Case Studies and Theory Development in the Social Sciences - Alexander L. George [140]
Quite early in pursuing the research program that has led to the publication of this book, we became aware of the need to move beyond structural-realist, rational-choice, and game theories. These deductive approaches “black-box” around both the process of policymaking and the strategic interaction between states that leads to foreign policy outcomes; they deal with these two processes by assumption and by refinement of assumptions. In our view, it is necessary to engage in the direct (and admittedly difficult) empirical study of decision-making processes and strategic interaction. However, deductive and empirical ways of developing knowledge and theory of international relations are hardly antithetical; as emphasized in Chapter 11 and elsewhere, it is desirable to link deductive and empirical approaches more closely together.521
Development of generic knowledge is not the only type of middle-range theory and knowledge that is relevant for policymakers. We shall note other types of scholarly research at the end of this chapter after discussing ways of developing generic knowledge of the tasks policymakers faced repeatedly (though in different contexts). (Throughout the chapter, we use the term “usable knowledge” as a synonym for policy-applicable theory and will comment later on the relationship of usable knowledge to scientific knowledge.522
What is Usable Knowledge?
We find it useful to address the challenge of developing usable knowledge by posing three more specific questions for discussion:523
1. What kinds of knowledge do practitioners need for dealing with different generic problems?
2. How can such knowledge be developed?
3. How can generic knowledge be used by policy specialists?
The answers to these three questions are interrelated. Scholars who acquire a realistic understanding of how generic knowledge can enter into the policy analysis that precedes and contributes to decision-making can better identify and develop the forms of knowledge that practitioners need. Policy specialists and decision-makers need much more specific information about particular situations they face than outside scholars possess. We must recognize, therefore, the limited usefulness of generic knowledge—but appreciate at the same time why it is indispensable for policymaking and how it should be used in policy analysis.
What Types of Knowledge Do Practitioners Need?
Turning to the first of the three questions posed above, policy specialists need a general conceptual model of every particular strategy or policy instrument that identifies the general logic associated with successful use of a policy tool. A discussion of the relatively simple concept of deterrence will illustrate this point. A threat to respond to actions an adversary may be contemplating is a critical component of general deterrence theory. The actions threatened for purposes of deterrence may or may not be ambiguous. In any event, the general logic of deterrence requires that threats to respond to possible provocations should be sufficiently credible and sufficiently potent to the adversary to persuade him or her that the costs and risks of the contemplated actions are likely to outweigh the expected gains. The logic of this abstract deterrence model, therefore, rests on a general assumption that one is dealing with a rational opponent who is able to calculate correctly the benefits, costs, and risks of actions he or she contemplates taking.
Two limitations of the usefulness of such abstract models for both theory development and policymaking should be noted. First, a general conceptual model is not itself a strategy but merely the starting point for constructing a strategy that fits a specific situation and is likely to influence a specific actor. The conceptual model identifies only the general logic—that is, the desired impact of a deterrent threat on an