Every day in the opinion columns of newspapers and blogs, editorial writers construct enemies of straw. They do this by simplifying and exaggerating opposing points of view, presenting them as senseless extremes with which most reasonable people would disagree, and which they can easily rout with their superior reasoning. By not naming or quoting their opponents, these columnists avoid the tiresome task of marshalling evidence against real-life adversaries whose subtle or complex arguments may be hard to defeat and who may demand a chance to defend themselves. Should scholars do the same?

Ha-Joon Chang's article (Chang, Reference Chang2011) is a case in point. It aims to ‘advance our understanding of institutional economics by critically examining the currently dominant discourse on institutions and economic development’ (ibid.: 1). This is a worthy goal, but the ‘currently dominant discourse’ is a straw man.

Chang's ‘dominant discourse’ is extremist: it ‘argues that institutions that maximize market freedom and most strongly protect private property rights are the best for economic development.’ (ibid.: 3, emphasis added) Chang says these institutions are ‘often called the Global Standard Institutions’ or GSIs (ibid.: 2). (By whom I wonder? The term is new to me.) Chang tells us that the GSIs are ‘typically found in Anglo-American countries’ and while there is no agreed definition, he says ‘the most frequently mentioned’ are a common law legal system, private ownership, a developed stock market, financial regulation that encourages prudence and stability, corporate governance based on shareholding to ensure that corporations are run for their owners, a flexible labor market, and a political system that restricts arbitrary action by political rulers and bureaucrats (ibid.). GSIs are ‘forced’ on developing countries by WTO (World Trade Organization) and NAFTA (North American Free Trade Agreement) rules, and pressure from World Bank and IMF (International Monetary Fund) loan conditions (ibid.). Chang criticizes the unnamed proponents of the GSIs for their support of unrestricted business freedom, their failure to recognize the need for third-party enforcement, their assumption that the relationship between the state and private property owners is always antagonistic, etc. He equates them with those benighted souls who once supported child labor and opposed patents.

Chang's ‘dominant discourse’ is simplistic: it assumes that the relationship between institutions and growth is linear and the same across societies (ibid.: 9). One group within the ‘dominant discourse’ assumes that institutions can be easily changed, ‘remolding non-GSIs into GSIs, or importing GSIs into countries with missing institutions’ (ibid.: 15). Another group is fatalistic, assuming that institutional change is nearly impossible. More on this later.

Seldom does Chang clearly identify the people responsible for or engaged in this discourse and nowhere does he quote from their offending works. Chang does say that the underlying academic research for the ‘dominant discourse’ was supplied by World Bank researchers and academic economists. Chang impugns the motives of these academic economists, saying they are responding not just to real-world circumstances, but also to academic fashion and ‘the high publishability of a relatively new research topic’ (ibid.: 3). (Strange, I always thought it was hard to publish institutional analysis in our neoclassically dominated journals. . . but I digress.) Who are these economists? Chang never specifies, although in another context he cites North (Reference North2005), La Porta et al. (Reference La Porta, Lopez-de-Silanes and Shleifer2008), and Acemoglu et al. (Reference Acemoglu, Johnson and Robinson2001, Reference Acemoglu, Johnson, Robinson, Aghion and Durlauf2005).

Why do I think the ‘dominant discourse’ is made of straw? Let us consider a specific person's point of view, as Chang has not. If there is one scholar whose work dominates the argument that institutions are an important determinant of development it is Douglass North. But North's views about institutions are largely at odds with Chang's ‘dominant discourse’. North argues that institutions are highly specific to time and to individual societies; that transplanting institutions from rich to poorer countries is unlikely to succeed and can result in violent civil strife; that institutional change is not easy and is usually gradual and incremental (see, e.g., North, Reference North1990, Reference North2005).

North does argue for stronger property rights and competitive markets but not in the simplistic and mindless way of Chang's ‘dominant discourse’. North states:

Chang (Reference Chang2011) critiques the ‘rational-choice’ framework of the GSI discourse, but one of the defining characteristics of new institutional economics is its rejection of neoclassical rational choice in favor of a more nuanced view. As North argues, the rationality assumption ‘may be true in evaluating opportunity costs at the supermarket, but it is wildly incorrect when it comes to making more complicated choices in a world of incomplete information and of subjective models used to interpret that incomplete information’ (North, Reference North2005: 64).

By lumping nameless scholars together as part of the ‘dominant discourse’ Chang ignores major disagreements among those studying institutions. Many of Chang's GSIs are controversial. For example, his list of the GSIs begins with a common law legal system, but the argument that a common law origin is always superior to civil law origins is highly disputed by many new institutional economists (and many legal scholars). Most of the essays on legal institutions in the Handbook of New Institutional Economics (Menard & Shirley, Reference Menard and Shirley2005) take issue with the legal origins argument. Chang says the GSIs are being forced on developing countries. Does this means that civil law countries are being required to adopt common law by aid conditions or trade rules? It seems patently ludicrous to make common law a condition of trade agreements or foreign assistance, since you cannot adopt customary law if you do not have the customs. It would strengthen Chang's argument if he were to cite some specific rules or projects that require countries to adopt common law.

I mentioned that Chang divides the ‘dominant discourse’ into two groups: the extreme voluntarists, who think institutions can easily be changed, and the extreme fatalists, who think institutional change is ‘nigh-impossible’. Chang does not name the voluntarists but he does name some fatalists: Stanley Engerman and Kenneth Sokoloff, and Daron Acemoglu, Simon Johnson and James Robinson (Chang, Reference Chang2011: 16–17). According to Chang, these fatalists believe that ‘institutions are determined by immutable things such as climate and culture, so they cannot be changed, except through some epoch-making external shocks, like colonization’ (ibid.: 16). Let us consider whether Engerman and Sokoloff fit the box that Chang has made for them; I do not have space to do the same for Acemoglu, Johnson and Robinson. Engerman and Sokoloff argue that factor endowments favoring plantation agriculture, extensive native labor and rich mineral resources contributed to differences in initial inequality in the Americas that ‘contributed to systematic differences in the way institutions evolved’ (Engerman and Sokoloff, Reference Engerman and Sokoloff2002: 45). Institutions change in Engerman and Sokoloff, but they change in ways that cause the initial extreme inequality to persist over time. In support of their argument they provide statistics on persistent differences in land ownership, suffrage, literacy rates, etc. I do not entirely agree with this argument, but I think it should be critiqued on its merits. Instead, Chang simplifies their argument beyond recognition and then equates it with the argument of the ‘many people’ who believe that Confucianism is pro-development, Islam is anti-development, and France is dirigiste, arguments that do not figure in Engerman and Sokoloff.

Surely I should be a poster child for Chang's ‘dominant discourse’, one of his ‘mainstream’ institutionalists. I am a former World Bank researcher, a card-carrying new institutionalist, a co-founder and past president of the International Society for New Institutional Economics, and a co-editor of the Handbook for New Institutional Economics (Menard and Shirley, Reference Menard and Shirley2005). Yet it may surprise the reader to learn that I share a number of Chang's concerns and discomforts. I too have been critical of aid agencies such as the World Bank for telling developing countries what institutions they need and for expecting that long-lived institutional frameworks can be changed by outsiders through foreign aid (Shirley, Reference Shirley2008). I too have criticized the broad abstract indices used to measure institutions, and I share Chang's concern about their use in cross-country growth regressions (ibid.). I agree with Chang (and North) that the same institutions may be good for one country or at one time, but not for another country or another time, and that institutional change is very country specific and unlikely to follow Western best practice (ibid.). If I do not fit into Chang's ‘dominant discourse’ – and I do not – then who does? Could it be that the ‘dominant discourse’ is a null set? I am not arguing that I or most new institutionalists agree with much of what Chang has written; far from it. Rather I am arguing that Chang has posited such simplistic and extremist views in the ‘dominant discourse’ that they bear little resemblance to what most new institutional economists espouse.

Chang is not alone in using the straw man approach to scholarly criticism. Other scholars, some quite prominent, regularly do the same. I have probably used it myself. And why not? It is much easier than debating real-life opponents who may use subtle reasoning, marshal strong empirical evidence and reply harshly to criticism. Well, the reason why not is because we are scholars, not opinion-editorial (op-ed) writers, and that requires us to battle in the field of ideas against real, fire-breathing dragons, not straw ones.

So what should be the standards for scholarly criticism? I do not pretend to answer that question, but I can suggest three ways to make a start. First, we should not critique opposing ideas by making them sound idiotic. Nor should we argue against our opponents’ ideas by equating them with discredited viewpoints that they do not espouse, such as support for child labor. We should honestly represent what others believe, rather than build an exaggerated, oversimplified caricature. Second, we should name those whose ideas we dispute, cite their works and quote their offending passages. We should refrain from using the ‘passive evasive’ (‘it is argued’) or the imprecise subject (‘many people argue’). Third, we should not lump together the ideas of scholars who disagree with one another and call that a paradigm or a ‘dominant discourse’.

An article that fits these standards well is Ronald Coase's ‘The Lighthouse in Economics’ (Reference Coase1974). Coase begins with quotes from Paul Samuelson and other economists about the lighthouse as a purely public good, provides contrary evidence of private lighthouses, and concludes by stating what he believes Samuelson has argued and why he disagrees. You do not have to agree with Coase's conclusions to see why this format is superior to the straw man approach. In ‘The Lighthouse in Economics’ the reader can judge for herself whether Coase has fairly represented Samuelson's views and presented sufficient contrary evidence to make a persuasive counterargument. Emulating Coase is not easy, but well worth the effort. We would all benefit if straw men were banished from scholarly discourse.