Final Reflections

Lawrence Harrison doesn’t have a hypothesis about culture and economic development. He has a hypothesis about the interaction between culture and institutions. Though he is not clear about this, this is what his argument implies, and this is the only thing that could explain why Chinese people have a good culture, yet China is extremely poor, and why the supposedly beneficial Chilean culture only became apparent in terms of economic performance after 1985.

This might be right, though I doubt it. How could one know if it were right? Not by quoting various authorities. The way we try to discover if a hypothesis is right in social science is to test it. To do so we’d need to be much more specific about how to conceptualize and measure culture. In my first reply I tried to raise this issue but the proponents of the cultural hypotheses seem to be remarkably uninterested in measurement and testing. They seem content with vague concepts, and untested arguments. I think this is why Greg Clark maintains that there has been so little progress on this topic and I agree with him. In the long list of books Lawrence Harrison mentioned I think that the one by my colleague Robert Putnam is the only one with a serious empirical analysis. Yet, though I like this book immensely, there are huge problems in using the data to identify a causal effect of culture on democracy and economy. Indeed, it was these problems that Guido Tabellini tried to address in his important paper, though my feeling is that he did not do so in a satisfactory way. I should add however that the sense in which Putnam talked about culture, or “social capital” is far from the emphasis on values and beliefs which Lawrence Harrison started out using as definitions of culture.

This lack of conceptual clarity is manifest in Stephen Lewis’ discussion of Botswana. I enjoyed these comments a lot, and I personally have learned a lot from Stephen Lewis’ work–he wrote a seminal study of economic policy in Botswana–and from him in person. He pointed out that at one level there was not much of a difference between the cultural and institutional interpretations of Botswanan success, but I think this reflects the fact that “culture” is used in very vague ways. I think I can describe what was important about the political institutions of the Tswana states and how this influenced the ways political elites behaved, though I admit there are many things I do not understand. This may be isomorphic to a claim about Tswana culture, but I find the latter much harder to measure or conceptualize.

Also from this issue

Lead Essay

  • In this month’s information-packed lead essay, Lawrence E. Harrison notes that the role of culture has been badly neglected in serious studies of economic devewlopment. But then, he asks, what explains “why, in multicultural countries where the economic opportunities and incentives are available to all, some ethnic or religious minorities do much better than majority populations?” Harrison reports some results of his recent Culture Matters Research Project, including the finding that “Protestant, Jewish, and Confucian societies do better than Catholic, Islamic, and Orthodox Christian societies…” Harrison provides a number of incisive country case studies, illustrating different ways pre-existing culture can produce economic results, and the ways policy and politics can transform culture.

Response Essays

  • In his reply to Harrison’s lead essay, University of California, Davis economist Gregory Clark writes, “I simultaneously want to endorse [Harrison’s] promotion of culture, and to run screaming from his lethal embrace.” While agreeing that the failure of purely institutional explanations of historical economic growth “opens the door … for culture,” Clark argues that “attempts to introduce culture into economic discussions so far have been generally either ad hoc, vacuous, blatantly false, or void of testability.” Clark points to great variation in economic performance within cultures and religions, and worries that Harrison’s “measures are not a pure probe into the essence of local cultures, but reflect institutions and economic environments that change the real possibilities for people.”

  • In his reply to Lawrence Harrison’s lead essay, George Mason University economist Peter J. Boettke argues that it is not culture but institutions—“the rules of the game that govern the way that people interact with one another”—that are the primary determinant of economic growth. However, culture may be crucial, Boettke argues, since it is “a tool for the self-regulation of behavior” that may raise or lower the cost of monitoring and enforcing compliance with “the rules of the game.” And that can make the difference between the success or failure of growth-conducive institutions and policies such as “private property, freedom on contract, limited scope of regulation, monetary restraint, fiscal responsibility, and open trade.”

  • James A. Robinson of the Harvard University Department of Government argues that Harrison’s measures are insufficient to establish that culture is the x-factor in economic development. For example, Robinson argues that the relative success of certain ethnic and religious minorities may be due to concessions from the majority group, and not the features of the minority culture. Also, Robinson asks, if the economic success of Chinese minorities in other countries is “because they have such a good culture, then why is China one of the world’s poorest countries?” And if Chile’s success lies in its distinctive culture, “then why did it manifest itself so recently?” Robinson concludes that “culture might matter, but doubters like me will not be convinced by the evidence here.”