Case Studies and Theory Development in the Social Sciences - Alexander L. George [213]
413
Chapter 2 illustrates the use of both methods in one research program.
414
Robert Bates et al., Analytic Narratives (Princeton, N.J.: Princeton University Press, 1998).
415
See, for example, Jack S. Levy, “The Role of Crisis Management in the Outbreak of World War I,” in Alexander L. George, ed., Avoiding War: Problems of Crisis Management (Boulder, Colo.: Westview Press, 1991), pp. 62-102; and Brent Sterling, “Policy Choice During Limited War” (Ph.D. dissertation, Georgetown University, Washington D.C., 1998).
416
Steven Weber’s book, summarized in the Appendix, “Studies That Illustrate Research Design,” illustrates how this can be done. See also Glenn Snyder and Paul Diesing, Conflict Among Nations (Princeton, N.J.: Princeton University Press, 1977).
417
For a detailed, rounded discussion of the similarities and differences between historical explanation and uses of history by political scientists to develop and test generalizations of theoretical interest, see “Symposium: History and Theory,” International Security, Vol. 22, No. 1 (Summer 1997), pp. 5-85.
418
Theories can be tested in two different ways: by assessing the ability of a theory to predict outcomes, and by assessing the ability of a theory to predict the intervening causal process that leads to outcomes (which we discuss in the present chapter).
419
Over the years, Jack Levy has published a number of articles that emphasize the failure of much early quantitative research on international relations to provide theoretical specification of possible intervening causal processes in correlational findings.
420
For a review of this literature, see Miriam Fendius Elman, ed., Paths to Peace: Is Democracy the Answer? (Cambridge, Mass.: MIT Press, 1997).
421
An example of research that makes this kind of contribution is Alexander George and Richard Smoke, Deterrence in American Foreign Policy, which is summarized in the Appendix, “Studies That Illustrate Research Design.”
422
This chapter cites a number of studies that have employed process-tracing; some thirty such examples are briefly summarized in the Appendix, “Studies That Illustrate Research Design.”
423
Harry Eckstein labels this type of study as “configurative-ideographic”; Arend Lijphart refers to it as an “atheoretical case study.” Harry Eckstein, “Case Study and Theory in Political Science,” in Fred I. Greenstein and Nelson W. Polsby, eds., Handbook of Political Science, Vol. 7 (Reading, Mass.: Addison-Wesley Press, 1973), pp. 79-138; and Arend Lijphart, “Comparative Politics and the Comparative Method,” American Political Science Review, Vol. 65, No. 3 (September 1971), pp. 682-693.
424
See Clayton Roberts, The Logic of Historical Explanation (University Park: Pennsylvania State University Press, 1996). See the discussion of historical explanation below.
425
Giovanni Sartori, “Concept Misformation in Comparative Politics,” American Political Science Review, Vol. 64, No. 4 (December 1970), pp. 853-864; and David Collier and Steven Levitsky, “Democracy With Adjectives: Concept Innovation in Comparative Research,” World Politics, Vol. 49, No. 3 (April 1997), pp. 430-451.
426
For a similar discussion of different types of causal relations, see Robert Jervis, System Effects: Complexity in Political and Social Life (Princeton, N.J.: Princeton University Press, 1997), pp. 34-60.
427
Levy (“The Role of Crisis Management in the Outbreak of World War I”) and Sterling (“Policy Choice During Limited War”) articles.
428
Both Eckstein and Lijphart offer typologies of case studies; their terminology differs but the types they identify are similar with two exceptions. Lijphart does not designate a category for Eckstein’s “plausibility probe,” and he adds the quite important “deviant case