Case Selection in Qualitative Research in Political Economy

|Peter Boettke|

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.

Comments

[Less Antman]

Beautifully written: in one essay you have briefly summarized the economic case for anarchy, identified the most important insights supporting it, and described the means by which we can persuade others.

Over the three decades I've been an activist, I've become persuaded that the hard cases are the most important to address in public discussions. Many times in recent years, I've discovered that some economist with a connection to GMU has written a paper that gives me just the ammunition I needed in such discussions. As your post has made clear, this is not a coincidence.

[Guillermo Barba]

Brilliant article.
Perhaps a complementary approach is to show, in as many cases as possible, how government inteventions to correct for 'market failures' are far worse that coordination resulted from relatively unhampered markets, especially in the medium and long term.

[Jerry O'Driscoll]

Critics need to adduce their own, contrary studies.

[Daniel Klein]

Smith's organon of sympathy is a sort of presumption of man inherently immersed in cooperation. His organon is to see the cooperation in any human action. (The disregard, even defiance, of this organon, btw, is why I disliked the film No Country for Old Men.) The issue is cooperation in what?

I submit that that perspective really alters the issues you are dealing with.

I haven't followed Pete L.'s work closely, but it is my impression that he really doesn't come to terms with the fact that for the most part pirates were cooperating in evil. Greg Clark's JEL review also raised this matter.

Here is Smith's sole remarks about pirates:

"Great warlike exploit, though undertaken contrary to every principle of justice, and carried on without any regard to humanity, sometimes interests us, and commands even some degree of a certain sort of esteem for the very worthless characters which conduct it. We are interested even in the exploits of the Buccaneers; and read with some sort of esteem and admiration, the history of the most worthless men, who, in pursuit of the most criminal purposes, endured greater hardships, surmounted greater difficulties, and encountered greater dangers, than, perhaps, any which the ordinary course of history gives an account of."

RE Chris's important work, my concern is how we deal with the counterfactual of a world without the US government raining its powers on regimes of evil cooperation. How much evil has "Team America" prevented? Chris shows that observed efforts at nation-building so often seem to fail. But does he show what would happen in a world without the US bully primed to beat on other bullies? I am entirely sympathetic to Chris's positions on these matters, but does he deal with this facet of the issue? That facet is really hard to deal with, of course, but does Chris adequately acknowledge it?

Again, Smith's organon is cooperative man, and the issue is cooperation in what. Hayek's work on man's instincts and our inheritance from the Paleolithic band fits nicely with this organon. Hayek says in FC that the "Hobbes problem" has all along been something of a red herring. Deirdre says the same. Yet you still seem to be treating the Hobbes problem as primary. "Cooperation" to you means voluntary association, but there are no grounds for confining the term is such fashion. We need subscripts on "cooperation." Smith is not about achieving cooperation, he is about discriminating among types of cooperation. The virtue of his organon is to refocus on that latter concern. It is a cultural project in making our cooperative penchants more enlightened. It is a cultural project, not in finding a sense of duty, which is essential to man, but in improving the character of our duties. (BTW, I just saw the film Roman Holiday -- an exquisite piece on doing our duty, advancing to a higher self-interest, and the poignant, becoming sacrifices that doing so entails. The becoming use of what is our own.)

Your words on this matter are laden with something of a presupposition that, in any instance, if govt power is stayed, then it is spontaneous voluntary cooperation that will follow. But if one bully is stayed maybe it is local, more grievous bullies that bubble up. Bottom-up evil.

The view of man as inherently cooperative puts Pete's work in a quite different light. If a relatively liberal bully is available, why not use it to put down pirates? Doing so can augment overall liberty, even though some liberty is sacrificed in the process. Pirates, or evil factions, are a primary justification for government power. Another question: By what cultural process would we anticipate pirates graduating toward liberal enlightenment? Maybe when the sea is tranquil they would hold collquia on Locke on toleration?

I think that the body of work you write of gets at good libertarian points on the hubris and folly of bullying, but that response has to do, not so much with the documenting of mere cooperation, but in understanding the evolution of what it is that people cooperate in.

[Peter Boettke]

Dan,

I think both Chris and Pete in various papers have addressed your comparative point --- see the work on Somalia as well as the work testing the "normal country" thesis, etc. But obviously thinking about the counterfactual is a point that we all think is important since we are doing comparative political economy. I've been pretty concerned about the counter-factual and the fallacy of "after this, therefore because of this" causing all sorts of interpretive problems for us.

Regarding the point about pirates, bandits, prisoners, black marketeers, etc. --- there are two issues there, first I don't think you are acknowledging the methodological point about complex coordination among actors that theory predicts will not be able to achieve such cooperation without third-party enforcement; second, we recognize your point about "graduation" --- the move from personal to impersonal exchange; the evolution of small scale trade and small scale capital accumulation to medium scale trade and medium size capital accumulation to large scale trade and large scale capital accumulation (from subsistence to exchange in Bauer's language). I tried to capture this under the rubric of scalability.

Anyway, we have a lot of work to do, but there are a lot of thinking behind what we are doing and why.

Pete

[Richard Ebeling]

I'm wondering if there is not complementarity and not competition between the two perspectives offered by Peter Boettke and Daniel Klein.

Pete's point, if I understand it correctly, is that the Mises/Hayek framework focuses on the potentials for cooperation and coordination that tends to develop "spontaneously" from any circumstance that seems to offer discovered mutual gains from trade, which, in turn, tends to generate rules of order among the participants for peace and predictability for their interactions.

The "case study" method offers a field for many examples of these processes at work that are able to build upon the a variety of "Austrian" presuppositions: intentionality and purposefulness in all conscious human action; explanation of the development of complex social orders from the actions and interactions of individuals; and emphasis on the qualitative processes through time through which such institutional orders emerge, take form, facilitate social life, and continue to evolve over time.

Thus, the Scottish philosophical and "Austrian" traditions are drawn upon to demonstrate the workings, logic, and historical reality of much in society that is "order without design."

Danny's point, drawing from aspects of Adam Smith's analysis, is that not all such orders are equal in terms of fostering certain types of rules, morals, attitudes and institutional arrangements for harmony and peace.

(This follows upon some earlier critics of Hayek who challenged what they interpreted as Hayek's view that all that developed "spontaneously" was, therefore, inherently superior and "good." I believe that this interpretation of Hayek, by the way, is incorrect. Hayek often admitted that certain spontaneous orders were less "progressive" than others and even potentially social dead-ends.)

Speaking very broadly, as an example, there is the comparison of ancient Athens and Sparta. Both "spontaneously" developed sets of social institutions and rules of social behavior and conduct. But during the high water-mark of Athenian civilization there was the fostering of a culture of reason, science, tolerance and a degree of freedom (though far from perfect for either free citizens or certainly for the many slaves). Sparta developed a more militaristic social order that placed greater emphasize on collective obedience and far less understanding of and tolerance for human autonomy and free discussion).

Not only is there the issue of what type of "virtues" different social orders develop, foster, and respect. There is also the dilemma that collectivist orders and systems of thought often have greater (destructive) potential and power in the short-run that threaten the stability and survivability of more individualistic orders and systems of thought.

While liberal pacifists debate the "pros" and "cons" of forms of self-defense, collectivist aggressors attack and destroy. While the French and the Belgians debate whether it is a violation of freedom of religion to issue traffic tickets to orthodox Muslim women who cover their faces while driving, Islamic religious fanatics debate the best targets to plant bombs in urban areas. While liberals lose sleep over the ethics and expediency of capital punishment, crazy men who hear voices in the air figure out where best to end their lives as suicide bombers.

Now, before any who knows my strong non-interventionist and laissez-faire views on all matters concerning both domestic and foreign policy start to believe that I am drifting towards the "dark side of the force," I am merely pointing out realities and not preferences.

The types of social orders and the systems of ideas that have evolved, taken form, and mold both thinking and action, therefore, matter.

Just as pirates had their evolved rules of social order, so did bootleg gangs during Prohibition. And its all fine and good to understand the methods and procedures that Al Capone and other gangland bosses developed to minimize conflict among themselves, their customers, and the law enforcement agencies.

But a superior social order is one in which those gangland "sub-orders" do not and will not have to emerge and develop. And, that I believe, is what Danny is reminding us of, and which should be part of the interpretative and evaluative process of social analysis.

Richard Ebeling

[liberty]

Your main point and concluding paragraph I agree with - "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. .. 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."

However, I wonder whether this is too narrow:

"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."

Perhaps this is the Mises program -- but I see the insights of Mises and Hayek as being applicable much more broadly. First, I question strict methodological individualism. I always have, but more and more I find issue with it, and think that Austrian-influenced economists should question their own assumptions in this regard. Yes, humans act - and we must include the individual human actor in our models - but the individual's preferences are influenced by others, by social environment, by legal and other kinds of institutions, etc. (For more on this, see for example Hodgson, "Institutional Economic Theory: The Old Versus The New," in Prychitko, Why Economists Disagree.)

Second, although spontaneous order, coordination without command, etc are at the heart of understanding economics, yes, is the core job of the economist only to explain this? Shouldn't economists (or some economists - division of labor) concentrate more on e.g. (1) What happens when certain policies are taken, policies which many voters desire, and whether the costs do or do not exceed the benefits; (2) The flaws in markets, which may in fact be fixable (see #1), which may occur if e.g., some individuals are unable to work and hence cannot demonstrate their preferences, or because of public goods problems or something else...

So, I think that your research program - while commendable - may be more narrow than it could (should?) be. I think, in addition to theory, it is wonderful to use case studies--this is very important, I think. I know some GMU-related economists also supplement with econometrics, which if done properly is also a good idea. So, in terms of methods, I think I am squarely in your camp. But methodology--pure methodological individual--and focus I think are too narrow.

[Qualitative Research]

thanks for sharing.. I like this, because it related to my research