Institutions and economic performance: the state of the art

Research on the empirical determinants of economic growth has exploded in the last decades. Indeed, hundreds of international studies have been undertaken to approach the ever-important but elusive question of what causes growth and the prosperity of nations.Footnote 12 Economists have traditionally focused on the effects of physical and human capital accumulation, total factor productivity, technological innovation, the process of knowledge creation and diffusion, and international economic integration.Footnote 13 However, it has been increasingly recognized that politics and institutions are central to the process of economic growth because they affect the incentives to accumulate, innovate and accommodate change. In this context, an active research programme has sought to model and test empirically the political determinants of growth. This section discusses one of the most fertile research streams in the political economy of growth: the literature on institutions, property rights and economic performance.Footnote 14

The first attempt to combine economic and political analyses to account for the determinants of economic growth studied the effect of political regimes on economic performance. Research looking at whether democracy promotes or hinders economic growth has produced three schools of thought: (1) the conflict school, arguing that democracy hinders economic growth, mainly in less developed countries, by creating consumption pressures, fuelling distributional conflicts and inhibiting capital accumulation; (2) the compatibility school, contending that democracy enhances economic growth, because the existence of fundamental civil liberties and political rights generates the social conditions most conducive to economic development; and (3) the sceptical school, claiming that there is no systematic relationship between democracy and economic development.Footnote 15 Typically, the advocates of the conflict school used the experiences of countries such as South Korea, Taiwan and Chile as evidence that ‘good-for-growth dictatorships’ can create the right conditions for growth by providing the sources of political order and social control. However, this argument seems to find little support in global statistical studies. For example, Przeworski, Alvarez, Cheibub and Limongi examine the experiences of 135 countries from 1950 to 1990 and conclude that ‘there is no trade-off between democracy and development, not even in poor countries’.Footnote 16

Although Przeworski et al. show, quite compellingly, that there is no evidence to support the claim that democracy needs to be sacrificed in the name of economic development, the empirical evidence on the relationship between political regimes and economic growth is still ambiguous. In a comprehensive review of the literature, Yi Feng argues that:

Having acknowledged that political regimes were not a key source of variation in observed growth rates, political economists began to investigate the economic effects of institutions and governance structures.Footnote 18

The ‘institutional turn’ in the political economy of growth gained momentum in the mid 1990s, especially with the publication of two pioneering articles: Stephen Knack and Philip Keefer’s ‘Institutions and Economic Performance’ and Paolo Mauro’s ‘Corruption and Growth’. The key contribution of these articles is that they develop novel measures of institutions and property rights, signalling the beginning of a new generation of econometric research on institutions and economic performance in an cross-national setting.Footnote 19 In contrast to the literature on political regimes and growth, the claim that ‘institutions do matter’ has received strong empirical support. Building on data for up to ninety-seven countries for the period 1974–89, Knack and Keefer find that the quality of institutions, operationalized as the security of property rights and the level of contract enforcement, is crucial to growth and investment.Footnote 20 Similarly, Mauro shows that subjective indexes of corruption are negatively linked with investment and economic growth.Footnote 21 In line with these results, Alesina demonstrates that institutional quality, as measured by bureaucratic efficiency, absence of corruption, protection of property rights, and the rule of law, is important for growth.Footnote 22 In the last decade, a large number of studies have provided empirical support for the proposition that institutions are a fundamental source of differences in economic development.Footnote 23

Unquestionably, this literature represented a breakthrough in the business of modelling and measuring institutions in the context of cross-country growth regressions. As far as analytical implications are concerned, its findings seem to provide empirical support for the ideas of North and Olson, who stress the role of secure property rights and effective contract enforcement in creating the political conditions for growth and prosperity. Moreover, the discovery of a link between institutions and growth appears to address some of the anomalies of old and new growth theories, which have not been able to accommodate easily the following stylized facts: (1) the absence of general convergence, and (2) why the most rapid growth tends to occur in a subset of low-income countries.Footnote 24 The new conventional wisdom is that institutions are the primary cause of economic development. Thus, the name of the game is now convergence, conditional on good institutions. However, how robust are the conceptual and empirical foundations of the institutions–growth causal linkage?

Per capita income and the quality of governance are strongly positively correlated across countries.Footnote 25 Although most of the above-mentioned papers used several methods to minimize reverse causation problems, causality is always an issue in growth regressions.Footnote 26 For example, Chong and Calderon apply a rigorous approach of causality to measures of institutional quality and economic growth, finding strong evidence of causation running in both directions, from institutions to growth, and from growth to institutional quality.Footnote 27 By contrast, building on a new dataset and using non-sample information, Kaufmann and Kraay find no evidence of a positive effect of incomes on the quality of institutions, calling into question the often-heard argument that only wealthy countries can afford good governance.Footnote 28 Acemoglu et al. claim to capture the causal effect of economic institutions on prosperity by ‘instrumenting’ the quality of current institutions with original variables measuring the historical conditions in the European colonies.Footnote 29 However, Glaeser et al. question the validity of these instruments and show that ‘the evidence that institutions cause economic growth, as opposed to growth improving institutions, is non-existent’.Footnote 30 Not surprisingly, the evidence on causality is inconclusive and the debate remains open.Footnote 31

Omitted variable biases may also explain the strong correlation between good governance and high growth. For example, Sachs shows that geography, reflected in weather conditions and access to maritime trade routes, plays a major role in countries’ economic success.Footnote 32 Other studies have documented the importance of factor endowments,Footnote 33 colonial heritages,Footnote 34 legal traditions,Footnote 35 ethnic conflicts and social heterogeneity,Footnote 36 and religious composition.Footnote 37 Most critically, the institutional economics literature has overlooked the role of the state in resolving co-ordination failures and facilitating private–public sector co-operation in early stages of industrialization.Footnote 38 The Olsonean problem of how interest-group formation intersects with institutions in affecting economic growth is also missing in most analyses of the institutional sources of development.Footnote 39

In various studies, Acemoglu, Johnson and Robinson (AJR) claim that once one controls properly for the effects of economic institutions, neither geography, nor the identity of the colonial power, nor the ethnic and religious composition of the population, nor religion, are deeper determinants of income per capita.Footnote 40 However, AJR also recognize that the very institutions that are central in determining relative prosperity are endogenous, and that ‘a theory of why different countries have different economic institutions must be based on politics, on the structure of political power, and the nature of political institutions’.Footnote 41 In this vein, Helpman argues that AJR indeed demonstrate that ‘variations in local conditions’ faced by European colonizers ‘played an independent role in the emergence of institutions and in their impact on long-term development’.Footnote 42 In this context, there are two problems with the ‘primacy of institutions’ thesis. The first is that the combination of the two central tenets of new institutionalism (‘institutions matter’ and ‘institutions are endogenous’) leads to circularity. As Przeworski clearly put it, if the endogeneity of institutions is sufficiently strong ‘then institutions have no autonomous role to play. Conditions shape institutions and institutions only transmit the causal effects of these conditions’.Footnote 43 The second potential problem is conceptual stretching. The institutions-rule literature ‘tends to see institutions everywhere’, thus downplaying the critical importance of developments at the level of ‘society-rooted politics’.Footnote 44

Critically, remaining problems in terms of conceptualization and measurement also call into question the conventional wisdom on good governance and development. Like the concept of democracy, governance, and by implication good governance, is an ‘essentially contested’ concept.Footnote 45 However, though there is an established literature dedicated to assessing the validity of alternative ways of conceptualizing and measuring democracy,Footnote 46 the discussion of the analytics and methods of good governance is still underdeveloped.Footnote 47 On the empirical front, Mauro and also Knack and Keefer revolutionized the art of measuring institutions by introducing the use of subjective indicators of corruption, governments’ repudiation of contracts and bureaucratic efficiency produced by private risk assessment agencies.Footnote 48 However, the indicators used to establish the proposition that institutions cause growth are constructed to be conceptually unsuitable for that purpose.Footnote 49 North defines institutions as the ‘humanly devised constraints that shape human interaction’.Footnote 50 However, most perception-based measures of institutional quality are not measuring institutions as constraints, but institutional outcomes/performance. Glaeser et al. show that widely-used governance indicators are (1) too volatile to reflect more or less permanent features of the political environment and (2) tend to rise with per capita income.Footnote 51 In their critique of the World Bank Governance Indicators, Kurtz and Schrank argue that the dominant measures of governance suffer from perceptual biases, adverse selection in sampling and conceptual conflation with policy choices.Footnote 52

Many empirical studies suggest that institutions do matter for economic performance. However, I have shown in this section that the proposition that good governance causes economic development can be contested on conceptual and methodological grounds. The econometrics of institutions and growth is still haunted by causality uncertainties, omitted variable biases, modelling/specification deficiencies, and most importantly, by problems of conceptualization and measurement. This should lead to serious concern about the empirical foundations of the prevailing ‘governance matters’ consensus. It may also lead to scepticism about the power of large-N cross-national studies in offering insights into the institutional foundations of economic development.Footnote 53 However, I believe that, whatever the remaining empirical problems, the main deficit of this research programme is analytical. Specifically, it relates to our imperfect understanding of how politics interacts with institutions in shaping development policies and outcomes. As Helpman and North highlight, we lack a developed body of theory capable of linking polities, institutional change and economic performance.Footnote 54 In the next section, I will discuss the analytical gaps in the comparative literature on institutions and development.

Analytical gaps in the comparative literature

In principle, the literature on institutions and growth can be criticized from two different analytical positions. On the one hand, one can question the validity of the governance-causes-growth consensus by challenging the core assumptions of new institutionalism. On the other hand, one can discuss the analytical problems of the literature by seeking further integration between growth empirics, institutional theory and political economy. In this section, I adopt the second strategy. I identify the analytical lacunae of existing empirical research by drawing lessons from leading scholars working in the political economy of institutions research tradition.

Policy implications of the good governance approach

The institutional political economy of growth is a research programme with very significant public policy implications. In this section, I will discuss the policy implications of the literature on institutions and economic performance in the light of the analytical critique presented above. In doing so, I will assess the plausibility of the good-governance approach to development implied in the discourse and actions of international organizations.

In the last decade, governance and, increasingly, good governance ‘have permeated development discourse and especially research agendas and other activities funded by public and private banks and bilateral donors’.Footnote 105 The good-governance approach to development implies the application of the institutions-do-matter proposition in the realm of development research and policy. The name of the good governance game is ‘convergence conditional on institutions’: relatively poor countries can enjoy rapid catch-up growth if they adopt good institutions. The growing attention to governance issues has to be understood in the context of the neo-institutional revolution in social sciences, on the one hand, and the concomitant proliferation of empirical studies claiming a positive link between institutions and development outcomes, on the other. Yet this story is incomplete without an understanding of the political context shaping the conditions for the rise of the good governance discourse in international public policy. Actually, an explicit concern with good governance and its implications for development was aroused in the framework of the reformulation of the Washington Consensus, prompted by some of the disappointing outcomes of the first generation of market-oriented reforms.Footnote 106 As Daniel Kaufmann acknowledged: Footnote 107

What are the policy implications of the comparative literature on good governance and development? Alberto Alesina’s article ‘The Political Economy of High and Low Growth’ is a good example of an empirical work on institutions with strong normative implications. Using information from a sample of 102 countries for the period 1960–90, Alesina claims that ‘institutional quality – as measured by bureaucratic efficiency, absence of corruption, the protection of property rights, and the rule of law – is important for growth’.Footnote 108 In the light of this finding, he makes the case for ‘institutional conditionality’:

Institutional conditionality is based on (1) the belief that, under certain conditions, good institutions are a factor influencing economic development (the institutionalist claim), and (2) the expectation that cutting off assistance to countries with poor governance records may create incentives for institutional strengthening via moral hazard considerations (the conditionality claim). The methodological validity of both claims can be questioned. In particular, one may object to the use of perceptions-based indexes of governance performance to demonstrate a causal relationship between institutions and growth. Similarly, it can be argued that such a strong normative inference is unwarranted given the limited time-frame of the study.Footnote 110 But let us assume, for the sake of the argument, that this study is free of any serious modelling/measurement problems. Even if this were the case, the conditionality claim would be problematic because moral hazard arguments do not travel well from textbooks to the reality of developing countries. Unintended consequences are likely to come into play. Given that the real incentive structure of a society is made up of formal, informal and cognitive constraints, it is difficult to predict the ultimate outcome of attempts to create institution-building incentives through the manipulation of financial aid. The challenge is to demonstrate that the cutting of assistance would increase the probability of governance-enhancing reforms becoming self-enforcing. As the history of foreign sanctions/interventions shows, the outcome very much depends on the way domestic politics mediates the process linking aid-related incentives with institutional reforms.Footnote 111

A critical issue here is the likely response of the political stability variable, which has significant effects in Alesina’s regressions. In some circumstances, one can expect that the application of conditionality may lead to the deterioration of both governance and economic performance through the political instability channel. This brings us to the work of Feng. This study stands as one of the most comprehensive and innovative approaches to the empirics of governance and economic performance. However, it also reveals the uncertainties associated with translating the findings of econometric studies into concrete lessons for institutional policy. The central policy message of this work is that ‘a nation can achieve “economic miracles” by reducing political instability, building policy certainty, and increasing political freedom’.Footnote 112 The puzzle is to identify what set of institutional reforms a misgoverned country should implement, and at which level, in order to strengthen the political foundations of its economic system. As suggested before, the real problem is that factors such as political stability and policy certainty are indeed outcome-variables, reflecting more the way in which the political game is played in different countries, rather than the actual institutional arrangements structuring that game.

Some of the measurement strategies chosen by Feng also suggest some pressing issues regarding policy implications. For example, he measures the level of policy certainty/uncertainty, one of the key variables driving his findings, using indicators of income distribution. The plausible assumption behind this choice is that a wide disparity of wealth polarizes societies and promotes conflict, thus undermining consensus about policies. This method of using measures of income inequality as proxies of governance-related concepts is common in the political economy of growth literature. There is nothing intrinsically wrong with such a measurement strategy provided it is supported by theory. The issue is, however, that a competing (i.e. less institutionalist) narrative about the political sources of underdevelopment can be constructed upon the observed relationship between income distribution and growth. Indeed, the problem of ‘same data, leading to different conclusions’ pervades the literature.Footnote 113 In this context, what are the actual policy lessons of Feng’s study? He finds, for example, that ‘political instability and policy uncertainty underscore the lack of economic performance in Latin America’.Footnote 114 However, we know that poor governance, high inequality and low rates of growth are inextricably linked in this region. Thus, it remains unclear whether Latin American countries should put the emphasis on institution-building or on delivering policies to tackle poverty and inequality.

Knack acknowledges that existing research showing that governance matters for development ‘does not often point the way toward specific reforms, because it is based largely on very broad and aggregated indicators of institutional performance’.Footnote 115 Aggregation might not be the real problem here. The real issue is that those aggregated indicators of governance are not even measuring institutions per se. Indeed, they are measuring at best ‘structure-induced equilibria’ and at worse ‘policy equilibria’. Take the example of corruption. Since the path-breaking work of Mauro,Footnote 116 several articles have documented that corruption has growth-retarding effects.Footnote 117 However, what are the policy lessons of this empirical finding? In many developing countries, high levels of corruption are equilibria induced not only by the formal rules of the game, but also (and fundamentally) by the incentives embedded in deeply rooted political practices, informal norms and players’ expectations. In addition, pervasive corruption is likely to flourish with the backdrop of poverty and inequality. In this context, it is not evident what kind of legal, institutional and public sector reform would be effective to curb corruption in developing countries.Footnote 118 As Rodrik highlights:Footnote 119

The case of corruption also underlines the role of special interests and distributional politics in conditioning the relationship between institutions and economic outcomes. In an explicit attempt at ‘rethinking governance’, Kaufmann recognizes that corruption is also promoted by the behaviour of multinationals and powerful economic elites seeking undue influence in the policy process.Footnote 120 Having acknowledged this point, he questions the effectiveness of a narrow technocratic approach to anti-corruption, proposing instead a bolder and more encompassing framework for confronting the challenge of ‘state capture’. His key assumption is that addressing governance issues requires an understanding of the ‘complex political dynamics’ in which ‘misgovernance’ takes place. In turn, this implies (1) moving beyond the blanket treatment of ‘politics-as-an-obstacle’ to governance reforms, and (2) abandoning the untenable assumption that the success of a given reform process depends on the ‘willingness’ of particular elites. Overlooking the political constraints shaping economic and institutional change is probably the most serious omission of the good governance approach to development. As Knack states: Footnote 121

The lesson is that we should be more active in studying the ‘possibilities and limits’ of institutional reforms.Footnote 122 The too-often-misinterpreted concept of path dependence may provide a useful starting point. According to North, path dependence simply means ‘constraints on the choice set’.Footnote 123 Politico-economic processes are path-dependent because large-scale changes undermine the interests of groups with high stakes in the status quo, but also because shared mental models limit actors’ choice set by defining the range of ‘perceived legitimate change’.Footnote 124 While scholars are increasingly looking at how distributive conflicts constrain good governance,Footnote 125 the relationship between cognition and institutions does not figure prominently in the good-governance agenda. The challenge is to understand the way in which evolving beliefs constrain the choice-set of the players engaged in the process of institutional change. Comprehending the way key actors perceive the world around them is critical for (1) identifying the sources of the imperfect enforcement of institutional reforms; (2) explaining why particular institutional paths are taken in the context of multiple equilibria; and (3) discovering the scope for institutional innovation.Footnote 126 In North’s words: Footnote 127

We can conclude this section by referring to one traditional normative issue in development policy: the role of institutional reform in the process of economic development. During the 1990s, most experts within international organizations endorsed a bold, comprehensive and orthodox approach to structural reforms. However, Rodrik reviews a number of country narratives and concludes that ‘the onset of economic growth does not require deep and extensive institutional reform’.Footnote 128 This raises questions regarding the scope, sequencing and pace of the reform effort needed to ignite and sustain long-term economic growth. Rodrik argues that the acquisition of good institutions often requires ‘experimentation, willingness to depart from orthodoxy, and attention to local conditions’.Footnote 129 There are signs that the game is moving towards heterodoxy. Arguably, the development community seems to be increasingly sceptical of top-down, universal and rule-based reforms.Footnote 130 In a way, the ‘rethinking of governance’ has been motivated by the following uncomfortable reality: despite the consolidation of the good-governance paradigm, developing countries have made scant progress in controlling corruption, strengthening the rule of law and improving institutional quality.Footnote 131

Institutions and development: lessons from East Asia and Latin America

The contrasting experiences of Latin America and East Asia are one of the highlights of growth econometrics. While seven of the eleven countries labelled by Knack as ‘catch-up countries’ in the period 1960–98 are from East Asia (Singapore, Hong Kong, Japan, Taiwan, Korea, Malaysia and Thailand), not even one is from Latin America. Not surprisingly, many Latin American countries have behaved as ‘fall-back countries’, with Argentina, Uruguay, Venezuela, Nicaragua and Peru showing the most depressing tales of relative economic decline.Footnote 132 The contrast is even starker when income distribution enters into the analysis. Whereas the ‘Asian tigers’ have achieved substantial redistribution in the context of a successful shared-growth strategy, Latin American countries are still at the top of world rankings of inequality.Footnote 133 I will not offer here a complete review of the political economy of Latin America and East Asia’s divergent growth paths.Footnote 134 Instead, the idea is to take up some issues from the comparative experiences of these regions and relate them to our discussion. More specifically, I will analyse whether the experiences of Latin America and East Asia in the post-war period tell us something about the relationship between governance, institutions and growth.

As mentioned above, the ‘institutions versus policies’ issue is at the core of the debate about the causes and implications of the East Asian miracle. On the one hand, some studies seem to suggest that good political institutions and secure property rights are the keys to economic success.Footnote 135 On the other hand, some authors argue that the impressive economic performance of a subset of Asian countries (including China) in the post-war period has been the result of the ‘policy choices’ made by their (mainly authoritarian) governments and not of the ‘institutional constraints’ placed on them.Footnote 136 This unsettled debate is relevant for contemporary Latin America, where in the context of disenchantment with neo-liberal reforms and left-leaning political realignment, the pendulum of the ‘war of ideas’ is swinging in direction of the policy approach.Footnote 137

The comparison between East Asia and Latin America also reveals the problem of reading too much institutionalism into the proposition that political institutions are the main sources of economic development. For example, political stability is one of the key variables explaining both relative economic growth in large-N studies and the contrasting performances of East Asia and Latin America.Footnote 138 But, as suggested before, analytical stretching may call into question the validity of this apparent institutions-to-growth mechanism. Political instability has evidently been one of the expected features of Latin American development. However, this outcome has been more associated with pervasive politico-economic disequilibria than with the features of the formal rules of the political game. Thus, it is unclear what analytical lessons can be drawn from the existence of a link between political stability and growth. This empirical finding should reactivate one of the oldest, and unresolved, debates in constitutional political economy. The challenge is assessing how alternative political systems (for example, forms of government and electoral rules) affect both political order and economic outcomes.Footnote 139

The case of Bolivia, one of the poorest and most unstable countries in Latin America, illustrates the puzzles underlying the relationship between stability and growth. It is clear that misgovernance has been both the source and the consequence of this country’s economic backwardness. However, what should a country like Bolivia learn from the literature on institutions and development? Mantzavinos provides one answer: ‘secure property rights and economic institutions maintaining open markets are the appropriate formal rules for a wealth-creating game’.Footnote 140 However, as North points out, the transplanting of market institutions to the Third World has frequently altered income distribution and generated political instability, leading to consequences that were the reverse of the intended objective.Footnote 141 And, indeed, Bolivia has recently undergone a new perverse cycle of political and economic instability, which was precisely rooted in political contestation against the elite class’s intentions to extend property rights in key areas such as mineral resources or water provision. This is one of the central unresolved tensions in the political economy of growth. There is an established literature showing that secure property rights provide incentives for growth by favouring production over predation.Footnote 142 However, another increasingly important research stream shows that social polarization reduces the security of contract rights and slows growth, and that inequality reduces growth in part through its effect on the security of property rights.Footnote 143

The key political question is how to make wealth-creating economic institutions self-enforcing. Henisz claims that an institutional environment with many ‘veto players’ induces economic growth by creating a credible commitment for non-interference with private investment and property rights.Footnote 144 However, this argument overlooks the historical dynamics of institutions and growth. As Tsebelis stresses, it is unclear whether many veto players will lead to higher or lower growth, ‘because they will lock a country to whatever policies they inherited, and it depends whether such policies induce or inhibit growth’.Footnote 145 In this context, there might be a potential conflict between the checks-and-balances type of accountability implicit in the notion of good governance, on the one hand, and executive effectiveness in implementing pro-growth institutional reforms, on the other.Footnote 146 Weingast argues that a constitution respecting the rule of law and the principles of market-preserving federalism has been the most frequent commitment device used to restrain governments from violating property rights.Footnote 147 The problem is that constitutional provisions as honoured in the United States have rarely been self-enforcing in Latin America, where the resilient political economy of patrimonialism has continued to structure the incentives of the game.

Institutional solutions to credible commitment are themselves subject to problems of credible commitment.Footnote 148 In order to get around this conceptual conundrum, some political economists emphasize modelling the conditions supporting self-enforcing institutions.Footnote 149 The implication is that research on the effects of institutions on growth must include an analysis of the conditions under which productivity-enhancing institutions are likely to emerge and be sustained in equilibrium. As Greif argues, evaluating the impact of particular institutions on economic growth requires ‘examining them in the broader context of the factors influencing the extent to which the political system is self-enforcing’.Footnote 150 The question of what conditions make possible the signalling/enforcement of alternative commitment strategies is key to the political economy of East Asian success. Two sets of supporting conditions are critical here: distributional mechanisms and international factors.

The case of East Asia stresses the interdependence of the processes of accumulation and distribution. One area of active research in political economy focuses on the effects of income inequality on growth via the implied pressure for redistribution.Footnote 151 The new conventional wisdom is that inequality harms growth through both economic and political channels. Recent scholarship has also reported a strong causal effect running from income inequality to bad governance.Footnote 152 In the light of this evidence, East Asia’s shared-growth model seems an attractive template for Latin American countries trapped in a bad equilibrium of ‘misrule’ of law, social polarization and stop-and-go growth. For example, progressive fiscal pacts may contribute to building political legitimacy and social consensus around a wealth-creating game. The puzzle is how to implement a shared-growth strategy that would not be undermined by the macroeconomics of populism; or how to introduce effective wealth-sharing mechanisms against a backdrop of pervasive clientelism. But, most importantly, there is an analytical lesson that should not be overshadowed by the shared-growth narrative. The discovery of virtuous circles between poverty reduction and growth should not lead to the de-politicization of the redistribution debate. Every transition towards a new distributive equilibrium is bound to be contentious and in conflict with the interests and beliefs of key elites.

The experience of East Asia also underscores the critical role of international conditions in supporting both accumulation and redistribution. For example, Campos and Root remark that ‘a geopolitical environment characterized by the high risk of insurrection explains the urgency of finding a cooperative solution to economic growth in East Asia’.Footnote 153 As a case in point, the threat of communist takeover, made credible by a successful peasant-based revolution in China, eased the implementation of wealth-sharing mechanisms (notably land reform).Footnote 154 It is indicative that Ireland, Spain and Greece are also catch-up countries in Knack’s analysis.Footnote 155 This suggests that the process of European integration has also provided a commitment device for countries seeking to reconcile growth and democracy. In the absence of these facilitating geopolitical conditions, the strategy of shared growth in East Asia, and the path of democratization with socioeconomic progress in the European periphery, could have been out of equilibrium.

The ongoing discussion underlines the importance of further integration of international political economy (IPE) into the debate on institutions and growth. Several scholars have stressed the role of external conditions in supporting the rise and sustainability of institutional equilibria.Footnote 156 In addition, an exciting research agenda in IPE explores how international–domestic interactions affect national politics and institutions.Footnote 157 In particular, scholars are looking at the way in which globalization impacts on national democratic governance,Footnote 158 the link between international integration and corruption,Footnote 159 the domestic repercussions of international crises,Footnote 160 and the economic and governance consequences of foreign aid.Footnote 161 All these research streams may offer interesting insights into the mechanisms of governance and development.Footnote 162 A critical question is how the clash between international and national constraints shapes the constellation of national institutions.Footnote 163 Another related question is whether ‘good governance’ institutions promote sustainable growth in the context of globalization. For example, some authors suggest that the relationship between international integration and development is conditional on the quality of domestic institutions.Footnote 164

In order to understand how domestic and international politics interact with institutional development in shaping the growth environment, we must look beyond the conventional property rights approach to institutions. Campos and Root’s analysis of East Asia shows that a variety of institutional schemes supports credible commitments between governments, elites and citizens. For one thing, the co-operation of powerful economic groups was sought by the creation of ‘institutions that facilitate the exchange of information between the public and private sectors’.Footnote 165 For another, East Asian leaders responded to the political challenges of economic growth by designing ‘institutions that spread the benefits of growth-enhancing policies widely’.Footnote 166 Similarly, Rodrik stresses the role of the ‘political institutions of conflict management’ in determining the extent to which domestic conflicts triggered by external shocks are translated into growth-collapses.Footnote 167 Other authors pinpoint the importance of ‘institutional complementarities’Footnote 168 and ‘institutional interdependence’.Footnote 169 In this context, there is a case for unbundling governance and looking at the trade-offs and synergies associated with alternative institutional configurations.

The good-governance approach also underestimates the importance of the political management of economic change. Recent crises in Argentina, Bolivia and Ecuador have eloquently shown that market-oriented reforms often imply a tension between economic efficiency and social cohesion. In the presence of this unpleasant trade-off, we should be cautious in interpreting some empirical findings. For example, some authors have found that economic freedom promotes economic growth.Footnote 170 However, as Karl Polanyi famously asserted: markets constitute a powerful source of socio-political change and produce equally powerful responses as societies attempt to protect themselves against market forces.Footnote 171 Thus, the expected benefits of governance-enhancing reforms are conditional on governments’ capacity to articulate them politically. Institutions may play a role in this process by affecting governments’ ability to accommodate change. Indeed, a key source of long-run economic success might be adaptive efficiency: ‘a set of institutions that readily adapt to the shocks, disturbances, and ubiquitous uncertainty that characterize every society over time’.Footnote 172 Good governance puts emphasis on the credibility/stability of the rules of the game. However, the notion of adaptive efficiency suggests that ‘institutions too have to change in order to promote growth’.Footnote 173 Indeed, institutional stability involves a great deal of ‘institutional adaptation’.Footnote 174 Latin America’s stop-and-go growth may be rooted in governments’ inability to accommodate the institutional stability/institutional adaptation trade-off.

We can conclude this section by emphasizing the value of understanding economic development as a process.Footnote 175 While many societies are still trapped in the deeply rooted equilibria of misgovernance, inequality and economic stagnation, other countries have managed to break with history. The crucial lesson of successful cases is that wealth-creating policies/institutions generally become self-enforcing in the context of specific conditions, including the memory of bloody civil wars and dictatorships, the threat of communist takeovers and processes of regional integration.Footnote 176 By stressing the dynamic features of development, I am not neglecting the policy relevance of theory-informed empirical research. Yet the challenge is striking the right balance between a naïve ‘one-size-fits-all’ approach, on the one hand, and an uninformative ‘everything-depends-on-local-conditions’ one, on the other. The conditions supporting credible commitment and growth in the past might not be replicable. Nevertheless, rigorous ‘contextualized comparisons’ can inform a discussion about the global and domestic governance changes required to ‘approximate’ the conditions facilitating a marriage between growth and social progress. For example, the East Asian experience shows that policy-generated political consensus matters for development, underlining the importance of addressing the ‘legitimacy gaps’ of pro-growth reforms in Latin American countries.

Conclusion: where do we go from here?

The literature on governance, institutions and growth is a progressive research programme seeking to identify the binding political and social constraints on development and to propose ways for overcoming them. The critique I have presented in this article should not lead to the conclusion that the conceptual lenses of new institutionalism/political economy have no leverage for uncovering some of the remaining puzzles of economic growth. In the same way, I do not want to discourage much-needed efforts in the areas of improving governance indicators and modelling the determinants/effects of institutions in cross-country regressions. On the contrary, I highlighted the analytical gaps in the literature with the aim of signalling ways of consolidating this fascinating and open-ended inquiry on the political foundations of development. In this conclusion, I shall outline the main arguments of the article, focusing on spelling out an agenda for further research.

At the analytical level, I claimed in this article that further comparative research on institutions and development should rest upon modelling/measurement strategies that (1) avoid the conflation of policies and institutions; (2) address the problem of conceptual stretching by unbundling the political and the institutional embedded in the notion of governance; (3) consider the role of the state and state formation in economic development; (4) discriminate between the rules of the game, the play of the game, and its outcomes; (5) take into account the difference between the structure, the structure-induced equilibrium and the policy equilibrium; (6) capture the way informal constraints and enforcement mechanisms mediate the relationship between formal institutions and outcomes; (7) recognize the critical importance of beliefs and human learning in the processes of economic and institutional change; (8) move beyond a narrow property rights approach to development; (9) stress the role of special interests and the distributional implications of institutions; (10) attempt to integrate the logics of accumulation/co-operation politics and distribution/contentious politics; (11) look at the sources of credible commitment and self-enforcing institutions; and (12) seek to reconcile theories of institutional effects and theories of endogenous institutions.

The good governance consensus faces a critical policy gap. Despite the growing acceptance that institutions matter for development, there has been little evidence of improvements in the performance of global governance in the last decade. Additionally, comparative studies on the economic effects of institutions rarely provide working principles for ‘institutional policy’. In principle, an institutionalist should demonstrate that the design/modification of the structure of certain institutions would produce better results given an evolving configuration of power/preferences. This task is more demanding from a Northean perspective, as the play of the game depends on an institutional matrix made up of formal and informal constraints, and on the way enforcement/cognitive mechanisms mediate the relationship between incentives and actual choices. In this context, attempting to improve the economic performance of a country by transplanting formal institutions is a recipe for disappointment, not to say disaster.Footnote 177 Clearly, more research on the politics of institutional change is in order. In particular, I believe future agenda should focus on (1) uncovering the domestic and international determinants of good/bad governance; (2) exploring the links between formal and informal constraints; (3) assessing the sources of imperfect enforcement; (4) investigating how special interests and evolving beliefs shape degrees of freedom in institutional design; (5) studying the trade-off between ‘institutional credibility’ and ‘institutional adaptation’; and (6) unbundling governance and analysing the role of ‘institutional complementarities’ and ‘interdependent institutions’.

Demonstrating that growth is a function of good institutions is of little help if this is not followed by an analysis of how such institutions emerge and become self-enforcing. Our knowledge about how rules arise, evolve and change is still insufficient.Footnote 178 As a result, the good-governance approach is handicapped by a limited understanding of how politics and social processes condition the relationship between institutions and outcomes. Our discussion of East Asia and Latin America offers important insights. First, it raises questions about the relative significance of institutional reforms vis-à-vis policy choices in the formulation of growth strategies.Footnote 179 Secondly, it stresses that foreign–domestic interactions seriously constrain institutional evolution and self-enforcing growth. Thirdly, it illustrates that a wide range of institutional mechanisms may support sustained growth by solving co-ordination failures, facilitating public–private co-operation, and articulating distributional struggles. Fourthly, it shows that the solution of multiple credible-commitment problems, regarding both accumulation and distribution, underpins the political conditions for economic growth. Last, but not least, it suggests that the way in which income inequality interacts with institutional development and growth is key to understanding divergent development paths. All these themes are likely to attract academic attention in the near future.

I would like to conclude by pointing out some methodological issues deserving attention. Growth econometrics will have to address the remaining modelling uncertainties. The quality of available measures should also improve. However, it is worth emphasizing that conceptualization must be the first step towards measurement. Taking seriously the issue of the formation of concepts is critically important for enhancing the validity of extant good-governance indicators.Footnote 180 Future research will also have to identify the channels through which institutions affect growth and the ways in which various institutions interact.Footnote 181 However, cross-national studies are not always good for uncovering the mechanisms of development.Footnote 182 This creates a demand for small-N studies looking at the institutional sources of self-enforcing growth.Footnote 183 Indeed, real progress in this field will require a closer dialogue between (1) the game-theoretic literature on institutions, (2) econometric studies on governance and development, (3) context-specific country narratives, and (4) the methodological literature on conceptualization and measurement.Footnote 184 Critically, this dialogue should put more emphasis on integrating economic and political ways of reasoning. After all, the only recipe for progress in solving the puzzles of institutional and economic change is a fruitful conversation between theory and empirics, and between economists and political scientists.