(a) Naive advocacy of liberalization reforms

It may be true that some important multilateral donor organizations are sometimes overly zealous in advocating some Washington consensus-type reforms and overly powerful in pushing their adoption by developing countries as a result of their financial power and their large research staffs. But, there are other large and powerful international organizations that are often critical of such reforms. Moreover, more careful investigations of the writings of any single such agency is likely to reveal a more nuanced picture. For example, while their annual reports often cover a single theme on which they may make several general policy proposals, the same reports frequently contain special ‘boxes’ citing implementation problems encountered or alternative reforms introduced in certain circumstances. In any case, beyond the confines of international agencies but still within I&D, there is no dearth of alternative institutional reform proposals.

(b) Overly simplistic rationale for such reforms

Contrary to assertion (b), I&D analysts are quick to admit the complexities of the issues and the hazards of generalization. For example, in the case of intellectual property rights it is widely admitted that the usefulness of such rights will depend heavily on considerations such as the level of development, the ability of its citizens and firms to undertake research and development, the sector composition of output, and the legal system's ability to enforce such rights (Allred and Park, Reference Allred and Park2007). Even in the case of land rights, I&D contributors have long pointed out complexities, such as the need to distinguish between (1) common property rights and open access (recognizing that for some activities and some environments common property rights may be more suitable than individual private property rights; Ostrom, Reference Ostrom1990; Nugent and Sanchez, Reference Nugent and Sanchez1993), and (2) the many different dimensions of land rights, e.g., rights to use, to sell, to rent, to inherit and bequeath and their scope of application (e.g., to land itself, to the air space above ground or the mineral rights below it; de Janvry et al., Reference De Janvry, Gordillo, Platteau and Sadoulet2001). Political constraints on standard property rights reforms are also widely recognized (e.g., when a tribal chief opposes private property rights because this would imply the loss of his power to zone or allocate, which in some cases can be justified on efficiency grounds). Moreover, I&D analysts have identified important instances, such as when these rights are highly concentrated in the hands of a few large politically powerful landowners, where private land ownership can be disadvantageous, leading to the undersupply of important public goods (Galor et al., Reference Galor, Moav and Vollrath2009).

(c) Excessive reliance on cross-section evidence

While much of the early evidence used to support one side or another of debates on property rights and liberalizing reforms was based on international cross-section evidence, this was the natural result of the only recent development of comprehensive institutional indicators based on contemporary surveys. Yet, with time, increasingly panel datasets have been constructed allowing panel data techniques to be employed, thereby mitigating endogeneity and omitted variable biases and allowing for tests of structural breaks and sample heterogeneity.

Even more importantly, I&D studies have been making extensive use of micro-level data and quasi-experimental methods that help relax many of the simplifying assumptions that Chang criticized. For example, the revolutionary change in China's property rights in land from collective ownership to quasi-individual ownership known as the Household Responsibility System was the result of its experimental use in one particular district of one province in 1978 which, when successful, led to demands for its use all over the country (Lin and Nugent, Reference Lin, Nugent, Schultz and Strauss1995). Other examples arise from the use of detailed surveys of individual farmers to identify their perceptions of the strength and scope of the various different land rights. Indeed, Tim Besley (Reference Besley1995) constructed quantitative measures of the security of these perceived rights for Ghana and showed them (1) to vary across villages, land type, even fields in a particular village and (2) to have different effects on farm investments depending on many factors. Erica Field (Reference Field2007) used a somewhat similar approach to show how varying perceptions of the security of land rights in Peruvian cities affected the extra-household labor supply in a somewhat more experimental setting. This kind of work has become quite standard and allows difficult selection issues to be treated.

While much of this work takes advantage of modern statistical and econometric techniques, contrary to Chang, appropriate case studies are much in use in I&D analysis. See, e.g., Platteau and Seki (Reference Platteau, Seki, Aoki and Hayami2001), who explain why fishermen in Japan's Toyoma Bay choose to fish cooperatively with full sharing of all costs and revenues in an otherwise rather capitalistic society, Nugent and Robinson (Reference Nugent and Robinson2010) who compare two countries in which coffee is produced almost entirely on large plantations with two others where coffee production is produced on smallholder farms and trace the consequences of these differences for institutional change, human capital and economic growth trajectories. For additional case studies applying a wide variety of institutional approaches to a wide variety of development issues, see Nabli and Nugent (Reference Nabli and Nugent1989).

(d) Ignorance and naiveté concerning the ability to change institutions

While early I&D studies focused on the effects of institutional changes, usually taking the latter as given, this was quite reasonable in view of the recognized costs of changing institutions. Only after determining which institutional changes matter and exert positive effects on development does it pay to investigate in detail the determinants of such changes. Nevertheless, especially for those institutional changes with demonstrated positive effects, very considerable efforts have been given to identifying and testing for the relative importance of a wide variety of determinants of these institutional/policy changes. Some of the factors affecting the likelihood of such institutional changes include pressures from donor agencies, large neighbors or trade partners, complementary institutions (which make for path dependence in institutions), and economic or political crises (Campos et al., Reference Campos, Hsiao and Nugent2010). To explain reform efforts and to distinguish sustainable from unsustainable reforms, economists, sociologists and political scientists have made innovative use of such diverse methods as collective action theory, applied game theoretic approaches to coalition formation, and club theory. Virtually all these approaches recognize that institutional change is far from costless, and sometimes feasible only in rather special circumstances.

(e) Failure to examine the determinants of institutional change and reverse causality

Naturally, once I&D studies began to examine systematic differences in the effects of institutions or institutional changes from one level of development (or even from one village to another) as in (b) above, and the determinants of institutions as in (d) above, it was easy to see that the causality could in principle be two-way or even one-way from development to institutions. In the case of property rights the importance of reverse causation has long been recognized in that it would only pay to develop property rights in a certain kind of asset when that asset had become sufficiently scarce and potentially valuable as to outweigh the costs of establishing a system of property rights and legal systems capable of enforcing such rights, conditions that are more likely to be fulfilled in higher-income countries.