Intelligence-policy relations and the problem of politicization

• Main Author: Rovner, Joshua Randall
• Other Authors: Harvey Sapolsky.
• Format: Others
• Language: English
• Published: Massachusetts Institute of Technology 2009
• Subjects: Political Science.
• Online Access: http://hdl.handle.net/1721.1/46633

Thesis (Ph. D.)--Massachusetts Institute of Technology, Dept. of Political Science, 2008. === Includes bibliographical references (p. 399-414). === A growing literature in international relations theory explores how domestic institutions filter and mediate international signals. The study of intelligence-policy relations fits naturally into this mold, because intelligence agencies are specifically designed to collect and interpret information about the international environment. This study provides a general framework for theorizing about intelligence-policy relations by exploring how leaders respond to new intelligence estimates. In addition to providing a deductive characterization of the intelligence-policy problem, the dissertation presents a model of politicization, defined as the manipulation of estimates to reflect policy preferences. When leaders commit themselves to controversial policies, they have strong domestic political incentives to put pressure on intelligence agencies to publicly support their decisions. Intelligence agencies control secret information and presumably have access to sources that are unavailable elsewhere. For this reason, the use of intelligence for policy advocacy is a uniquely persuasive kind of policy oversell. The dissertation tests the model in a series of pair-wise comparisons. The first pair of cases explains why the Johnson administration first ignored and later politicized intelligence on Vietnam. The second pair explains why, despite their differences, the Nixon and Ford administrations both ended up politicizing intelligence on the Soviet strategic threat. The last pair of cases compares the U.S. and British responses to intelligence before the recent war in Iraq. The results of the study show that domestic variables identified in the oversell model strongly affect the likelihood of politicization. Organizational and individual-level explanations are less satisfying. === by Joshua Rovner. === Ph.D.