Institutions,social change,and economic development in the periphery: A confrontation between neo-institutionalism and Arrighi and Piselli’s essay on Calabria

Comparative historical institutional analyses of economic development

Comparative historical institutional analyses of economic development (henceforth, CIAED) were proposed in a series of works within political science or political economy that appeared in the 1980s and 1990s to reverse the then-influential view of economic development as an outcome of market-driven policies, associated with the so-called ‘Washington consensus’ (Williamson 1990). With a focus on specific, then-well performing developing economies, such as Brazil in the 1970s (Evans 1979) or East Asian economies, such as South Korea (Amsden 1989) or Taiwan (Wade 1990) in the 1980s, these works, loosely connected to the ‘developmental state’ tradition in comparative political economy, epitomized by Chalmer Johnson’s (1982) famous book on industrial policy in post-war Japan, ⁴ all highlighted the important role played by state policies in fostering economic development where local private entrepreneurs or capitalists were absent or lacking. Thus, CIAED (as a loosely connected cluster of neo-institutional works) had a strong focus on ‘hard’, state-centered institutions: state agencies and bureaucracy and state policies. It can be related, therefore, to a broader project in political science: that of ‘bringing the state back in’ the analysis of political and social phenomena, as suggested by Peter Evans et al. (1985) in a programmatic 1985 volume. This overall project does not imply a dichotomous, ‘state versus market’ agenda: on the contrary, several authors within this tradition emphasized the strong interconnections between active government policy and vibrant markets. As Robert Wade (1990) put it, for instance, the economic successes of Taiwan, Japan and South Korea in the 1980s might be similarly ascribed to ‘an intense and almost unequivocal commitment on the part of government to build up the international competitiveness of domestic industry’ (p. 7). ⁵

Works within the CIAED framework emphasized the operational and strategic autonomy of political institutions – while carefully observing that such autonomy is no synonym for separatedness; as the editors of the already cited 1985 volume argued in their introduction, indeed, ‘[h]euristically, at least, it is fruitful to assume both that states are potentially autonomous and, conversely, that socioeconomic relations influence and limit state structures and activities’ (Evans et al. 1985: viii). This double assumption informs most subsequent studies within this tradition (see, in particular, Amsden 1989; Wade 1990) and turned into the ‘embedded autonomy’ hypothesis formulated by Peter Evans in a later volume, dedicated to South-East Asian development (Evans 1995). Again, however, such autonomy is not presented as absolute and is seen as the outcome of specific political economic dynamics. By contrast with cruder, earlier elaborations, ⁶ Evans described developmental states as organizationally coherent (Evans speaks of ‘corporate coherence’), and it is this coherence that constitutes the basis for their operational and strategic autonomy. However, at the same time, developmental states are ‘embedded in a concrete set of social ties that binds the state to society and provides institutionalized channels for the continual negotiation and renegotiation of goals and policies’ (Evans 1995: 12). Thus, the institutional system culminating in state’s economic interventionism simultaneously relies on and shapes specific state–society relations premised on certain social structures. The (limited) variety of ideal-typical state models in development (predatory, developmental, ‘intermediate’) implies a parallel variety of state–society relations: ‘Developmental states played a central role in producing the organized industrial classes they needed as counterparts. The Zairian state also helped produce the counterpart that it needed: a disorganized and divided civil society incapable of resisting predation’ (Evans 1995: 72). Finally, a specific social structure (‘an organized class of industrialists’: Evans 1995, 228) is required to complete this virtuous configuration. Indeed, these various degrees of embeddedness and autonomy have a direct impact on economic success: ‘Efficacious states combine well-developed, bureaucratic internal organization with dense public-private ties’ (Evans 1995: 72).

The original institutional focus of works within the CIAED framework was solidly state-centered, given, first, the desire of some of the authors whose works are cited above to criticize and go beyond the then-widespread view of economic growth as a market-driven process; given, second, the general belief that state institutions and policies should be brought back at the forefront of the analysis of economic development; and given, third, the cases themselves – countries whose impressive economic performance was indeed correlated with intense state involvement in the economy. However, subsequent works within the same tradition have somehow shed this state-centric bias. This is the case, for instance, of Richard Doner’s (2009) work on Thailand, where the author considers a much larger set of institutions: as Doner (2009) puts it, ‘Different levels of development require goodness of fit between the tasks involved and the capacities of institutions – the norms, rules, and organizations that “govern” economic activities’ (p. 4). Interestingly, in his book, Doner explicitly locates his argument within the theoretical framework of new institutional economics, thus exemplifying a recent convergence between CIAED and neo-Northian works.

Varieties of capitalism

The second cluster of neo-institutional works in the field of socio-economic development consists of studies within the ‘VoC’ tradition, associated with Peter Hall and David Soskice’s (2001) seminal eponymous 2001 volume. In their book, Hall and Soskice aimed to develop a ‘new framework’ to think about cross-country institutional differences and similarities that could explain a broad range of issues, including the variety of economic performance across nations. Such framework relies on a now well-accepted notion in contemporary social sciences, as mentioned in the ‘Introduction’ section: the notion that institutions play a key role in bringing about social and economic phenomena. To that notion, the ‘VoC’ framework adds two hypotheses: first, the hypothesis that individual political and economic institutions are interrelated and form, together with other interrelating institutions, an institutional system that is characterized by ‘institutional interlocks’ fueled by ‘complementarities’ between institutions; second, the hypothesis that such interlocks make radical institutional change generally very unlikely – institutions and institutional systems are path-dependent (Hall & Soskice 2001). ⁷

In their volume, Hall and Soskice (2001) distinguish their framework from approaches that emphasize the function of institutions as, respectively, ‘socializing agencies that instill a particular set of norms or attitudes in those who operate within them’ (p. 5), sources of power that endow certain actors with additional resources based on authority, or a ‘matrix of sanctions of incentives’ to which political and economic actors are called to respond to. By contrast, Hall and Soskice (2001) conceive of institutions as ‘conditioning’ the ‘strategic interactions’ deemed ‘central to the behavior of economic actors’ (p. 5). This conceptualization reflects Hall and Soskice’s aim to provide micro, game-theoretic foundations to a more macro theory of institutional change. Hall and Soskice (2001) also present their framework as a ‘firm-centered political economy’ – given their view of firms as the ‘crucial actors in a capitalist economy’ (p. 5). Thus, institutions provide ‘support for the relationships firms develop to resolve coordination problems’ (Hall & Soskice 2001: 9). ⁸ The kind of support Hall and Soskice (2001) have in mind consists in ‘providing capacities for the exchange of information, monitoring, and the sanctioning of defections relevant to cooperative behavior among firms and other actors’ (pp. 10–11). Such support-providing institutions include product and factor markets; the legal system; the financial system; and the industrial relations system. In addition, given the pervasiveness of distributive conflicts Hall and Soskice identify in the political economy of nations, ‘deliberative institutions’ also play an important role in bringing about coordination outcomes.

Although the VoC framework was originally proposed to re-think the different trajectories of change and economic performance among industrialized countries, it has subsequently been used to characterize differences in economic growth among developing and emerging economies as well: in Eastern Europe (Nölke & Vliegenthart 2009) or South America (Schneider 2009).

Neo-Northian economic approaches

The third cluster, which we have termed the ‘neo-Northian’ economic approaches to development (henceforth, NEAD), is even more heterogenous than the previous two. We call it ‘neo-Northian’ because of its strong theoretical links to the works of Douglass C. North and colleagues, and their brand of new institutional economics as applied to economic history (see, for a synthetic statement of this approach, North 1991). ⁹ Within the Northian framework, institutions do not only shape the economic incentives of individual and collective agents; they ‘become embedded in the assumptions and behavior of economic actors’ (Evans 2005: 95).

Two groups of works may be identified within this third cluster: a first one that may be called ‘hardcore neo-institutional economics’, which mostly focuses on and emphasizes a narrow set of indicators consisting of property rights, transaction costs and contracts. These indicators are supposed to, in line with writings by Ronald Coase, Oliver Williamson and Douglass North, encapsulate the key relationships between institutions and economic development. Such is the perspective taken by authors like Claude Ménard and Mary Shirley (see Shirley, 2005 and see Ménard and Shirley, 2005, 2014; Shirley, 2005, 2008). This approach is strongly market oriented. As Shirley (2005) writes, for instance, to

Meet the challenge of development countries need an institutional framework that supports the market economy, which includes two distinct and not necessarily complementary sets of institutions: (i) those that foster exchange by lowering transaction costs and encouraging trust, and (ii) those that influence the state and other powerful actors to protect private property and persons rather than expropriate and subjugate them. (p. 611)

However, despite the continuous insistence on pro-market institutions by authors, such as Shirley, the ‘institutional turn’ (Evans) in development theory has gone somewhat a long way since its Northian inception, in particular through development beyond ‘property rights’ as an originally central institutional mechanism, which, in Peter Evans’ (2007) words, ‘was forced to carry far too heavy an explanatory burden’ (p. 36). Indeed, as Evans (2007) further argues, ‘the idea that historically specific property rights arrangements can be arrayed along a simple ordinal scale is anything but plausible’ (p. 37).

A second group of works within the same broadly defined cluster presents a more sophisticated, and perhaps more durable, analysis of institutions, and in particular of the interplay between political and economic institutions. The work of Daron Acemoglu et al. (2001) is often cited for their emphasis of the role European colonization played in creating sets or systems of institutions more or less conducive to economic development. More important is, however, Acemoglu et al.’s (2005) work on political institutions, which are then causally related to economic institutions. In a 2000 study, Acemoglu and Robinson (2000) purported to show how democratization in countries, such as the United Kingdom occurred as elites attempted to face off threats of revolution in the 19th century. In particular, the extension of the franchise (through successive waves of democratization in the British case) was theorized by Acemoglu and Robinson (2000) as a commitment device used by elites to convince ‘the poor’ that the former were serious in their commitment to redistribute income through taxation. Acemoglu and Robinson then plugged this theorization of political institutions into a comprehensive framework aimed at explaining the relationships between institutions and development. This framework is represented in Figure 1, building on Acemoglu et al. (2005).

OPEN IN VIEWER

Figure 1.

AJR’s theoretical framework.

At one end of the framework – probably the most questionable – Acemoglu, Johnson and Robinson adopt the traditional Northian ideas: economic institutions, may them be property rights, labor market regulations, or financial regulations, shape the incentives of economic agents; these agents are thus encouraged to behave in certain ways and, in particular, invest certain amounts of economic (monetary and non-monetary) resources in certain activities; ¹⁰ economic performance ensues. If incentives are poor (good), then agents will make poor (good) investment decisions, which will result in negative (positive) economic outcomes. At that end of the argument, Acemoglu and colleagues do innovate by mentioning the impact of agents’ behavior on the future distribution of resources, which will feed back into the causal framework from the opposite end.

Working backwards, the theoretical framework proposed by Acemoglu, Johnson and Robinson proposes that economic institutions result from ‘political power’, which is implicitly assumed to be an attribute rather than a relation. Indeed, given the conflict of interests characterizing economic institutions (deriving from the latter’s distributive effects), it is not efficiency that guides the choice of economic institutions, but political power; and political power, in turn, depends on two factors: de jure political power is generated by existing political institutions, while de facto political power depends on the existing distribution of resources. Thus, there is a second important feedback mechanism in Acemoglu et al.’s model, in addition to the distribution of resources: political institutions, which affect political power and also results from the latter.

Furthermore, Acemoglu et al. (2014) have proposed to distinguish two kinds of (economic) institutions according to their impact on economic development: ‘inclusive’ institutions (such as private property rights), on one hand, which encourage innovation and growth and ‘extractive’ institutions, that extract resources from the majority of economic agents to the benefit of an elite. This dichotomous analytical framework developed by Acemoglu and colleagues somehow echoes Peter Evans’ (1995) distinction between ‘predatory’ and ‘developmental’ states – only the latter conform to the ‘embedded autonomy’ Evans views as instrumental to economic development. It stands at the center of Acemoglu and Robinson’s (2012) book, Why Nations Fail, conceived to present their ideas to the non-academic public. This dichotomy is Acemoglu, Johnson and Robinson’s (AJR) way to conceptualize the ultimate relationship between economic institutions and economic outcomes.

The three clusters’ key assumptions and theoretical claims are summarized in Table 1. All three clusters of neo-institutionalism have raised substantial amounts of criticism. The ‘hardcore’ neo-Northian approach to the role of institutions in economic development has been questioned for its axiological reliance on pro-market assumptions, as well as its neglect of reverse causality between institutions (see, for instance, Chang 2011; and, for an explicit Marxist critique of new institutional economics, Ankarloo & Palermo 2004). Neo-institutional approaches in general are regularly challenged by other approaches that downplay the causal centrality of institutions as envisioned by neo-institutionalists. Not only is geography often mentioned as an alternative cause for economic development (McCord & Sachs 2013; Mitton 2016). Institutional factors may be mitigated or reinforced by non-institutional factors as well. For instance, in a recent study on the economic development of former European colonies, William Easterly and Ross Levine (2016) show that ‘any adverse effects of extractive institutions associated with small European settlements were, even at low levels of colonial European settlement, more than offset by other things that Europeans brought, such as human capital and technology’ (p. 225). Finally, a growing empirical literature investigates the factors that may condition the impact of institutions on subsequent development. For instance, in a recent work, Bennett et al. (2017) have found that colonial settlement conditions and the identity of the colonizer were important factors in shaping the post-colonial institutional environment, which in turn relates to subsequent socio-economic development.

OPEN IN VIEWER

Table 1.

Summary characteristics of institutional analysis applied to economic development.

Approach	CIAED	VoC	Neo-Northian
			Hardcore NIE	AJR
Level of analysis	National economies	National economies/comparative	Comparative cross-country	Comparative cross-country
Institutional focus	State institutions and policies	Economic and political institutions	Economic institutions	Economic and political institutions
Institutions as	Power configurations/policies	Strategic framework	Rules of the game	Matrixes of incentives and constraints
Systemic properties of institutions	Not spelled out (government bias)	Interdependent (institutional interlocks) and path-dependent	Not spelled out	Not spelled out
View of economic development	Economic performance (growth, exports . . .)	Economic performance (growth, specialization . . .)	Economic growth	Economic growth

CIAED: comparative historical institutional analyses of economic development; VoC: varieties of capitalism; AJR: Acemoglu, Johnson and Robinson; NIE: new institutional economics.

However, the aim of the article is not to assess the theoretical or empirical validity of neo-institutional approaches to socio-economic development. It is, rather, to compare such approaches with Arrighi and Piselli’s and derive from such comparison possible suggestions for future research in the analysis of the causes of (under) development.

Epistemological divergences

Perhaps the least remediable, most definitive divergences between CDHE and neo-institutional works are epistemological divergences, that is, meta-theoretical and meta-methodological choices that radically differentiate the kind of contribution to social sciences these various authors purport to make. Three epistemological divergences stand out: (a) the choice of the dependent or outcome variable; (b) the role of geography and the level of analysis; and (c) the conception of causality, especially with regard to the role of contingency in historical processes.

The first divergence concerns the dependent or outcome variable – or, more profoundly, the fundamental research question tackled by Arrighi and Piselli, on one hand, and neo-institutionalist scholars, on the other hand. Arrighi and Piselli are concerned with the emergence of what they call a ‘social formation’ (i.e. wage labor), in correspondence with a specific type of development of productive forces, which they call ‘capitalist development’; furthermore, they emphasize the divergent and then convergent roads to wage labor within a broader locale (a region of Italy) that occupies a specific position within the world division of labor (peripheral). By contrast, neo-institutional scholars are mainly interested in economic performance, which is more or less broadly defined as long-term economic growth or growth in per capita income (Wade 1990), but may be approached as ‘prosperity’ (Acemoglu & Robinson 2012); or economic development, narrowly conceived. Some of the works focus on narrower measures of economic development, such as industrial innovation (Hall & Soskice 2001; Wade 1990).

Yet, as Arrighi and Piselli (1987) write in the conclusion of their essay,

[t]he concept of ‘economic development’ is often used to refer to two related but distinct processes. One is the process of social change, through which the organization of economic life of a given territory is transformed. The other is the process of economic progress, through which the absolute and relative command of the residents of that territory over economic resources (that is, their ‘wealth’) is increased. (p. 735 – italics in the original)

Some neo-institutionalists have attempted to broaden their take on development. Thus, Peter Evans (2005) has written that ‘the full flowering of the institutional turn depends on escaping [the] restrictive vision [of development]’, vision that relies on simple metrics such as the rate of growth of incomes or production (p. 95). As an alternative, Evans proposes to use Amartya Sen’s capability approach. ¹⁷ Even so, we would still be far away from Arrighi and Piselli’s concern with the qualitative transformation of social structures that may be associated with economic development.

This focus on different dependent variables creates obstacles for a commensurate treatment of CDHE and neo-institutional works, given (i) the complex relationship between ‘capitalist development’ and ‘economic performance’ – as Arrighi and Piselli (1987) argue, ‘different ways of organizing economic life [. . .] have no necessary relation to economic progress’ (pp. 735–736); ¹⁸ and (ii) the different methodological requirements for addressing each research question – in particular, given the higher degree of analytical nuance required to explain the emergence of particular social formations at particular moments of time.

The second epistemological divergence between the two perspectives concerns the level of analysis. CDHE, and Arrighi’s (1990, 2009) production more generally, is firmly rooted in a world-system analysis, whereby capitalism is seen as ‘an evolutionary system in which the stability of the whole is premised on the perennial change in and of the parts’ (p. 15). Moreover, the core-periphery asymmetry that characterizes the world-system is not merely a feature or the outcome of unequal exchange or the particular configuration of trade networks: it is ‘the systemic outcome of the perennial gale of creative and not-so-creative destruction engendered by the struggle over the benefits of the world division of labor’ (Arrighi 1990: 15). Finally, capitalist development ‘continuously isolate a minority of activities, actors, and locales from the competitive pressures of the world-economy, and simultaneously intensify those pressures on all other activities, actors and locales’ (Arrigi & Piselli 1987: footnote, p. 687). ¹⁹ In the case of Calabria analyzed by Arrighi and Piselli (1987), this has two effects: first, the peripheral position of Calabria was sustained by a continuous flow of factors and resources between Calabria and the rest of the world-economy; second, these flows also sustained the three different social formations during the first half of the 20th century: ‘their viability was dependent on the flows of labor that linked them’ (p. 685).

By contrast, neo-institutional analyses of economic development and growth mostly consider national states, or upper or lower units (see, for instance, Mitton 2016), as autonomous economic systems: systems with greater or lower connection to the world markets, but whose set of institutions (in particular political institutions) is (i) coherent ²⁰ and (ii) shielded from the country’s wider geo-economic position. This has two consequences: first, neo-institutional analyses almost completely neglect a central element in the political economy of world-system analysis: namely, the cross-country or trans-regional flows of goods, capital, and labor that so profoundly shape, according to the latter’s perspective, the trajectories of national or sub-national economic development. ²¹ Such flows, we may add, matter even beyond the core-periphery dichotomy emphasized in traditional world-system analysis. In fact, Arrighi and Piselli carefully insist on the differential importance of these flows in shaping the development of similarly peripheral areas of the world-economy. The second consequence lies in the different treatment of causality: institutions cause outcomes regardless of the position of the unit of analysis within the world-economy. This assumption is shared, in particular, across VoC and NEAD studies. In Hall and Soskice (2001), for instance, globalization is viewed as a competitive game pitting nations against other nations, which draw on their institutional resources to offer the best response to external challenges: ‘much of the adjustment process will be oriented to the institutional recreation of comparative advantage’ (p. 63).

A third epistemological divergence concerns the role of history and contingency in the trajectory of economic and social development. Here one may identify three different approaches: that of Arrighi and Piselli; a first, contingent-sensitive model of causality in comparative historical institutional analyses (CIAED and VoC); and a second, more structural approach to causality in cross-country institutional analyses (NEAD). In the case of CDHE, its two authors explicitly incorporate contingency into their theoretical framework; the three roads to wage labor in the early 20th-century Calabria represent, indeed,

the different outcomes of the struggles of landowners and other surplus appropriators or would-be appropriators to establish large-scale commodity production in response to a particular stage of development and to a particular conjuncture of the world-economy on the one hand and to the formation of the Italian nation-state on the other hand. (Arrighi & Piselli 1987: 685)

As hinted above, there are significant differences between neo-institutionalist approaches as far as the treatment of history is concerned. Virtually, all contemporary neo-institutionalist studies of economic development incorporate time variables in their analysis – even works within the NEAD tradition, which is more structural than the other neo-institutional variants. However, time does not equate history. The latter is laden with contingencies that cannot be easily conceptualized within a theoretical framework. Even the most ‘historical’ versions of neo-institutionalism still purports to make a contribution to social science, predicated upon timeless (or, rather, contextless) generalizations. For instance, in a 2010 volume on institutional change not specifically related to any of the neo-institutional traditions in the study of economic development, political scientists James Mahoney and Kathleen Thelen (2010) write that ‘institutional change occurs precisely when problems of rule interpretation and enforcement open up space for actors to implement existing rules in new ways’ (p. 4). Thus, history is just a (sometimes deeply researched) canvas against which theoretical models of (institutional) change and causality are tested. Social scientists are no historians, very much so as historians are not social scientists. There is, in the historians’ work, as E.P. Thompson (1995 [1975]) famously put it, a ‘distinct logic, appropriate to the historian’s materials [. . .]’, ‘[. . .] designed as far as possible to test hypotheses as to structure, causation, and so on’ (pp. 61–62). While Thompson, in his 1978 pamphlet, took aim at the works of Marxist structuralists, there is a brand of neo-Marxist scholars, such as Giovanni Arrighi himself, whose incorporation of historical contingency goes a long way toward bridging the gap between social sciences and history – much further in any case than neo-institutional scholars.

These divergences not only pit CDHE starkly against neo-institutionalist works but also make comparison and dialogue difficult, as each ‘side’ may see itself engaged in scientific efforts that are implicitly incommensurable with others. As Arrighi (2009) himself reflected at the end of his long career, in an interview with David Harvey published in 2009, he ‘abandoned abstract modeling for the concrete, empirically and historically grounded theory of social anthropology’, thus beginning his ‘long march from neo-classical economics to comparative-historical sociology’ (p. 62). These stark epistemological divergences feed into methodological divergences that are, however, less clear-cut.

Methodological differences

They are two methodological differences between CDHE on one hand and neo-institutional works on the other; although, as will become immediately clear, the differences are greater within the various neo-institutional approaches presented in the ‘Institutions and development: current debates’ section. The first difference is the treatment of time. As pointed out above, Arrighi and Piselli’s study is much more sensitive than all neo-institutional approaches to contingency. In addition, their analysis of the 100-year transformation of ‘social formations’ in Calabria is throughout framed by (what they see as) the actual historical phases undergone by such transformation. ²² In other words, Arrighi and Piselli focus on change; while neo-institutionalist scholars are interested in the generation of specific outcomes. ²³

The second difference regards empirical data: both its nature and treatment. Given the different focus of CDHE and neo-institutional works, it may not come as a surprise that the former contains none of the more or less sophisticated methodologies used to treat large sets of quantitative data, especially widespread in current works within the neo-Northian tradition (see for instance, Mitton 2016; Acemoglu et al. 2001). Again, as reminded above, Arrighi, while trained as an economist in a neo-classical mold in the late 1950s, broke free early from such a tight framework; his joint work with Piselli on CDHE builds on qualitative sociological and ethnological studies of several communities in the Calabrese hinterland.

However, here the differences may not be that insurmountable. First, there are wide methodological differences within neo-institutional approaches as well: the qualitative, small-n comparisons traditionally used within the CIAED approach are also at odds with quantitative, large-n studies, popular both within VoC and NEAD approaches. More importantly, however, in the latter two, an increasing number and proportion of studies actually rely on mixed methods including, besides large-n quantitative analysis, small-n longitudinal work. Mixed methods are, indeed, more in sync with complex causal frameworks that aim to capture historical transformations. Indeed, as James Robinson (2013) has argued, ‘many institutions come as a bundle’ (p. 29); therefore, (neo) institutional works require a methodological mix including carefully crafted case studies alongside quantitative analysis of large datasets.

Theoretical disagreements

Given the absence of any direct confrontation or cross-references between, on one hand, Arrighi and Piselli’s essay (together with Arrighi’s subsequent works) and, on the other hand, neo-institutional analyses of socio-economic development, the multiple theoretical differences one may identify between the two approaches are mostly implicit. In addition, these differences are compounded by the varying degree of theorization in these different works. For instance, early neo-Northian works overly relied on limited theoretical notions, such as property rights in general. As Peter Evans (2007) has observed about works such as those by Acemoglu and Robinson, the latter are seemingly aware of ‘the gap between their measure and their theory and the extent to which the concept of “property rights institutions” is underspecified’ (p. 38).

Beyond this arguably minor issue, there are at least four major theoretical disagreements between the two approaches. The first theoretical disagreement concerns the causal operation of institutions. What do institutions do? In most neo-institutional accounts examined in the ‘Institutions and development: current debates’ section, this is very clear – and in line with Douglass North’s (1991) famous proposition that institutions are ‘the humanly devised constraints that structure political, economic and social interactions’ (p. 97). More broadly – and in accordance with North’s later views on the subject, institutions operate by shaping individual or collective agents’ behavior, through (i) constraints; (ii) incentives; and (iii) expectations. The third mechanism corresponds to the more recently explored nature of institutions as shared meanings (Denzau & North 1994; see also von Staden & Bruce 2015). Together, these three mechanisms form what may be conceptualized as a ‘functional-internal’ view of institutions. Arrighi and Piselli are much less explicit in their view of the causal nature of institutions (for the reasons stated in the ‘Institutions in Arrighi and Piselli’s essay: a missing link?’ section); they do imply, however, that both economic and political institutions offer temporary solutions to unstable or conflictual relations of production or social structures by stabilizing behavior. Theirs may be recast as a ‘functional-external’ view of institutions – a view that pays more attention to the establishment of institutions than to their effects, which are implicit in their establishment. In the new institutional economics literature, a similar opposition may be found between an ‘institutions-as-rules’ view (associated with North) and an ‘institutions-as-equilibria’ view. The key distinction between the two views is the nature of enforcement: exogenous (to institutions) in the former, endogenous in the latter (Greif & Kingston 2011). Despite the fact that CDHE does not, of course, share the rational-choice foundation of the ‘institutions-as-equilibria’ view, one could consider the two views to similarly consider enforcement as endogenous. However, here again the difference may not be that large, given that the current new institutional economics often combines the two views (Greif & Kingston 2011); and given that CDHE does include institutions in a wider causal argument similar to neo-institutional approaches. If indeed, as Arrighi and Piselli (1987) argue, the path of social change ‘determine[s] the distribution of such command within the territory, and therefore the welfare of its population’ (p. 36, italics in the original); then, institutions sustaining particular paths of social change may, indeed, be correlated with particular distributive and redistributive outcomes. Here the gap between the neo-Marxist framework presented by Arrighi and Piselli, on one hand, and neo-institutional works, on the other hand, is not so hard to bridge. ²⁴

A second implicit theoretical disagreement concerns the role and the identity of the actors of the institutional world. Despite the crude modeling of a country’s social structure in the early works by Acemoglu and Robinson (2000), the latter may be more congenial to Arrighi and Piselli’s approach than the other two clusters of neo-institutional works identified in the ‘Institutions and development: current debates’ section. What kinds of social agents matter for the establishment of institutions or the generation of institutional change? Both VoC and CIAED seem to have definitive answers. In VoC’s case, Hall and Soskice choose, as discussed in the ‘Institutions and development: current debates’ section, to emphasize firms’ coordination problems. In Arrighi and Piselli’s approach, by contrast, firms are not privileged actors: the relevant actors are social groups that may or may not operate and interact from within economic agents such as firms. In other words, the importance of firms as economic and political actors is contingent upon the particular institutional or historical configuration. The same assumption regarding pre-constituted sets of relevant actors surfaces in CIAED. As Peter Evans (1995) candidly observed in his book, ‘[i]n developmental states, connectedness has meant ties with industrial elites’ By contrast, neo-Northian works do not make assumptions about key actors in their social world – probably because actors are thought out as abstract agents, whose position in the specific relations of production or the social structure is never spelled out. Here, a potentially minor theoretical disagreement between CDHE and one brand of new institutionalism actually reveals a much deeper epistemological chasm.

The third theoretical disagreement follows from the second one, and concerns both the autonomy of institutions and the interrelations between institutional and non-institutional factors. Regarding the first issue, many neo-institutional works distinguish between political and economic institutions – varying degrees of autonomy are then attributed to each kind. In the neo-Northian analyses of economic development that give theoretical preeminence to political institutions, economic institutions appear as mere instruments of the latter (Acemoglu et al. 2005). If we could plot various institutions on an instrumental scale – with the least instrumental on one end, and the most instrumental on the other – economic institutions would be located on the opposite end to political institutions. This ‘over-determination’ of economic institutions is problematic in that it assumes a specific causal relationship between the two types of institutions instead of leaving the relationship to the empirical work – there is clearly a difference with the prospective of Arrighi and Piselli, as can be seen by confronting Figures 1 and 2.

In addition, the ‘least instrumental’ treatment of political institutions by Acemoglu and colleagues is closely and axiomatically associated with an elevated degree of agentic autonomy. In other words, wherever institutions are conceived as sources, rather than outcomes, of collective behavior, they are supposed to operate with an elevate degree of autonomy. This does not have to be institutions that may operate as agents (and shape other agents’ behavior) with varying degrees of autonomy from other agents’ behavior. To be fair, not all neo-institutional approaches share this rigid causal hierarchy between economic and political institutions. ²⁵ However, when considering the broader set of relations between institutional and non-institutional factors, most neo-institutional works, across the three clusters identified in the ‘Institutions and development: current debates’ section, share the view that institutions may be effective precisely because of their operational autonomy from non-institutional forces. ²⁶ The argument of autonomy logically implies a historical theory of institutions, even when it is actually underdeveloped in many studies. For instance, Evans (2007) compares the perspectives by Acemoglu and colleagues with that proposed by political scientist James Mahoney, emphasizing the latter’s thesis that institutions ‘emerge out of uncertain, politically motivated choices, made primarily during “critical junctures” when development possibilities are in flux’ (p. 43). Then path dependence kicks in, and institutions produce effects autonomous from the causes of their emergence.

By contrast, in CDHE institutional factors never operate in isolation from non-institutional factors. This appears more clearly perhaps in Arrighi and Piselli’s analysis of the post-World War II transformation of capitalist development in Calabria, where three key factors are singled out: the active role of the Italian State, successive waves of mass migration, and social conflict. The latter two are, evidently, non-institutional. In neo-institutionalist works, given the causal preeminence given to institutions, the interrelationships between institutional and non-institutional factors is asymmetrical. By contrast, in CDHE the reasoning is much more circular (less linear), and the question of autonomy is not tackled upfront. In addition, CDHE does not contain an explicit theory of the durability of institutions – institutions are not path-dependent, and given the high degree of instability of relations of production and the high degree of contingency in the trajectories of change, institutional change may be radical.

This difference overlaps with the fourth major theoretical disagreement one may, at first, find between CDHE and neo-institutional approaches: the question of endogeneity and conflict. In ‘hardcore new institutional economics’ approaches, economic institutions are mostly exogenous – this view is clearly at odds with Arrighi and Piselli’s, since, as it has already been pointed out above, CDHE presents an analysis of institutions entangled within a tight causal web of institutional and non-institutional factors. However, other neo-institutional works are keen to endogenize institutions – thus, explicitly incorporating into their analytical framework, the distributional aspects of economic institutions, together with the ‘power-shaping’ aspects of political institutions. As Acemoglu et al. (2005) note in their work, ‘[t]he distribution of resources in society is an inherently conflictual, and therefore political, decision’ (p. 394). In the neo-Northian works, endogeneity reaches a high level that makes Peter Evans (2007) observe that ‘[i]t might even be argued that, for Nugent and Robinson, institutions are not real “causes” at all but simply “transmission belts” which instantiate the effects of pre-existing economic and political interests’ (pp. 41–42).

This view does not seem far remote from the view, exposed in CDHE, that institutions develop differentially according to pre-existing conditions. In the case analyzed by Arrighi and Piselli (1987), the three roads to capitalist development ‘represented different outcomes of the struggle of capitalists and would-be capitalists [. . .] to overcome [sociological and ecological] obstacles [to profit-oriented activities]’ (p. 681). In other words, the varying ‘social ecology’ of the terrain determined the institutional configuration of particular social formations. In Acemoglu and Robinson’s writings, geography, too (but not the social factors adjoined to geography by CDHE), determined the viability of particular (political) institutions.

However, two further theoretical elaborations on this issue bring forward, again, this theoretical disagreement about endogeneity. First, the endogenous nature of political institutions leads some neo-institutional scholars to analyze the distributional conflicts that generate institutional responses. One may, for instance, superficially consider Acemoglu et al.’s theoretical construction of political institutions (or political institutional change) as endogenous solutions to social conflicts (see, for instance, Acemoglu & Robinson 2000, 2001). However, these social conflicts are mostly potential conflicts; institutions are built/changed precisely so as to avoid social conflict. By contrast, in CDHE political and economic institutional change reflect and result from both actual social conflicts and potential ones. The highly abstract notion of social conflict in neo-Northian works reflects their rational-choice roots; while the privileging of certain pre-defined actors in VoC and CIAED prevent the latter to come up with an analysis of the relationship between social struggles and institutions as sophisticated as Arrighi and Piselli’s.

Second, and finally, the similar degree of institutional endogeneity displayed by CDHE and neo-institutional approaches cannot dispel one more difference, which lies in the ‘voluntaristic’ nature of institutions – at least in neo-Northian works (in line with the rational-choice framework, mentioned above), and in some of CIAED as well. Institutions can be chosen, in a quasi-deliberative process that selects the best (i.e. more sustainable over time) institutional solutions to coordination problems or distributive struggles. As Peter Evans (2007) argues, ‘[d]enying the possibility of political choice would negate the very historical narratives on which the extension of the institutional turn is predicated’ (p. 49). Of course, ‘political choice’ may be understood in various ways; but the intentional or voluntaristic act of choosing an institution seems at odds with a neo-Marxist understanding of institutions such as the one permeating CDHE. However, here again, all depends on the notion of ‘choice’ one has. Clearly, Arrighi and Piselli’s work contrasts with the notion that institutions may be chosen in a pre-set and deliberative context. But local economic institutions, in CDHE, may be thought of as the crystallized forms of collective choice – choice expressed through a variety of mechanisms such as voice (conflict), exit (migration), or loyalty (obedience and submission to a repressive apparatus).

A room for dialogue?

Given all these divergences, differences and disagreements, is there any room for a (constructive) dialogue, that is, a dialogue that may inform and improve our current social scientific explanations of how institutions affect socio-economic development? The position held in this article is that there are, indeed, avenues for a fruitful dialogue between the two perspectives. These avenues cut across the broad, ongoing debates within neo-institutionalism itself: as hinted above, many of the methodological and theoretical differences highlighted between CDHE and neo-institutionalists actually exist between the various clusters of neo-institutionalism as well. In addition, the evolutionary nature of social scientific knowledge enables to see new possibilities for theoretical developments within neo-institutionalism that bring us back to the kind of perspective delineated by Arrighi and Piselli’s (1987) essay. The next section aims to provide a very brief review of the spaces for such encounters, and the questions these encounters may help address.

How CDHE may contribute to the improvement of neo-institutional analysis

A first path for a neo-institutional approach based on foundations laid by Arrighi and Piselli concerns the causal role of institutions within world-system dynamics. Indeed, institutions may not necessarily simply and linearly reflect elements characterizing locales/local units of analysis, as is implied in most of the neo-institutional literature. They may, by contrast and in line with CDHE, act as interfaces between world-economy and economic development – or as the double locus where (i) local relations of production interact with the local social structures and (ii) local relations of production and social structures interact with world-economy dynamics and macro level political institutions. A world-system perspective may enrich neo-institutional analyses of change by framing new research developments. First, the process of peripheralization may entail different kinds of institutions (with respect to core regions) and institutionalization processes. Second, the degree and nature of institutional endogeneity may actually vary with the locale’s position within the world-economy.

Third, and perhaps more importantly, specific institutional resources may help offset peripheralization:

Key to counteracting peripheralization is the capability of a given actor or locale to participate in as wide a division of labor as possible and still be able to isolate itself temporarily from the competitive pressures of the world-economy. (Arrighi & Piselli 1987: 696)

Rather than the problematic dichotomy between ‘extractive’ and ‘inclusive’ institutions used in a linear causal model to explain (the lack of) prosperity (in the neo-Northian works of Acemoglu and colleagues), one may conceive a continuous distinction between institutions on the basis of their ability, contingent on the specific political and economic context, to counteract peripheralization, or to assuage the effects of peripheralization on local social structures. Institutions, in other words, may help tame the ‘circular and cumulative’ causation identified, since Myrdal, with peripheralization.

A second area where neo-institutional analyses of economic development may benefit from the analytical framework presented in CDHE is the area of institutional change. As mentioned in the ‘Institutions in Arrighi and Piselli’s essay: a missing link?’ and ‘Arrighi and Piselli versus new institutionalism: a meaningful confrontation?’ sections, Arrighi and Piselli’s understanding of institutions is essentially dynamic: institutions crystallize the evolving social structure and relations of production at time t. Thus, an understanding of institutional change is immediately related to the continuous transformation or structural imbalance of social structures and relations of production within capitalism. The key role of social struggles within Arrighi and Piselli also provide accounts of institutional change with a powerful causal engine: Institutionalization can be seen as the crystallization and near-complete autonomization of distributive and redistributive struggles. The same analytical building block in CDHE – that is, the interrelationship between institutional and non-institutional factors – may also be used to shed light on a separate but related concern in the neo-institutional literature: namely, the need to distinguish between de facto and de jure (political) institutions (Voigt 2013). As Foldvari (2017) has shown in a recent study, the tightness of the relationship between de facto and de jure political institutions may actually vary over time. The appearance of gaps between de facto and de jure institutions may be attributed to, respectively, the occurrence of strengthening of distributive or distributive struggles; or the unraveling of the world-economic bases of local economic institutions.

How neo-institutionalist insights may improve our understanding of capitalist development

Institutional change is also the terrain where the questions addressed by Arrighi and Piselli may be re-framed using neo-institutional concepts and theories. This potential contribution of the latter owes to the greater tendency of the most recent neo-institutional works to dynamically analyze the interrelationship between development and institutional change. ²⁷ In particular, new institutionalism in political science has now tackled with great analytical detail the dynamics of institutional change (Mahoney & Thelen 2010). However, the path dependency of political and economic institutions may shed new light on the later stages of capitalist development in Calabria, where socio-economic institutions may reflect past structures and thus contain a potential repertoire of behavioral responses to external stimuli or shocks. Institutions, in other words, may be imbued with certain characteristics that shape their evolution (or lack thereof) over time – an autonomous (to a certain degree) institutional logic that interferes with changes in the social structures or relations of production. Regardless of the specific causal role attributed to institutional factors in the first place, neo-Marxist accounts of ‘capitalist development’ may benefit significantly from incorporating in their analysis, the insights provided by the now-large literature on institutional change – which, to be fair, is not central in neo-institutional analyses of economic development at the core of the present study. In particular, institutions may be more explicitly treated as key components of the ‘hostile environment’ peripheral areas of the world-economy may find themselves confronted with; thus the processes of social change associated with capitalist development in such areas may be more explicitly linked to the difficulties associated with combating institutional inertia.

Furthermore, neo-institutional insights may be applied to the broader framework delineated by Giovanni Arrighi in his later, post-CDHE works. In particular, the analysis of long-term shifts in macro regimes of accumulation presented in The Long Twentieth Century may benefit from more sophisticated notions of institutional change offered by the neo-institutional literature. Thus the ‘stickiness’ of one hegemonic system between the ‘signal crisis’ and the ‘terminal crisis’ of a given regime of accumulation (Arrighi 2010 [1994]) may be attributed to institutional path dependence or ‘institutional drift’ (Mahoney & Thelen 2010), whereby institutions conceived to perform a certain function gradually and tacitly undertake a fundamentally different role.

Another, related potential avenue for a fruitful integration of neo-institutional concepts and methods into an Arrighian framework concerns the hypothesis of ‘institutional interlock’ found by Hall and Soskice (2001) to be at the basis of institutional status at systemic level. In a nutshell, institutional stasis occurs when individual institutional change becomes hindered by strong interdependencies between individual institutions. It is this varying degree of interdependence that may increase or decrease individual institutions’ resistance to social change and, ultimately, capitalist development. ²⁸

A second area where a more active dialogue between the analysis presented by Arrighi and Piselli in CDHE and neo-institutional approaches lies in the exploration of the informal aspects of the process of institutionalization. As pointed out above, Arrighi and Piselli do lend some importance to informal institutions when they discuss the role of the new norms of behavior adopted by migrant workers from Calabria – such as the ‘consumption norm’ in the 1960s – that then participate in the unfolding of the old relations of production and a model of development (or survival) centered around land in their region of origin. We may extend this insight and look for other, similar informal institutions (norms of behavior) that have accompanied the century-long transition to wage labor in Calabria, either on their own (i.e. self-standing norms of behavior) or in association with formal institutions. For instance, the growing presence of the state may, at some level, trigger a process of institutionalization through which certain norms of behavior are internalized. Yet, in Arrighi and Piselli these aspects are not dealt with. To a certain extent, then, one may wish to enrich Arrighi and Piselli’s analytical framework by incorporating an explicit and systematic treatment of changing norms of behavior implied by the process of institutionalization. Such an extended framework may benefit from the analytical insights generated by yet another brand of neo-institutionalism, what we may call ‘cultural’ or ‘cognitive’ institutionalism (see, for instance, Conrad 2006), which may explain, again, the persistence or stickiness of both certain formations and the strategic choices made by key actors.

Of course, the latter also suffers from key shortcomings and cannot therefore be exhaustively relied upon. In particular, it mostly ignores the contradictions between cultural institutions, or groups of norms and behavior associated with development. Norms of behavior may, indeed, collide – precisely when new norms are imported from another locale, such as was the case in the consumption norm mentioned above, clashing with old norms – such as the one attributing to land (its ownership, its possession, and its cultivation) a key role in sustaining households’ living standards. But an extended version of the CDHE framework may precisely deal satisfactorily with this issue. Another limitation of cultural institutionalism is the sidestepping of distributive conflicts at the origins of institutional emergence. This, however, is a characteristic more or less common to other brands of neo-institutionalism that may be accurately dispensed with through the detailed analysis of social struggles that characterizes CDHE.

A final area where neo-Arrighian or, more broadly, world-systemic analyses of economic development may gain from more direct engagement with the neo-institutional literature is the area of choice/strategy. Again, this does not imply that Marxian scholars should turn into rational-choice theorists; however, it may be useful to explicitly address the issue of strategic behavior in the formation, development, and supersession of both political and economic institutions. In particular, political institutions may be instrumental in setting the strategic expectations of actors of social change. Given the centrality of social conflict (class struggle) in determining the process of social change in CDHE, it may be useful and interesting to analyze the ways in which political institutions, in particular deliberative ones, have governed the flows of labor and capital surplus between Calabria and other Southern regions and the rest of Italy, in the years following the period studied by Arrighi and Piselli.

As mentioned above, neo-institutional analysis of strategic interactions between institutions and social actors may be unnecessarily burdened by an assumption of ‘free choice’ by autonomous, atomized actors. But Marxian scholars or comparative historical sociologists need not accept that assumption. On the contrary, one may explicitly historicize the deliberative world often assumed in neo-institutional works – a deliberative world that may, at specific points of time, have an effect on social actors’ capacity to shape their institutional context.