2. Institutions: why they matter

Consider first points (i) and (iv) above. It is true that the concept of institutions is a source of confusion and has not been used consistently in the literature. As Glaeser et al. (Reference Glaeser, La Porta, Lopez de Silanes and Shleifer2004) have pointed out, the standard measures of economic institutions used in the early empirical literature refer to policy outcomes, let alone the rules of the game mentioned by McCloskey. These outcomes reflect a combination of factors that are hard to sort out, such as the quality of the public administration, underlying cultural attitudes, and detailed policy choices. The same critique applies to measures of political institutions that often combine rules of the political game and political outcomes.

McCloskey is also right to point out that the existing empirical literature on political institutions has not been able to precisely point out how changes in political institutions lead to subsequent economic improvements of the observed order of magnitude. A very large literature has attempted to quantify the economic effects of democratic transitions. These effects are surprisingly small. Many studies have argued that democracy is not helpful for growth. According to Persson and Tabellini (Reference Persson and Tabellini2006), the long run effect of becoming a democracy is to raise GDP per capita by about 12% in the long run. Exploiting better data and more sophisticated econometric techniques, Acemoglu et al. (Reference Acemoglu, Naidu, Restrepo and Robinson2014) raise the estimate to 20% in 30 years. True, a democracy is also more involved in redistribution, which can be a source of distortions. Moreover, a single indicator of democracy does not adequately capture what may be the key differences in political institutions. But certainly these estimates point to the fact that being a democracy does not explain why some countries are rich and others are poor.

Nevertheless, research on institutions and comparative development is not as simplistic as McCloskey portrays it. The argument in this literature is not that better rules of the game improve equilibrium allocations and get rid of Harberger's triangles. The argument is that political freedoms, the rule of law, checks and balances against government abuse, allow the economy to exploit dynamic gains from enterprise, investment and innovation. More open and inclusive political institutions reduce opposition to innovation and facilitate an intellectual climate of openness to innovation, critical thinking, and the emergence of a market for ideas. As suggested by Joel Mokyr (Reference Mokyr2012), this is just what happened before the industrial revolution. Of course liberal ideas in turn influence the functioning of existing political institutions, so causation is certainly not running in only one direction. But innovation does not emerge in a political vacuum, and can easily be stifled by oppressive and authoritarian regimes, even in modern times. Bad political institutions can block innovation and they have often done so in the past, as emphasized by Acemoglu and Robinson (Reference Acemoglu and Robinson2012).

Existing research on institutions and economic development is also exploring how the organization of the state – its ability to establish order and deter violence, to enforce contracts, to provide public goods – interacts with economic development (see for instance Besley and Persson, Reference Besley and Persson2011). This literature acknowledges that economic development and state formation go hand in hand. It asks what brings about specific changes in state organizations, and how these in turn affect economic development. Although it is still too early to assess how lasting the contribution of this line of research will be, it certainly goes beyond the simplistic view that economic development is just a matter of ‘adding institutions and stirring’.

3. Ideas: which ones matter and where do they come from?

Deirdre McCloskey emphasizes that critical turning points in economic and political development are often associated with the emergence of important new ideas. This emphasis is certainly right, and it is often missing from the economic determinism of some recent contributions, and in particular from Acemoglu and Robinson (Reference Acemoglu and Robinson2012). Yet we need to go beyond the claim that ideas matter, and identify which ideas are so important, why, and how they emerge.

In this paper and in other scholarly work, McCloskey points to the emergence of the bourgeois ethic as a key turning point. The consolidation and diffusion of the goals and values of the urban middle classes, it is argued, can explain both the great divergence between Europe and the rest of the world that started in the XVIIIth century, as well as the more recent rapid convergence of China and India.

This may well be, and space constraints prevent me from adequately discussing this conjecture. But any explanation of economic or political change that rests on the importance of specific ideas faces two difficulties. First, why emphasize one particular set of ideas rather than others? Specifically, why is this particular cultural innovation so important, as opposed to others that a priori seem equally if not more relevant? There are two other cultural traits, in particular, that a priori seem equally or more central in the process of economic development, and that have also attracted the attention of economists.

The first one concerns attitudes towards innovation. Rapid economic growth is often correlated in time and space with creativity in many other dimensions, scientific and artistic. This is true of Italy in the XVIIth century, of Vienna in the late XIXth century, of the US today. There are many possible explanations of this phenomenon, including reverse causation (economic growth attracts individual talents or generates a demand for amenities). One such explanation, however, is that economic growth, scientific innovation and artistic creativity all reflect an underlying culture of openness to innovation and change (any innovation, not just an economically valuable innovation). Benabou et al. (Reference Benabou, Ticchi and Vindigni2014), for instance, document a strong and robust negative correlation between religiosity and measures of scientific innovation across US states. In subsequent work, exploiting the individual opinions collected in the World Value Surveys, Benabou et al. (Reference Benabou, Ticchi and Vindigni2015) show that individual religiosity is also strongly negatively correlated with cultural indicators of openness to innovation (e.g. attitudes towards science, new versus old ideas, risk propensity and so on). Similar findings have been provided by others with different methods and data sets. How do we know that it is the diffusion of bourgeois ethics (and the associated appreciation of innovation with commercial value), rather than openness to innovation in general, that is responsible for rapid economic development?

A second cultural trait that has attracted much attention by economists working on comparative economic development is the distinction between generalized versus limited morality. Namely between norms of good behavior that are deemed to apply towards everybody versus a narrow group of relatives or friends (see for instance Platteau, Reference Platteau2000). The diffusion of generalized morality facilitates social interactions in many ways, both in economic and political domains, with obvious implications for economic development. In my own work, I have attempted to exploit this distinction to explain the functioning of administrative and political institutions within Italy, as well as differences in economic development across countries or regions (see Tabellini, Reference Tabellini2008a). In ongoing work with Avner Greif, we seek to explain the different evolution of social organizations in China and Europe on the basis of this distinction: China's culture of limited morality led to the emergence of the clan as a key organization, whereas the diffusion of the corporation in Europe was facilitated by the underlying principles of generalized morality imbedded in the Christian religion (Greif and Tabellini, Reference Greif and Tabellini2015).

A second obvious difficulty with an ideas-based explanation of economic and political development is that ideas are endogenous. Where do these ideas come from, and why do they spread so rapidly in some places or moments in time and not others? Why do bourgeois values spring up in India in the early 1990s and not sooner or later? If we cannot adequately answer these questions, our explanation is incomplete.

This is not to detract from the importance of the middle class bourgeois values in the process of economic development. Perhaps they are equally or more important than other cultural traits. But any claim to dominance of one particular set of historical ideas or values rests on fragile foundations, given the measurement difficulties, and given that ideas are certainly co-determined with the same economic and institutional processes that they seek to explain.

4. The method of economics

In the final part of the paper, McCloskey argues that the method of economics prevents us from correctly analyzing institutions and many other crucially important issues in the social sciences. The reason is that the almost exclusive focus on individual optimization implies that economic theories are forced to abstract from notions of ethical valence and other key concepts, such as obligations, rights, power relations between individuals.

I disagree with these statements. Economics relies on methodological individualism (cf. Popper, Reference Popper1945 and Hodgson, Reference Hodgson2007 for a more extensive and critical discussion). This in turn is built on two principles. First, we ought to explain social phenomena starting from the decisions of individuals. That is, to explain group behavior, or aggregate phenomena, we need to explain the behavior of the individuals composing the aggregate, and the interactions between them. Second, individual behavior is explained in terms of what Popper called the situational logic. That is, we reconstruct the decision making situation in which the individual finds himself/herself, and we apply the methodological postulate that he/she behaves appropriately to that situation. Individual behavior is explained by explaining the situation (constraints, beliefs, preferences, incentives). Maximization is often used as a simplifying assumption, but of course the postulate of ‘appropriateness to the situation’ can be formulated in ways other than simple maximization, as witnessed by the rapidly growing literature in behavioral economics.Footnote ²

The simplified reconstruction of the decision making situation, however, does not force us to abstract from the concepts that McCloskey rightly considers so important. Although this is often left implicit in many economic studies, individual beliefs can also refer to the intentions of others, and such intentions have implications for how I feel and for what I do. The literature on psychological games deals precisely with these issues (see Battigalli and Dufwemberg, Reference Battigalli and Dufwenberg2009). Similarly, the notion of preferences in the situational logic can be extended to include values and concepts such as fairness, obligations, rights. Several papers for instance stress the importance of reference points as drivers of individual behavior (e.g. Hart and Moore, Reference Hart and Moore2008). Such values and preferences, moreover, need not be taken as exogenous. On the contrary, we can apply standard tools of economics to study how parents transmit values to their children, or influence their beliefs (as in Benabou and Tirole, Reference Benabou and Tirole2006; Bisin and Verdier, Reference Bisin and Verdier2001; Tabellini, Reference Tabellini2008b). Perhaps this literature is not as large as it should be, and its results are still preliminary or unsatisfactory. But there is no methodological constraint that prevents economics from addressing these important problems.

Although there is much to learn from other disciplines, including the humanities, methodological individualism has proved to be extremely successful and there is no reason to abandon it. Of course the world is complex, and our abstractions can only capture a small part of reality. Deirdre McCloskey is right to point out that economists should pay much more attention to ideas, values and culture, and how these interact with the functioning of institutions. But to explain is also to simplify, and the simplifications used by economists are often more profound and revealing compared to those implicit in other methods of analysis.