate broader research agendas that are taken up by scholars of diverse methodological orientations. For example, consider the massive research agenda stimulated by Amsden's book or by a work such as Chalmers Johnson's MITI and the Japanese Miracle (1982). Johnson's work in fact influenced generations of scholars and reached deeply into the policy community as well (see Haggard forthcoming).
In line with its focus on specific cases, CHA research is problem driven in the sense that it is animated by real-world questions: CHA scholars explore empirical puzzles anchored in particular times and places (Pierson and Skocpol 2002; Mahoney and Rueschemeyer 2003). The kinds of questions they typically ask are different in basic ways from the kinds of questions that large-N scholars typically ask. CHA scholars ask questions about the causes of specific outcomes in particular cases, such as why Korea and Taiwan experienced sustained high growth or why Costa Rica experienced better developmental outcomes than the rest of Central America. By contrast, large-N researchers typically ask questions about the average effect of variables within a population of cases, such as whether GDP per capita increases health outcomes across countries.
Both kinds of questions are valuable. But the empirical puzzle-driven questions of CHA are especially well suited for the explanation of outcomes of exceptional interest, such as cases of revolutions, sustained episodes of very high growth, and strong social development in the midst of economic poverty. These outcomes are outliers in statistical analyses or, alternatively, cases with such extreme values on the dependent variable that they are not well explained by the statistical model. CHA has a comparative advantage for cases with outcomes that are exceptional and otherwise not easily explained as part of a general population trend. Its mode of explanation is designed precisely to account for why specific cases experience particular outcomes, including outcomes that seem puzzling or unusual within a larger population of cases.
A Focus on Context and Configurations
CHA research explains developmental outcomes by considering how multiple factors combine to form coherent larger combinations or packages, using what can be called configurational analysis (Ragin 1987). This approach to causation assumes that interactions among variables are common and that individual causal factors usually must be analyzed as parts of larger packages. For example, to understand the effect of investment on growth for one or more cases, CHA suggests that ordinarily one will need to consider the values of other variables. At the level of the individual case, investment may have a positive effect, a negative effect, or no effect on growth, depending on the values of these other variables. These other variables work with investment as part of a larger configuration to shape growth.
Specifying causal configurations is one important way in which CHA researchers bring “context” into the analysis of development. Variables within a given causal configuration pick up features of the background and the setting that make up the context of a case. For example, in the study of rapid economic growth in the Asian newly industrialized countries (Deyo 1987; Koo 1987), scholars argued that labor repression contributed to high growth. When bivariate statistical studies then showed that labor repression is not correlated with growth (Geddes 2003), these scholars were not surprised (see Goertz and Mahoney 2012:179–81). They argued that the effect of labor repression on growth depends on the values of other variables. In particular, labor repression can contribute to growth when accompanied by an emerging developmental state, an abundant supply of cheap labor, and a strong market for labor-intensive exports—but that in other contexts it will not have this effect.
The strength of configurational explanations, then, is closely linked to recognizing that the effects of individual variables usually depend on the values of other variables. CHA scholars take context seriously because doing so is necessary for valid explanation, including understanding when individual causal factors are likely to have certain effects. However, the use of configurational explanation means that CHA arguments often cannot be well summarized as simple relationships or even necessarily divorced from the case under study. Instead, CHA findings are frequently more “qualified” generalizations in the sense that they assume that a given variable value may contribute to an outcome in one context but not another. For example, whereas a strategy of primary product exportation may contribute to growth if the exporter is endowed with natural resources that are needed for manufacturing by a rising economic hegemon, the same strategy may be a formula for economic stagnation if the exporter lacks the appropriate resources or is linked to a declining metropolis with weak and declining demand.
A Concern with Mechanisms and Temporal Flow
CHA explanations focus attention on the mechanisms through which causes and causal configurations exert effects within particular cases. Scholars of development in the CHA tradition distinguish incidental correlations from causal associations in part on the basis of whether mechanisms can be identified to explain the associations. When mechanisms cannot be found, they may be skeptical of a potential explanation. They believe a convincing explanation should reveal why a correlation exists by opening up the black box and identifying the steps that connect observed causes to observed outcomes.
As an example of this kind of explanation, consider Dietrich Rueschemeyer, Evelyne Stephens, and John D. Stephens's (1992) Capitalist Development and Democracy. These authors follow a long line of research in noting an association between capitalist development and democracy. However, on the basis of the close analysis of individual countries, they find that the effects of development on democracy depend on intervening mechanisms. For instance, development was associated with democracy in many cases of historical Europe because it strengthened prodemocratic working classes and weakened antidemocratic landed elites. In Latin America, however, the organized working class was not always tied to democratic leaders and movements, so development was not associated with democracy. By uncovering the mechanisms through which development and democracy are associated, Rueschemeyer, Stephens, and Stephens provide insights about when the two phenomena will and will not go together.
In CHA, the concern with temporal flow allows researchers to pay close attention to the sequence in which variables appear and their timing relative to one another. Placing an explanation “in context” often means situating variables in a particular temporal setting. CHA scholars are concerned with temporal setting in part because the effect of a variable on developmental outcomes can change, depending on its temporal location (e.g., Pierson 2004). For example, in Falleti's (2010) study of decentralization reforms in Latin America, administrative decentralization reforms can either empower governors and mayors or weaken them, depending on their position within an overall sequence of decentralization reforms.
A focus on mechanisms and temporal processes also allows for the study of phenomena such as path dependence in long-run development. With a path-dependent pattern, a key explanatory task is identifying the historical sources of the pattern. CHA works often focus on historical turning points or critical junctures when initial decisions or events occur to launch these sequences. These historical periods hold the selection processes and root causes that explain how the long-run pattern was started in the first place. For example, if one seeks to explain why western Europe is relatively rich and Africa is relatively poor, a good explanation in CHA demands that one recognizes path dependence in developmental outcomes and find the “original” causes that placed Europe and Africa on different paths of development. Contemporary differences between Europe and Africa—such as differences in demography, consumption, and savings—are very likely symptoms or effects of path-dependent development outcomes, not root causes of these outcomes. The causal roots of the different outcomes probably go back several centuries to preindustrial times, when the developmental differences between the continents were not so large.
This example is intended to illustrate how an approach that looks at temporal processes over long periods of time can “see” causal patterns that would be invisible with approaches that adopt only short time horizons. Because of its long time horizons, CHA is well suited to uncover otherwise invisible historical sources of development outcomes.
WHAT HAVE WE LEARNED SO FAR?
CHA studies in the field of sociology are responsible for a good many of our stable findings about the sources and consequences of development. As with any research domain in the social sciences, some of these findings are more conclusive than others; and none of them are uncontroversial or beyond falsification. Nevertheless, by looking across various CHA works, we can summarize several major results that make up important aspects of what we know so far about development. These findings are not exhaustive, but they are suggestive of CHA's contribution to knowledge gains.
States and Development
State Intervention Plays a Leading Role in Both Promoting and Hindering Development
CHA studies have been at the forefront of rejecting the idea that state intervention is in general “good” or “bad” for development. Likewise, they have been central to the rejection of the hypothesis that a small or minimal state is the best apparatus for development (or that an especially large amount of state invention is required for rapid development). Instead, they have shown that state intervention can either help or hinder development, depending on the specific kind of intervention and the context in which it occurs (see especially Rueschemeyer and Evans 1985; Evans 1995).
Perhaps most importantly, the findings generated by this literature propose a set of conditions under which the state can play a positive role in promoting development. This positive role is especially likely if the following three conditions are found together in combination.
Autonomy from powerful economic classes: To promote development, state actors need to have policy-making autonomy from economic elites. This autonomy enables officeholders to formulate and implement development projects that are not simply reflections of the interests of elites. These projects include the provision of public goods that do not maximize short-term revenue gains for either the officeholders or the economic elites. In some cases, states must carry out policies that clash with the economic interests of elites to achieve broad-based development.
Internal cohesion and bureaucracy: To promote development, states require a coherent, meritocratic bureaucracy in which officeholders operate according to rules and established norms. With a meritocratic bureaucracy, officeholders can and often will seek to maximize their individual interests, but they do so by working within the predefined system of bureaucratic rules. The presence of a cohesive bureaucracy is especially important in those parts of the state responsible for growth and development.
Strong external networks connecting state and civil society: To promote development, officeholders must be connected to important civil society groups through external networks. These networks may be informal ties that link officeholders to capitalist elites oriented toward economic growth. They may also entail linkages between state actors and other societal groups, such as organized workers or civil associations.
While any one of these conditions may be associated with development on average, the key finding of the CHA literature is that the combination of these three conditions is an especially powerful determinant of growth. Because this configuration or “recipe” for effective states was derived mainly from CHA analyses of sustained high growth in cases such as Korea and Taiwan, one can raise questions about the extent to which it will continue to apply in the twenty-first century. Nevertheless, it stands as one of the more important findings about the sources of contemporary development.
Historical Exogenous Processes Cause Developmental States
The CHA literature has also been at the forefront in teaching us about the causes of effective states in developing countries (Rueschemeyer and Evans 1985; Chibber 2003; Kohli 2004; Lange 2009). A core lesson of this literature is that such states are extremely difficult or impossible to create by willful design. Instead, these states are generally the products of historical processes that originate from outside of the state's borders. The literature suggests two exogenous processes that have been especially important in the creation of effective developmental states:
Grave external threats: Political leaders generally build developmental states only when they face dire political threats, typically emanating from abroad, and lack easy access to resources. When leaders face a national security crisis, and when they do not have easy access to resources, they are forced by circumstances to forge the broad coalitions and pursue the high-growth policies that underpin developmental states. In the absence of these imperatives, political leaders are prone to cultivate narrow coalitions or to draw on readily available resources to deal with external threats.
Developmentalist colonial experiences: Colonial experiences in which the colonizer pursues intensive state-led development policies generally leave behind favorable conditions for a postcolonial developmentalist state. Colonial policies themselves may move the state in the direction of a meritocratic bureaucracy and bring some coherence to agencies managing the economy. In addition, colonial rule may dismantle existing power holders in a way that fosters conditions of state autonomy. The classic example is Japanese colonialism in Korea and Taiwan, under which Japan promoted manufacturing colonies that it imagined would serve its interests over the long run.
The processes that tend to bring into being developmental states mostly lie outside of the agency of officeholders. State leaders cannot generate grave external national security crises or change colonial history. Thus the existing CHA literature leaves us with rather pessimistic conclusions about the likelihood of creating developmental states. However, this literature does suggest several tangible goals for officeholders: cultivate agencies that can pursue collective goals; devote resources toward the promotion of competent, coherent bureaucracies; and expand and encourage networks to key civil society groups. Moreover, the CHA literature has been influential in thinking about alternative recipes for manufacturing sustained high-growth formulas that are variations on the developmental state model (e.g., Evans and Heller forthcoming). CHA work on the extent to which the recent high growth in China conforms to the developmental state literature is one example of this kind of analysis (e.g., Evans 2008; Knight and Ding 2012).
Colonialism Is a Fundamental Cause of Long-Run Development Outcomes
A long line of CHA work dating back to the 1970s calls attention to the centrality of colonialism in shaping development outcomes throughout much of Latin America, Africa, and Asia (for recent contributions and literature discussions, see Kohli 2004; Krieckhaus 2006; Lange 2009; Mahoney 2010; Go 2011; Steinmetz 2007). This work suggests that variations in the form of colonialism are a fundamental cause of variations in relative levels of political colonial development. While colonialism may not always be useful in understanding variations in postcolonial growth trajectories, it goes a long way toward explaining why some countries became richer than others in the first place.
One core component of the finding about colonialism entails a path-dependent argument: initial postcolonial levels of development tend to be sticky. Thus countries that emerge from colonialism wealthier tend to stay wealthier, and analogously for poorer countries. This stickiness is explained in the CHA literature in part by the fact that developmental states are very hard to construct, making sustained high growth a rare phenomenon. More generally, the stickiness has been understood as a product of path dependence and as linked to increasing returns to power for elites (e.g., Mahoney 2010).
The CHA literature has been concerned not only with the effects of colonialism on development but with the kinds of colonialism that bring countries to higher or lower levels of development. The literature finds that colonial institutions were crucial and that the nature of these institutions varied across different types of colonies. For example, British settler colonies laid down institutions that yielded rich postcolonial countries, whereas indirect British colonialism supported institutions that led to developmental disasters in the aftermath of colonialism. By contrast, the Spanish colonies with heavy European settlement created institutions poorly suited to development, whereas the marginal Spanish colonies were sometimes comparatively more successful. The implication is not that either British or Spanish colonialism is, on average, more favorable for development. Rather, the implication is that the effect of British and Spanish colonialism on development varies depending on the intensity of colonialism.
The CHA literature yields other important general findings about the effects of colonialism. This literature shows that while colonialism was always traumatic for indigenous groups, it could be favorable for postcolonial development if the colonizing nation was both a rising world hegemon and oriented toward long-run settlement. This fact helps explain why the British settler colonies emerged independent as some of the wealthier countries in the world. By contrast, colonialism was nearly certain to yield underdevelopment when it was led by a hegemon that pursued indirect rule or when the colonizing nation was a fallen hegemon oriented toward settlement. This fact helps explain why the core Spanish colonies, such as Mexico and Peru, and especially the indirectly ruled British colonies in Africa, such as Nigeria and Sierra Leone, emerged independent as quite poor countries.
World-Historical Time Matters for Policy
Different world-historical times are marked by differences in technology, modes of production, levels of global and regional trade, norms about human rights, and many other international-contextual processes that can shape and limit the possibilities for development. The CHA literature has found that the effects of particular policies and development strategies vary depending on the world-historical time in which they are pursued.
One example is the varying results of primary product exporting as a growth strategy over time and place. In sub-Saharan Africa, this strategy was relatively successful from the 1960s until the mid-1970s (e.g., Arrighi 2002), failed from the mid-1970s until the early 2000s, and became a source of growth for several countries over the last decade. In Latin America, primary product exporting was associated with significant growth from 1870 to 1930 (Furtado 1976). But problems with this model became obvious with the Great Depression, leading the major economies of the region to move toward industrialization strategies. In recent years, with the rise of China, primary product exporting has once again become a source of growth for many countries in the region. The clear lesson that follows is that growth policies succeed or fail depending, in significant part, on the economic opportunities available in the world economy. Because world-economic context matters, one must be suspicious of one-size-fits-all growth solutions.
Another finding from the CHA literature concerns Korea's and Taiwan's successful outward-oriented industrialization strategy initiated in the early 1960s. The timing of this strategy was important: the world capitalist economy was growing during this era. As one author concludes, “Core capitalists were eager to invest, Euromoney was easy to borrow, and interest rates were relatively low; there were fewer barriers against Third World manufactures and little competition from other labor-rich developing countries” (Koo 1987:169). By contrast, when South American countries pursued a similar growth strategy in the 1970s, they encountered a very different and far less auspicious world economy. The conclusion from the CHA literature is that an outward-oriented strategy is not inherently superior to more inward-looking strategies (e.g., import substitution industrialization); inward-oriented strategies can be appropriate, depending on context.
Domestic Social Structures
Powerful Landlords Hinder Development
The CHA literature suggests that, in modern times, the weakening or elimination of powerful landlords is nearly prerequisite for sustained high economic growth and greatly contributes to social development outcomes as well. Landlords are powerful when they exercise substantial local and national political influence, when they employ or control large quantities of subordinate labor, and when they own large estates. Powerful landlords hinder development by undermining state autonomy, blocking the creation of public goods, and thwarting economic change. Their labor-dependent agricultural estates depress domestic demand, trap labor and hinder mobility, and discourage innovation and modernization. Powerful landlords generally stunt development even when rural production is highly profitable in the world market and primary product exportation is a favorable growth strategy.
The corollary of this finding is that the absence of powerful landlords is an almost necessary condition for sustained high growth in modern times. In some cases, powerful landlord classes may not have formed historically, so that this obstacle does not need to be removed. But most countries have historically featured this group, and thus sustained high growth requires its weakening or elimination. The elimination of powerful landlords is rare, and it typically occurs through events such as colonialism, social revolutions, and wars—which can sometimes rapidly remove large estates, free dependent rural labor, and block political access for rural elites. South Korea and Taiwan are examples where powerful landlord classes were undercut by colonial policy and successful land reforms. Social revolutions in Mexico and China also weakened landlord classes, allowing for the possibility of more sustained and equitable growth in the future.
Ethnic Divisions Hinder Social Development
CHA works underscore the importance of ethnic cooperation for the distribution of public goods and the promotion of social development outcomes (e.g., Lieberman and Singh 2012; Mahoney 2010; Sandbrook et al. 2007). While economic development is strongly correlated with social development, there are outlier cases that are economically poor but perform rather well on social indicators (e.g., Costa Rica) or that are fairly rich but underperform on social indicators (e.g., South Africa). In the latter cases, CHA studies suggest that historical ethnic divisions have prohibited the dissemination of wealth to a broad spectrum of society, thereby undercutting social development progress even in the presence of economic wealth. By contrast, the former cases nearly always have found ways to overcome ethnic divisions, including by imagining the nation as ethnically homogeneous in such a way that all citizens are entitled to the fruits of wealth. As a result, these cases have achieved gains in social development even though they are not rich.
It bears emphasis that the key finding is not that ethnic heterogeneity per se undercuts social development. Rather, the finding is that polarized ethnic identities undercut social development. Or, said differently, the absence of polarized ethnic identities is a nearly necessary condition for social development (though not economic development). While ethnic heterogeneity may make ethnic divisions more likely, the source of the problem is the divisions, not the heterogeneity itself. For example, while Costa Rica features a diverse range of ethnic groups, it lacks intense ethnic divisions and performs well on standard social indicators. Conversely, even very rich countries—such as the United States—will underperform on these indicators if their history has produced important and persisting ethnic divisions.
AGENDAS FOR FUTURE RESEARCH
CHA works grow in significant measure out of real-world problems, and real-world changes in the political economies of nations and the international system will dictate many of the new empirical agendas that emerge in this field in the coming years. Nevertheless, some trends in scholarship can be discerned and briefly discussed.
While the concept of development has always been contested, until recently most definitions emphasized primarily or exclusively economic growth, often measured as GDP per capita. This conceptualization has come under scrutiny in the last 15 years, partly because of the publication of works such as Amartya Sen's Development as Freedom (1999). Scholars of development now often emphasize the expansion of human capabilities as the core of development. This reconceptualization puts the spotlight more on social development, as measured by indicators such as infant mortality, literacy, and life expectancy. These indicators better capture the concept of “development,” if it is taken to mean quality of life, including an ability of people to lead the kind of life they value and have reason to value (Sen 1999).
The understanding of development as the expansion of human capabilities puts more focus on civil society than in the existing state-centric literature. For example, while promoting education and health care certainly requires state capacity, it also demands assistance from civil society groups who can help overcome resistance from traditional forms of authority threatened by the changes (Evans and Heller forthcoming). The kinds of civil society groups that can support these interventions are not likely to be the economic elites emphasized in the literature on embedded autonomy. Instead, they are more likely to be subordinate actors who have the greatest stake in the interventions. As the concept of development becomes more focused on ideas of expanding human capabilities, it seems likely that discussions and possible measurements of levels of development will increasingly look to the situations of civil society organizations.
Ethnic and Identity Institutions
CHA works on development continue to focus on historically constructed institutions as causes and outcomes to be explained. However, CHA researchers are increasingly paying attention to the institutions of ethnic stratification as well as the institutions of private property (e.g., Lieberman and Singh 2012), the latter of which have received more attention to date. Whereas the institutions of private property (or, simply, “property rights”) affect access to resources through rules for regulating economic resources (e.g., labor and land), the institutions of ethnic stratification shape access to resources through rules for regulating subjective identities. Ethnic institutions are the rules that classify categories of people as ethnicities and then regulate behaviors within and across these categories. They include various informal cultural codes that are never written down and that may be simply taken for granted. They also include laws for regulating behavior, which often do not refer specifically to ethnic groups but that nonetheless shape ethnic stratification.
We can expect CHA researchers to increasingly analyze ethnic institutions and other identity institutions, paying attention to the rules governing identification and behavior across different identity communities and within particular identity communities. The rules shaping cross-identity relationships often create asymmetrical categories in which whole collectivities of individuals fall into dominant and subordinate positions. The challenge for CHA analysts of development will be to elucidate the ways in which these formal and informal rules create and distribute resources to promote human well-being. With respect to within-community analysis, researchers must look at institutionally driven resource flows among a group of individuals who share identification. These within-community resource flows may or may not reinforce existing identities; and they may or may not facilitate collective action. The task for CHA scholars of development is to explore precisely these questions in order to make sense of development outcomes in ways that works focused only on property institutions cannot.
Subnational, National, and Supernational Studies
To date, it seems fair to say that most CHA works on development have focused on nation-states. To be sure, some works have always operated at a higher level of analysis, including famously Immanuel Wallerstein's (1974, 1980, 1989, 2011) studies of the modern world system. Likewise, some important CHA works have linked higher and lower levels of analysis, such as Saskia Sassen's Cities in a World Economy (1994) and Janet Abu-Lughod's Before European Hegemony: The World System, A.D. 1250–1350 (1991). In the coming years, however, even more work is likely to shift analysis to higher and lower levels.
On the one hand, as CHA scholars of development take up questions related to globalization, they move the focus of analysis above the nation-state. While few scholars believe that globalization implies the complete withering or outright retreat of the state (Evans 1997; O'Riain 2000; Sassen 1996, 2008; Strange 1996), it does force analysts to think about groups, classes, and movements in a global perspective. The tools of CHA, including its concern with configurational analysis, are well suited for the study of an interconnected world economy. This is especially true insofar as one asks questions about real-world outcomes located in specific times and (supranational) places.
On the other hand, development scholars are likely to pay increasing attention to the vast regional differences within nation-states. Key questions include: Why are individuals in some subnational regions of a given state more capable of leading healthy, informed lives than those in other subnational regions? Why are certain subnational regions in one state similar to subnational regions in another state? How and why do subnational groups shape development outcomes at local, national, and global levels? The methodological tools of CHA are once more well suited to addressing these questions by virtue of their capacity to take on focused comparisons of concrete development experiences.
In the future, CHA work on development will surely continue to embrace the core characteristics that distinguish this approach: case-based research, a focus on context and configurations, and temporally oriented analysis. These characteristics have enabled CHA works to contribute important findings that tell us much about what we currently know concerning development. Yet we can also anticipate some important changes. New work is likely to focus more on social development outcomes than on economic development. This work is likely to emphasize identity institutions more than previous work, which focused primarily on material institutions. And the new work is likely to call increasing attention to levels of analysis both above and below that of the nation-state. The result of these new emphases will doubtless be important new insights about the sources and consequences of development that complement those that we already have.