Case Studies and Theory Development in the Social Sciences - Alexander L. George [145]
It is often assumed that policymakers do not make use of generalizations in diagnosing and prescribing. This view is mistaken. Indeed, one of the major tasks of policy-oriented scholars is to discourage the decision-maker from applying oversimplified generalizations for purposes of policymaking. It is not only academic researchers with a passion for correlating only two variables who can be charged with engaging in “crude empiricism;” the policymaker, too, is often a crude empiricist. He or she can make highly dubious use of univariate propositions of the form: “if A, then B”—for example: “If appeasement, then World War III.” However, the decision-maker does not always operate as a crude empiricist. He or she often goes beyond available generalizations to note, in addition, what is special about the case at hand. We need to study and learn more about what a person does when he or she “goes beyond” available generalizations to deal with a single case. Evidently, he or she is trying to assess other relevant variables—not included in the generalizations—and the possible interactions among these many variables to make a judgment about the present case that goes beyond a crude probabilistic treatment.
Figure 12.1. Knowledge and Judgments for Policymakers.
Figure 12.1 indicates how the three types of knowledge that policymakers need—abstract conceptual models of strategies, generic knowledge of strategies, and actor-specific behavioral models of adversaries—enter into policy analysis within the government .532 The diagram also reflects the fact that policy analysis requires specific intelligence and information about the situation, and that the decision finally taken by the top policymaker is influenced by one or more of the seven types of judgment that need to be made in choosing among available options.
Implications for Scholarly Research and Policymaking
The preceding analysis has the following six implications for researchers. First, however desirable in its own right, theory and knowledge that fully meet scientific standards are not essential for the sensible conduct of foreign policy.533 As early as thirty years ago the distinguished psychologist Donald Campbell noted that “we must not suppose that scientific knowing replaces common-sense knowing.” Similarly, James March has endorsed “a perception of theory as contributing marginally to ordinary knowledge rather than summarizing all knowledge.”534
Just as intelligent people generally manage the many chores of everyday life reasonably well without benefit of fully developed scientific knowledge, so too can intelligent policymakers use the best available knowledge of different aspects of international affairs. To be useful in policymaking, conditional generalizations about the efficacy of a strategy need not satisfy the high degree of verification associated with scientific knowledge. Of course, policymakers would like the general knowledge on which they base decisions to be as valid as possible, but in practice they will settle for more modest levels of precision. When verification of conditional generalizations is limited, policymakers can still make use of them, even though the generalizations have limited empirical support and therefore are only plausible.535 By drawing on available information about a particular case, policymakers can often judge whether the plausible generalization is likely to hold for that situation.
Second, academic scholars should include manipulable variables, variables over which policymakers have some leverage, in their research design. Strategy is just such a variable. We noted earlier the limited relevance for policymaking of quantitative correlational research on international relations that deals only with non-decision-making variables in attempting to account for variance in foreign policy outcomes.
The importance of “leverage variables” is noted by sophisticated policymakers, who are at times explicit about their need for information about the critical variables in a situation. Dean Acheson,