I am heading to Harvard tomorrow for a conference, hosted by the Harvard Negotiation Law Review, on the value of case studies and the role of lawyers in deal making. Vic at the Glom has blogged about the conference here. The conference is organized around Vic Fleischer’s case study on the MasterCard IPO, and David Millstone & Guhan Subramanian’s study of the Oracle/Peoplesoft takeover bid. I am very much looking forward to hearing the presentations. My paper discusses the antitrust implications of MasterCard’s IPO. More specifically, I attempt to pin down precisely what MasterCard’s new governance structure purchases (and at what price) in terms of reduced antitrust exposure, and touches more generally upon the role of the antitrust lawyer in and after such governance decisions. I will post my draft to SSRN shortly, and perhaps do some blogging after the conference.
While the MasterCard IPO raises some interesting antitrust issues, Vic’s post raises some tough questions about the value of case studies in legal scholarship that I hope to discuss along with the conference participants. Vic writes:

One of the challenges of using case studies is figuring out what lessons to draw from them. At a minimum, qualitative empirical work can be useful for generating hypotheses. I look for unusual cases and try to find patterns and theories that might explain what’s going on. But when we present case studies as scholarship, or use them in the classroom, aren’t we implicitly suggesting that they are representative of a broader pattern? But how do we know if these case studies are a sample of a broader phenomenon, or the universe of such cases? Should/must case studies be accompanied by proper quantitative research before they are taken seriously? If not, aren’t we encouraging our students and fellow scholars to violate the rules of inference? When one stumbles across an interesting case like MasterCard, or Google, or Ben & Jerry’s, what’s the best way to proceed? I’m looking forward to the discussion Wednesday as I try to shape my summer research agenda.

These are all good questions. I certainly don’t have all the answers. But I do have some initial reactions. The dangers of generalizing too broadly from single observation case studies are well known and intuitively obvious. Researchers ought to be clear about what their case study illustrates, what it does not, and what questions it raises. But let me at least take a stab at one of the questions and give a direct answer: No, case studies need not be accompanied by proper quantitative research before they are taken seriously.

There are obvious tradeoffs to be made between what is sometimes referred to as “quantitative” empirical work (i.e. regression) and “qualitative” empirics (i.e. case studies). One advantage of case studies is that they typically allow the researcher a greater understanding of the institutional details underlying relationships of economic interest. Case studies, for example, have had significant influence on shaping modern antitrust policy, i.e. the work of Benjamin Klein and others in vertical restraints. The Federal Trade Commission makes much use of case studies in evaluating the competitive effects of restraints in particular industries, which exhibit great variance in institutional detail. The theory of the firm, including Coase’s Theory of the Firm, and Klein’s (and Klein, Crawford and Alchian) Fisher-GM work, was also greatly advanced by the case study approach. Of course, there has been much progress in these fields (to name two) both in advancing the theory and empirically testing hypotheses derived from the theory.

I do not mean to suggest that broader quantitative analysis is not useful. Of course it is. It offers many advantages (namely, general application) that detailed case studies do not. I guess that brings us back to tradeoffs. The bottom line is that New Institutional Economics (and transactions cost economics) has taught the lesson that institutions (i.e. “the rules of the game in a society,” as Douglass North defines the term) matter. If this lesson is true (of course it is), case studies can be quite valuable with or without prior quantitative research.