Enemies of Intelligence: Knowledge and Power in American National Security
Richard Betts. New York: Columbia University Press, 2007. 193 pages plus notes, index.
Reviewed by Nicholas Dujmovic
Twenty years ago, the economist and essayist Thomas Sowell published a persuasive treatise on the history of ideas, A Conflict of Visions, in which he posited that political differences and policy preferences stem not so much from one's political priorities as from one's view of the essential nature of man.1 There are two kinds of people in the world, Sowell argued: those with a "constrained vision" see mankind as unchanging and imperfect, while those with an "unconstrained vision" view man as a malleable, changeable, even perfectible animal. Policy choices follow accordingly.
Likewise with the field of intelligence. Whether as process, product, or profession, it comes down to people, and one's view of intelligence and what ought to be done about it ultimately is shaped by how one views people, with all their virtues and achievements, vices and shortcomings.
In this important collection of essays, Dr. Richard K. Betts of Columbia University demonstrates the consistently clear thinking that has marked his writings on intelligence for the past three decades.2
Though he does not mention Sowell's work, Betts's work is fundamentally an application of Sowell's thesis to intelligence. Betts seems to suggest that, when it comes to intelligence, there also are two kinds of people: those who believe intelligence can be made perfect or nearly so--all we need is the right reform package--and those who doubt that any kind or degree of reform can prevent failures.
Betts addresses what too often has been lacking in the national debate about intelligence and its reform since the attacks of 11 September 2001: a sober and realistic assessment of what intelligence can be expected to do and, more importantly, what it cannot reasonably be expected to do because of its built-in, and therefore unavoidable, limitations.
These limitations, which Betts calls "enemies of intelligence," are behind all intelligence failures, so they bear scrutiny. Betts groups them into three categories. The first is obvious: "outside enemies" are literally our national enemies--the foreign adversaries whose capabilities we must divine, whose plans we must thwart, and whose allies here must be caught.
Betts's second category, "innocent enemies," consists of organizational shortcomings that cause failure, including institutional myopia, negligence in standard procedures, gaps in coverage, inefficiencies caused by organizational redundancies, the lack of particular skill sets--the kinds of things that bureaucracies, particularly government bureaucracies, do or don't do out of institutional legacies or laziness. Betts finds that most of the debate about intelligence focuses on this category of enemies: if we hire better people and organize them properly, it is widely assumed, we can prevent intelligence failures.3 Fix the wiring diagram of US intelligence and all will be well.
Betts is having none of this; he is skeptical about the efficacy of organizational reform in eliminating failure not only because intelligence is genuinely challenged by the guile of outside enemies but because of "inherent enemies," his third category of enemies. These are the limitations that are part of the human condition and that exist in the nature of the practice of intelligence itself. They "pervade the process no matter who is involved, and they intrude time and time again. Although not immune to defeat, they are extraordinarily resistant."4
Critics of intelligence often do not appreciate, usually because they've had no relevant experience, that the demands placed on the human brain by the requirements of intelligence operations and analysis quite literally can go beyond reasonable expectations regarding perception, reasoning, memory, and imagination. As Betts observes, "cognition cannot be altered by legislation."
Critics also often fail to recognize that the practice of intelligence always involves trade-offs: analytic accuracy versus timeliness, organizational centralization versus pluralism, the need to share information versus the imperative for security, the benefits of expertise versus the fresh views of non-experts, to mention just a few of the prominent dilemmas. The WMD Commission in 2005, for example, recommended that the Intelligence Community create more centers to achieve "fusion" in analysis; Betts points out the downsides of such centers, including diminished competitive analysis, creation of new "stovepipes" for information, and entrenched large, new bureaucracies.
Expectations must take these realities into account. Unrealistic expectations, like a "zero defects" standard, will hinder effective reform. Betts is particularly critical, and with good reason, of those who think that we can prevent intelligence failures by improving analytic procedures. Tweaking processes, of course, can improve analysis, but such improvement will be marginal rather than radical, according to Betts, for the simple reason that intelligence failures most often result not from analytic error but from mistakes by policymakers: "Failures occur more often at the consuming than the producing end of intelligence."
Ultimately, Betts's argument seems to be directed more at policymakers and other consumers of intelligence than at practitioners of intelligence, who already know their craft's inherent limitations. Policymakers, says Betts, can't have it all when faced with the dilemmas practitioners face: they cannot insist on the benefits of multiple opinions and on a single, coordinated analytic line; or force a consolidation of resources while expecting intelligence to maintain worldwide coverage; or demand analysis that is honest in describing complex realities and unambiguous. Policymakers have the responsibility to learn about intelligence and its limitations, to provide guidance, and to understand that intelligence paradoxically provides the unwanted news the policymaker needs, and should want, to hear. Above all, Betts wants policymakers to set priorities and accept, simply as a mature, realistic stance, that there will be the occasional and unambiguous disaster and that, by contrast, intelligence successes are difficult to measure and assess (and therefore to take credit for). We tend not to notice the disasters that did not happen.
None of this is to let intelligence professionals off the hook. Improvement should always be worked for, and this requires intelligence officers to press policymakers for guidance, to show some independence in implementing requirements, and even, on occasion, to present unwelcome news in such a way that it cannot be ignored--what Betts calls "grabbing the lapels" of the policymaker. National security is not served when intelligence considers "the customer" always right, or says "yes" when a "no" is warranted. Though Betts does not say so directly, this was the major problem of the directorship of George Tenet, as one can conclude from Tenet's recent memoir.5
The most controversial part of this book concerns Betts's conclusions about the proper balance between security and freedom. When it comes to collecting intelligence to prevent another 9/11, Americans need to realize that the right to privacy is not on the same plane as other civil liberties: "It is reasonable to invade the privacy of some citizens in order to gain information that might help protect the lives of all citizens." Many will disagree, and Betts admits his argument derives from a practical sense of what needs to be done rather than a legally recognized "hierarchy" of rights in the Constitution.6 Nonetheless, Betts deserves attention on this score. The panic that would result from another dramatic attack, he argues, will "sweep away" our devotion for civil liberties generally--Betts is particularly concerned about due process for citizens--unless careful steps are taken beforehand to reduce the expectation of privacy, with proper precautions to limit the government's use of such information, not its ability to collect it.7
A brief review cannot do justice to this rich and densely argued book. Betts uses historical cases well, particularly on the issue of politicization during the Vietnam war, the Team A & B controversy over Soviet strategic forces in the 1970s, and Iraq analysis in 2002 and 2003. Other important points that Betts makes include:
It is possible to be wrong for the right reasons and not because analysts were somehow derelict; we have to avoid the temptation to define "sound" judgments solely as those that turn out to be accurate.
The main intelligence failure before the 9/11 attacks was the "insufficient collection of unambiguous information," but the trade-offs from a maximum effort includes the loss of focus and the risks taken by deployed collection platforms (the Pueblo, e.g.). A realistic reform proposal would be to institutionalize an analytic process that included the views of a generalist (or non-specialist) known for "exceptional thinking" to argue the case for discontinuity in any major estimate or analytic assessment.8
The twin failures of 9/11 and the 2003 Iraqi WMD estimate made intelligence reform politically imperative, but the resulting structural change left basic questions unanswered: for example, the DNI is basically the same figure as the DCI, but DCIs could have had more authority if only presidents were willing to give it. Will presidents or the Congress give the DNI the authority he needs, that is, over military agencies?
Whether one agrees with all of Betts's conclusions, this illuminating discussion of intelligence in the post-Cold War age is necessary reading for the intelligence professional, and for those served by the profession. I would also recommend its use in academic courses dealing with intelligence and reform.
2. Among Betts's voluminous writings on national security, two of his works on intelligence, separated by a quarter century, stand out as classic essays: "Why Intelligence Failures are Inevitable," World Politics (October 1978); and "Fixing Intelligence," Foreign Affairs (January-February 2002).
3. The organizational focus of the history of intelligence reform proposals and prescriptions can be seen in Michael Warner and J. Kenneth Macdonald, US Intelligence Community Reform Studies Since 1947 (Washington, DC: Center for the Study of Intelligence, 2005).
6. One reviewer, Steven Aftergood in his blog on the Federation of American Scientists Web site (www.fas.org/blog/secrecy/2007/09/in_print_enemies_of_intelligen.html), asserts that Betts's argument would turn the US Constitution itself into an "enemy of intelligence." This is a misreading that ignores the boundaries of Betts's categories and his advocacy of ways to preserve civil liberties against government encroachment.
7. Others have argued for the need for "rules of engagement," both to protect American civil liberties and to allow intelligence officers to do their jobs. See David Robarge's review of James Olsen's Fair Play: The Moral Dilemmas of Spying (Washington, DC: Potomac, 2006), in Studies in Intelligence 51 no. 1 (2007).
8. This is hardly a fresh idea (see Bruce Reidel's discussion of the Israeli use of such a system in the Winter 1986 issue of Studies in Intelligence), but sometimes new situations require fresh looks at old ideas.