The Role of the National Security Adviser

Since the Kennedy administration, the assistant to the president for national security affairs (a.k.a. "the national security adviser") has played two roles: manager ("honest broker") of the day-to-day policy process and substantive policy adviser. Presidents clearly want both, but the roles are in tension. Specifically, an assistant who pushes his own views too strongly risks losing the trust of a secretary of state (or defense) who has a different opinion. Some national security advisers have balanced these roles adroitly. Others have not, generating discord within the president's senior advisory team.

In this NSC oral history roundtable, five former national security advisers, serving six presidents from Lyndon Johnson to Bill Clinton, discuss how they personally dealt with this tension, how they played the role more generally, and how (with the benefit of hindsight) they think it should be performed. The discussion, chaired by Brookings Institution President Michael Armacost, addressed a set of questions (Appendix A) circulated in advance. But the discussion also ranged well beyond the questions themselves, as the participants offered a wide range of insights. In addition to the roundtable, we were able to interview, for this record, three national security advisers who were unable to participate. These interviews are reprinted as appendices.

This roundtable is the fourth in a series held by the National Security Council Project, co-sponsored by the Center for International and Security Studies at the Maryland School of Public Affairs and the Foreign Policy Studies Program of the Brookings Institution. Transcripts of three previous roundtables — on the Nixon NSC, on the role of the NSC in international economic policymaking, and on the Bush NSC—have been published previously and are available on the Internet. Two additional transcripts — on the NSC and U.S. Policy toward China and on the NSC and arms control policy — will be published in the near future. Other roundtables are planned for the fall of 2000. These seminars have been conducted for their own independent value. They also will provide useful input to a report on the NSC we plan to publish in the fall of 2000 and a book to be published in 2001

We are grateful to the participants for coming and talking with candor and insight. We would also particularly like to thank Karla Nieting for her help in organizing the roundtable and her work with the participants in bringing this edited version of the proceedings to publication. Responsibility for any remaining errors rests with us.