Quantitative Approaches to Political Intelligence: The CIA Experience

From the Author\nIn the early 1970s, CIA formed a methodology office to examine quantitative methods that had been developed in the 1960s during what was called the behavioral revolution in academic political science, and to test how these methods could be applied in intelligence analysis. The creation of this office was ordered by CIA Director William Colby in response to criticism by prominent academics that the agency's analytic methods were way out of date.\nI chaired a panel that presented examples of this office's work at the 1977 meeting of the International Studies Association. This book publishes those panel presentations. My opening chapter summarizes our experience and some of the lessons learned. It describes how CIA's political research objectives condition its methodological procedures and what kinds of methods and techniques are particularly well suited to the CIA's needs. It became apparent that social scientists commonly define policy-relevant research far more broadly than the foreign policy community does, and that there were not a great many relevant methods and proven models ready-made for our use.\nThis chapter discusses fundamental differences in perspective between the intelligence analyst and the academic researcher. For example, the academic researcher is relatively free to define a problem in his or her own terms, while our research problems are generally defined by the requirements of U.S. foreign policy. The academic researcher chooses a topic for which data are available, whereas it is often new problems (or old problems defined in new ways) for which the policymaker requires intelligence analysis. Quantitatively oriented scholars can easily limit their work to those variables that can be operationalized, but the government analyst seldom enjoys that luxury. Government analysts are far more concerned with matters of presentation. They are writing for an audience that, by and large, does not understand the procedures or tolerate the jargon of social science methodology, and they must keep their presentation brief if they want it read by persons in authority.\nThe discussion of these differences is as relevant today as it was 32 years ago. A copy of this initial chapter is available from the author.