2. Dominant discourse on institutions and economic development

With this background, let me turn to Professor Ha-Joon Chang's paper (Chang, Reference Chang2011). As in his many other papers, he brings widespread knowledge and prodigious energy to challenge accepted interpretations. However, as I read this and his other work, I often find myself rising to the defense of mainstream economics against what I consider to be ill-considered or unfair attacks. Let us consider some of the points made in this article.

The paper first argues that the dominant discourse on institutions and economic development suffers from three theoretical problems: (a) its neglect of the causality running from development to institutions; (b) its inability to see the impossibility of a free market; and (c) its belief that the freest market and the strongest possible protection of private property rights are best for economic development. I do not find myself persuaded by any of these claims.

(a) The literature is full of observations on the causality running from development to institutions; there may have been more emphasis on the causal effect of institutions on development, but that is understandable given the intellectual background, as described above. The idea that the growth of commerce and the political power of the business class reinforced each other is a prominent theme in North (1981). The effect of the size of the market on the evolution of institutions favorable to commerce is described throughout the book by Nathan Rosenberg and L. E. Birdzell (Reference Rosenberg and Birdzell1986). If there has been less progress in explaining the emergence of development-promoting institutions than in making the case that institutions matter, that is because it is a much tougher nut to crack.

(b) I agree that there is no objective definition of the ‘free market’; but so what? Of course, the assessment of the roles of the market, state, and other institutions in the economy depends on value judgments. Who denies that? I find the concept of the free market to be very useful, even if its definition is fuzzy around the edges, but in using the term I do not imply that the free market is the most efficient arrangement. Nor does mainstream or even neoclassical economics.Footnote ¹

(c) The dominant discourse in institutional economics certainly does not claim that maximal protection of private property rights is best for development, nor does it typically see the relationship between private property owners and the state as an antagonistic one (Rodrik, Reference Rodrik and Rodrik2007). It is true that a series of articles by Rafael La Porta et al. (for a nice review, see Nunn, Reference Nunn2009) has made strong claims for the superiority of the common law legal system, but these claims have not gone unchallenged. Chang is on weak ground when he takes provocative new claims by leading researchers to be the ‘dominant discourse’.

3. Institutions

Next, the paper argues that ‘the supposed evidence showing the superiority of “liberalized” institutions relies too much on cross-section econometric studies, which suffer from defective concepts, flawed measurements and heterogeneous samples’ (Chang, Reference Chang2011: 1). Granted, it is difficult to find empirical counterparts to theoretical concepts, in institutional analysis as in other areas. Despite the weaknesses of the measures, I do not think there is much doubt about the claim that institutions differ enormously across countries and that institutions in a broad sense matter greatly for development, as the thought experiment indicates. Moreover, the literature is moving forward on the challenging task of explaining how institutions come about and how they change. The amount and types of data have increased substantially since the early 1990s. The early studies made use of the judgments (typically made by ‘knowledgeable experts’) by for-profit credit-rating agencies that sold their ratings to companies doing or contemplating business in emerging economies. Since then, the World Bank, the Gallup poll, the World Values Survey and other organizations have conducted surveys of business people, government officials, and ordinary citizens aimed at illuminating institutional differences among countries and over time. Detailed political databases have been compiled, and analysis is being conducted among groups of similar countries and among regions at the sub-national level, where some of the ambiguities in the variables are less severe (see Menard and Shirley, Reference Menard and Shirley2008). The ‘observer bias’ becomes less severe, as early measures of institutions can be related to later measures of performance. At the same time as measures of institutions have been accumulating and improving, the economics profession has paid much more attention to econometric attempts to determine causality, through instrumental variables, event studies, regression discontinuity designs, and differences-in-differences techniques.Footnote ²

4. Climate-culture and global standard institutions

The final section of the paper says that the dominant discourse takes one of two positions: (a) the ‘climate-culture’ school believes that institutional arrangements are inherited from history and are virtually impossible to change, and (b) the ‘global standard institutions’ school believes that institutions can easily be changed if politicians display political will.

(a) It is surely unwarranted to say that a researcher who identifies a cultural or climatic determinant of institutions through historical investigation is claiming that history is destiny. Stanley Engerman and Kenneth Sokoloff, and Daron Acemoglu and his co-authors have pointed out reasons why institutions, both functional ones and dysfunctional ones, tend to persist, but they certainly do not deny that institutions have changed and are changeable through the actions of political leaders. A good discussion of the persistence and changeability of institutions is in the textbook by David Weil (2009), a mainstream economist who is not particularly of the institutionalist persuasion. If the points that Chang is making are being explained to undergraduates in mainstream textbooks, where is the dominant discourse?

(b) I am quite sympathetic to the points that Chang makes about the extreme voluntarism of some of the commentary on institutions and economic development. In my view the criticism of extreme voluntarism is not to be directed against the academic literature, which is generally highly cognizant of the forces causing institutions to persist, but against the pronouncements of international organizations, the media, business groups, and activists promoting one or another cause. These actors have incentives to simplify their messages. Moreover, I think there is a natural tendency, stemming from intellectual laziness, to which many of us are prone, to identify a single, key obstacle to development. The differences in institutions between rich and poor countries are stark, and it is easy to persuade oneself that ‘if only’ this particular pattern of behavior could change from the way it is in India to the way it is in the USA, then all would be well. But the message of institutional thinking is that the complex of behavior, expectations, beliefs, and values that support an institutional equilibrium have to evolve through changes in these variables at the individual level. You have to start from where you are. Evolutionary thinking does not preclude rapid evolutionary change,Footnote ³ but changes in behavior need to make sense at the individual level. Moreover, while institutions matter, they are not the only determinant of growth and development, and economic development generated by other causes can often bring desirable kinds of institutional change. On this important message I agree with Chang.

5. Concluding remarks

Economics has changed enormously since Chang burst onto the scene in the early 1990s. His critiques of mainstream economics are far more applicable to the discipline at that time than they are now. Whatever the shortcomings of the current literature on institutions and development, progress is not being held back by conceptual blinders imposed by economic orthodoxy. If you read through the Handbook of New Institutional Economics (Menard and Shirley, Reference Menard and Shirley2008), you find consideration of all types of causal channels, from policies to institutions to values and world views and back again. Game theory and behavioral economics have enriched the menu of models acceptable in the professional journals. More important, empirical standards have greatly advanced. Behavioral models are being confronted with computerized games and even with in-the-field experiments, especially in poor countries where recruiting experimental subjects is cheaper (Duflo et al., Reference Duflo, Glennester, Kremer, Paul Schultz and Strauss2006; Banerjee and Duflo, Reference Banerjee and Duflo2008). The recent literature does not merely undermine traditional assumptions;Footnote ⁴ it offers parsimonious interpretations of behavior that help us to understand the economy and the polity better than before. To be sure, progress has consisted and will continue to consist in the accumulation of insights about particular situations; society is too complex and cultures and institutions are too diverse to expect simple new conclusions.

The fact that intellectual progress is occurring does not imply that there will be breakthroughs in the practical means by which outsiders can help poor countries to develop. Institutional deficiencies are a major barrier to progress, yet institutions are not easily reformed, especially through outside intervention. I imagine that in many cases outsiders can help most, not by a direct assault on the most important institutional obstacles, but by working around them.