2. Institutionalism: a reluctant defense

Chang's critique of the dominant discourse in institutional economics has a theoretical and an empirical pillar. Theoretically, the argument is that the discourse (a) exclusively focuses on causality from institutions to development, (b) refuses to see the impossibility of the free market, and (c) erroneously claims that the freest market and most secure property rights are best for development. Empirically, (d) support for these claims comes from cross-section studies that are rather shaky.

It is hard to argue with most of these objections. One may question, however, whether they invalidate the institutional discourse. Take the reverse-causality issue. Chang is to some extent correct in arguing that the discussion mostly focuses on the causality from institutions to development.Footnote ¹ However, it is not at all evident that taking reverse causality into account would somehow change or invalidate the insights gained so far from focusing on one side of the relation. A possible defense of the approach by institutionalists such as Daron Acemoglu is that they address one issue at the time, gradually adding complexity to expand our understanding. After showing that at least part of the correlation between institutions and development is attributable to causality from institutions to development (Acemoglu et al., Reference Acemoglu, Johnson and Robinson2001), later works address reverse channels and interaction effects between various institutions (Acemoglu et al., Reference Acemoglu, Johnson, Robinson, Aghion and Durlauf2005; Acemoglu, Reference Acemoglu2010). None of this later work repudiates earlier findings.

Similarly, Chang (Reference Chang2011) rightly argues that economic freedom is impossible to define and not self-evidently good for development. But this critique of the vagueness, subjectivity and ambiguity of a concept like ‘economic freedom’ has little to do with the literature relating much more specific institutions, such as contract enforcement and property rights, to specific economic outcomes. Giving maximal ‘economic freedom’ to business firms is indeed not evidently good for development (ibid.: 7); creating dependable and transparent contract enforcement probably is. A few ideologues apart, it is at this lower level of abstraction that the discussion mainly takes place.

At this level of abstraction, benefits of institutions such as property rights are in the end empirical issues. Chang (ibid., 10–14) identifies important problems with the empirics of institutionalism, targeting its use of cross-section studies lumping all kinds of societies together and of questionable indices to measure institutional quality. Most of this critique is well deserved. But the question remains: were we to add time-series and better measures, do we have reason to believe insights would change? Or would it lead to the same results with more certainty? In the latter case, Chang's argument falls flat. To argue the former, we need to take the critique further, identifying problems in institutionalist reasoning that cause structural biases in its conclusions.

3. Institutionalism: methodological critique

I argue that such structural biases are present in the dominant discourse and that they arise from the fact that institutional economics is essentially a comparative economics (Djankov et al., Reference Djankov, Glaeser, La Porta, Lopez-de-Silanes and Shleifer2003). Institutionalism attempts to fit the diversity in the way economies function into a single model, which allows us to compare societies with varying degrees of economic success with each other. This methodology creates pitfalls and limitations.

First, it results in a focus on differences of degree, neglecting differences of kind. Meaningful comparisons require factors held in common by all cases; otherwise we end up comparing apples with oranges. Chang's (2011: 8) criticism of the property rights discourse for failing to allow other forms than, private, state, and open access illustrates this. All forms of property rights are forced onto a few dimensions; aspects of property rights regimes that cannot be fitted are ignored.

Second, a strict comparative focus results in a neglect of potential interactions. Differences in economic performance between societies are attributed to the differential effects of their respective institutions. But what if performance and institutions of one society are linked to the performance and institutions of another? What if the effect of certain institutions is dependent upon the position of a society in the world? The classic study by Alice Amsden (Reference Amsden1989), for example, suggests that late-comer development requires different institutions than early development. Also, the fact that emerging economies nowadays seem to benefit from so-called market-friendly institutions is probably related to the fact that today's rich countries are market economies. Would market-friendly institutions still be good for development if the Soviet Union had won the Cold War and communist countries dominated the world?

What an open eye for interactions teaches us is that the relationships we observe in institutional studies are not necessarily universal, but depend on the context in which they occur. This brings us to the third and most important pitfall of institutionalism: a neglect of the context-laden character of the perspective from which we conduct comparisons. As institutionalists, we recognize diversity in the laws and mechanisms by which societies are governed. As economists, we seek to develop universally applicable theories which cover all possible institutional settings. As a result of this basic tension, institutional economists have the tendency to adopt a partial perspective and assume it to be universally relevant (Beugelsdijk and Maseland, Reference Beugelsdijk and Maseland2011). The civil versus common-law opposition (LaPorta et al., Reference LaPorta, Lopez-de-Silanes and Shleifer2008), for example, is taken from a Western context and adopted as the lens through which to interpret all societies. Had we started out from an Asian context, we would have probably come up with different dimensions and results.

Other aspects of the institutional-comparative framework are similarly vulnerable to partiality. For example, researchers often use nation-states as units, not recognizing that these are relatively recent phenomena. The ‘historical experiment’ literature, all the craze in institutional economics nowadays, illustrates this. It focuses on finding factors that exhibit variation across countries centuries ago to relate them to (institutional) divergence between countries today (Acemoglu et al., Reference Acemoglu, Johnson and Robinson2001; Easterly and Levine, Reference Easterly and Levine2003; LaPorta et al., Reference LaPorta, Lopez-de-Silanes and Shleifer2008; Haber and Menaldo, Reference Haber and Menaldo2010). But that logic only works as long as one assumes a fixed one-to-one relationship between territory and society, something that is more nationalistic myth than historical fact.

All these issues – a focus on differences of degree, a neglect of interactions, and the tendency to apply institutionally specific perspectives – work together to create a discourse in which currently successful societies are often held as universal and objective ‘best practices’. Others are represented only from the perspective of their dissimilarity to these winners – their otherness being put forward as explanation for their lesser performance. One does not have to be a cynic to see in the many institutional measures developed over the years measures of Anglo-Saxonness.

There is a pressing need to be more aware of the partiality and limitations of the current institutional discourse. Economists need to come to terms with the fact that the kind of universal frameworks and single-peak thinking that we have grown used to in economics simply do not fit the fundamental diversity inherent in institutionalism. Unfortunately, the level of reflection still present in the work of the frontrunners in the debate (North, Reference North2005; Rodrik, Reference Rodrik2006) is lacking in that of many of their followers. The result is an often naive equation of developed-country institutions with good institutions.