At the recent IHS conference there was a student presentation challenging the work that Pete Leeson and I have done on "Robust Political Economy". During my lecture at Comparative Historical research as well as my discussion group on anarchism as a progressive research program, there were questions about case selection and comparative assessment.
I am not satisfied with the answers that I gave to these very thoughtful criticisms. I view this as a failure in teaching on my part because Pete and I (and Chris Coyne (and Ed Stringham, Ben Powell, Scott Beaulier, Bob Subrick, Frederic Sautet, Virgil Storr, Dragos Aligica, Anthony Evans, Dan D'Amico, Adam Martin, and David Skarbek)) have actually thought about this a lot and written methodological papers to address concerns and justify our shared research program. But after the IHS experience, I started thinking that our essays have fallen short of capturing our actual position. We talk about it among ourselves, but we stress the actual writing up of the work and publication in the professional journals rather than the reasons either for the work or how it is that we decided to write on Soviet black marketeers, birth of the Amsterdam Stock Exchange, Somalia, Sunnies and Shiites, pre-colonial Africa trade, 16th century Scottish borderland, the law merchant and international trade, pirates, la cosa nostra, chicano prison gangs, San Pedro prison, Gypsies, etc.
What is critical in this work and what unites it methodologically is not so much topic choice, but why those topics are worthy of serious study. Some (such as the issue of post-conflict efforts to rebuild a country) justify themselves, but most of the studies in this line of research could be interpreted as a form of "cuteonomics" or "cleveronomics" (a GMU version of "freakonomics"). Now I have often said that economics is both a deadly serious subject that addresses deadly serious topics, and a wonderfully entertaining and enjoyable subject that should be read with great pleasure as well as benefit. And the examples of these two styles of work that I give are Chris Coyne's After War and Pete Leeson's The Invisible Hook. And we should never back off pushing both styles of economics and political economy research and publishing. But this is not the point that I want to stress at this point. The GMU brand of "cleveronomics" is not merely clever, but smart and good political economy.
The main point is that these cases are all selected for a reason. What unites our joint research program is a shared commitment to two core ideas from Ludwig von Mises's Human Action. First, the methodological commitment to the universal application of a rational actor/methodological individualist perspective to the analysis of social problems. Second, that the core analytical problem of economics is one of explaining how social cooperation under the division of labor is possible let alone achieved without any central direction/command.
With these two ideas firmly embedded in our consciousness, the first path is an analytical one of squaring both these ideas derived from Mises. In other words, how would a rational actor approach explain the mechanisms by which social cooperation under the division of labor would result without central command. In Mises's own brilliant discussion of these issues Ricardo's Law of Association does the work --- recognition of the gains from specialization (principle of comparative advantage) and the gains from trade. Mises's analysis takes place within an environment of private property rights, freedom of trade, free prices, and rational economic calculation. The flip-side of Mises's positive assessment of Ricardo's law of association, is his equally brilliant analysis of the problem of economic calculation in the absence of well-defined and strictly enforced property rights.
Social cooperation without command; economic calculation in generating complex coordination of economic activities are the animating ideas. There is an easy or ideal environment to demonstrate these ideas, but this is not that persuasive to those who are not already predisposed. In fact, if you take a critic of laissez faire such as Joseph Stiglitz what he has tried to do is to demonstrate that the Walrasian case of the efficacy of the market economy is vulnerable to slight deviations from ideal conditions. If the informational assumptions are slightly relaxed, or the market structure assumptions are slightly relaxed, then the Walrasian argument disappears. In other words, if we are dealing with agents who are imperfectly informed, and they are interacting in an imperfect market, then arguments for laissez faire disappear. Of course, for followers of Mises-Hayek, this doesn't make sense. Markets work in realizing the gains from trade and the gains from innovation precisely because economic actors have different information, and the competitive process is an ongoing rivalry for market power (and where price is a variable rather than a parameter). In the Mises-Hayek analysis of the market, dispersed (and often divergent) knowledge is utilized by individuals throughout the competitive market economy to pursue (discover) opportunities for mutual gain from trade and innovation and generate wealth and generalized prosperity. In other words, it is the current inefficiencies that spur the entrepreneurial discoveries of arbitrage opportunities, cost saving technological innovations, or new product offerings to consumers. Today's inefficiency is tomorrow's profit opportunity for that enterprise individual or group of individuals who figure out how to effectively address it.
So part of the Mises-Hayek argument is to demonstrate that Stiglitz overstates the fragility of the case for economic laissez-faire, and that instead complex coordination is more robust than Walrasian theory would suggest. However, a flip-side of the argument is also to demonstrate (with the aid of Buchanan-Tullock) that Stiglitz's own case for government intervention to correct for what he has identified as 'market-failures' due to problems of economic calculation, problems of disperse information, problems of bureaucratic incentives, and problems of interest group manipulations. In other words, government intervention designed as a corrective to social ills is fragile to deviations from ideal conditions of government policy. So the first attempts at developing "robust political economy" were related to the assumptions of omniscience and benevolence, and the argument (as laid out by me in my review of Stiglitz in the JEL and then developed further in my introduction to The Intellectual Legacy of F. A. Hayek (3 volumes)) was that if we relax the assumptions of omniscience and benevolence, economic liberalism proves to be robust against deviations from ideal conditions, while socialism proves to be extremely fragile against slight deviations from ideal conditions.
What Pete and I did in following up on this argument is try to develop the argument that for comparative analysis it was most effective to assume best case assumptions for intellectual opponents, and worst case assumptions for yourself. In other words, as Hayek argued in "Individualism: True and False" is that liberalism was a political economy doctrine that tried to take men as given (neither saintly nor brilliant). The goal was to find a set of institutions in the political-legal nexus where bad men could do least harm. A system that didn't require perfect men for its functioning, or require men to become perfect in order for the system to operate effectively. Socialism, on the other hand, did require such perfection for its functioning. The Mises-Hayek analysis assumed benevolence, but challenged the assumption of omniscience, while the Buchanan-Tullock analysis assumed omniscience, but challenged the assumption of benevolence.
Simultaneously with this attempt to lay out a framework for comparative analysis based on robust political economy, we were also exploring how robust the claims for Ricardo's Law of Association were in environments that were less than ideal. Developments in economics in the 1990s had demonstrated that cooperation without central command, and in the absence of a working legal authority, could be proven to emerge in small group settings of homogenous agents who possess low discount rates. In that population environment, reputation and ostracism would be enough to ensure social cooperation. But we also knew that the benefits of the division of labor and the gains from trade grow as social distance in the population gets greater. In other words, it is precisely cooperation among strangers and the coordination of economic activities over great social and physical distance. The classic example of Adam Smith was the number of exchanges (which would exceed all efforts at computation) required to produce even a common-woolen coat; Leonard Read (and Milton Friedman) made this point in the modern economic discourse with the example of a #2 pencil. Individuals who do not know each other, will never know each other, who are separated by great physical distance, and often speak different languages, hold different religious beliefs, etc. cooperate with one another without each others knowledge of what the other is doing and why it matters to produce products which are common-place and often essential tools in our daily lives. Think parts of a car; think parts of a computer. Even a tee shirt travels the globe before being produced. Paul Seabright wrote about this in The Company of Strangers but our emphasis was slightly different (though complimentary).
How do individuals turn situations of conflict into opportunities for cooperation even in the most unlikely of circumstances? Pirates were not studied by Pete only because he went to Disney World as a kid, or that popular movies make Pirates sexy, or that Pirates have recently emerged in the news again. All of these reasons are reasons enough for any scholar to study whatever subject they pursue. But there is something else going on as well. In Pete's studies of pirates, the population under examination is actually quite large, consists of heterogeneous actors, and almost by definition (since they are outlaws) have high discount rates. Yet, as Pete demonstrates, these pirate societies which operated outside any third party enforcement were able to achieve an amazing level of cooperation and coordination.
The question is by what mechanism are groups like the pirate societies able to achieve cooperation and complex coordination. These mechanisms come in the form of signaling strategies within games and commitment devices within games. It is through signaling and commitment that situations of conflict are turned into opportunities for cooperation. Social cooperation under the division of labor emerges and the complex coordination of economic activities is realized even under the most undesirable of circumstances. That is the point.
The group of scholars that first coalesced under me, and now are more often coalescing under Pete and Chris (rightfully so) pick cases for which their priors about the possibility of cooperation without command and in the absence of a third-party enforcer are theoretically least likely to result. Yet, they demonstrate that even under these unfavorable circumstances, social cooperation under the division of labor results. This is the comparative science of association that I have tried to pursue from my studies on how the Soviet economy operated to my exercise in intellectual history on the work of Lin and Vincent Ostrom. This is what I mean when I advise PhD students to "look out the window" and to explore cases where "history appears to defy what logic dictates". It is a Misesian project, both in inspiration and in method. But it is also an empirical project that could not get off the ground without the theory of social cooperation under the division of labor as developed from Adam Smith to Mises-Hayek as well as modern developments in political economy.
So what is the central point --- pick cases that are most unfavorable to the argument you want to make, not those that are most favorable. Don't just look for social cooperation without command and without 3rd party enforcement among a population of small groups of homogenous agents with low discount rates; instead look at populations of large groups of heterogeneous agents with high discount rates. If you can find social cooperation in that world, then you have unearthed a "robust" result. What challenges remain for this approach? There are many, but perhaps the most challenging are (a) the problem of scalability, and (b) the problems of conceptual clarity and empirical generalizability. But we see these challenges are opportunities for future research of a theoretical and empirical nature. However, it is important to realize as we step back that there is a methodological argument for the comparative historical approach to political economy and the research program of the positive political economy of anarchism.