New Paradigms Forum Proliferation Issues and Much More …


Relations Between Intelligence Analysts and Policymakers: Lessons of Iraq


This is the text of a Hudson Institute publication also available through the Hudson website.

It is important to return to a sense of balance in the relationship between intelligence analysts and policymaking intelligence consumers. Both parties need to understand the complex parameters of this relationship, finding the right sort of constructive tension at appropriate points while still respecting the special strengths and institution al roles each side brings to the table. This lesson of Iraq will be an important one for President Obama and his national security team to learn if they wish to govern competently and effectively.

No single person could possibly grasp—and keep up with—all the information necessary as a basis for decision-making in the foreign policy and national security arena, and this is more true the higher the rank of the officials in question.  Senior policymakers are thus inescapably dependent upon those who provide them with information, filtering the blizzard of data available and explaining to them the meaning of what small fraction is passed along.  In foreign policy and national security, this means, among other things, that officials cannot avoid being enormously dependent upon the professional analysts of the U.S. Intelligence Community.

But all is not right with the intelligence processupon which President Obama and his national security team will have to rely. They will need to fix this process if he wishes to govern as the effective and transformative leader he was elected to be.  The fundamental relationship between intelligence analyst and policymaking customer today seems to be broken, and perhaps perilously so.

The problem grows out of the Iraq war.  If there is blame to be cast, it lies most with the self-exonerating senior intelligence bureaucrats and the opportunistic political and media critics outside the Executive Branch who invented and popularized the notion that the problem with pre-war Iraqi weapons of mass destruction (WMD) analysis was that policymaking customers distorted intelligence for political ends.  It is an alliance between these groups, after all, that has worked very hard to enshrine a new, post-Iraq conventional wisdom that the problem with the relationship between analysts and customers has been that the latter did not give enough unquestioning deference to the former.  But assigning blame is at best a secondary point.Fixing things is more important.  The key point is to realize that this new conventional wisdom is both false and dangerous.

Intelligence analyses of Iraqi weapons of mass destruction (WMD) before the U.S. invasion were famously awful.But America needs to learn the right lesson from that debacle. At present, we seem well on the way to learning the wrong one.

The first problem with analysis of Iraq’s WMD was just that the analysts got things wrong. That is not necessarily unforgivable, however, for intelligence information is routinely highly incomplete or ambiguous, and generally concerns matters that others work very hard to conceal. We needanalysts to be able to be wrong, for they cannot do their job if they are paralyzed by a fear of error.

Less forgivable is the secondIraq-related failing: the breakdown of the analyst/customer relationship and the dysfunctional mythology that has been built upon its ruins. In this regard, the problem with pre-war WMD intelligence was not that the policy community compelled flawed analysis in order to justify predetermined policy goals. It was that the analysts themselvesoffered mistaken conclusions and the policy community accepted these conclusions without a second thought.  That’s just what I thought, the response seems to have been, so let’s go!

Interestingly, by comparison with the sorry tale of WMD analysis, pre-war analysis of Iraqi state-sponsored terrorist activityseems, in retrospect, to have been pretty good. This offers us important lessons.

Analytical products on Saddam’s terrorist connections were greeted with some skepticism in the policy community. Intelligence customers were apparently surprised that there weren’t moresuch connections, and felt that the analysts must have missed something. There thus developed an iterated process of substantive engagement between analysts and customers, in which they debated the facts and the meaning of available information. Policymakers in some cases took the time themselves to become familiar with the underlying intelligence reporting, and were not shy about questioning assumptions and forcing analysts to defend their data and their reasoning. As a result, the analysis improved. Analysts re-checked their work, some additional activities and connections were identified, and their reasoning was sharpened.

At the same time, customers came to understand that there really wasn’t evidence for the sort of deep and sweeping terrorist connections to the Iraqi regime that had been suspected, and they came to have greater confidence in the resulting work. The resultwas a body of analysis that apparently got things about right.

So what should all this teach us?  The lesson is that it is vital to maintain analyst/customer relationships characterized by a slightly skeptical, probing sort of substantive engagement. This requires effort and understanding from both sides.

Analysts must expect and even welcome this engagement, for it is essential to their doing their job.  It is not “politicization” to force a briefer to explain precisely why he or she has come to those conclusions, to worry about the reliability of the underlying reporting, to question analytical assumptions, to consider the possibility that collectors and analysts may be the victims of denial or deception techniques, or simply to demand more information before making up one’s mind. Ultimately, in fact, it is the policymaker’s prerogative entirely to rejectintelligence analysis if, to that policymaker, the conclusions seem mistaken. (If the Iraqi WMD example shows anything, after all, it is that analysts can be quite wrong. After such a dramatic series of analytical failures, would it not be perverse—or worse—to require that all future Intelligence Community work product be accepted at face value?)  Analysts need to remember both that they do not have a monopoly upon evaluating the facts, and that it is the exclusive prerogative of the policymaker to draw policy conclusions based upon whatever facts appear creditable.

For their part, intelligence customers must become as familiar as possible with available intelligence reporting and methods, as well as with the information that is being reported.  No one expects them to know as much as a professional analyst, but policymakers who wholly outsource critical thinking about their global environment are not doing their job. It is hard workto be a good intelligence consumer.

Policymakers who do not somehow make the time regularly to read intelligence reporting—and not just finished conclusions but, wherever possible and on the most important issues, the underlying reporting cables themselves—are not doing their job. Policymakers who cannot to some degree converse with their intelligence briefers about the strengths and weaknesses of various types of collection, or the assumptions that underlie analytical conclusions, are policymakers who are unprepared to evaluate what they are told.

It appears, however, that Washington is notlearning these lessons very well. A mythology has arisen around the Iraq WMD intelligence imbroglio that treats intelligence analysis as the secular equivalent of Revealed Truth: something to which the policy community must give absolute and uncritical deference.  Exacerbated by the eagerness of policymakers, legislators, and intelligence bureaucrats alike to toss declassified versions of analytical products into the hurly-burly of pub lic debate—an eagerness which has done more to make intelligence conclusions into political footballs and distort the policymaking process than any effort by any party actually to skew the conclusions themselves—the conventional wisdom now has it that policymakers must accept intelligence conclusions unquestioningly. To hear some tell it, the Intelligence Community now enjoys a privileged position as the exclusive “finder of fact” in the national security bureaucracy, being now afforded a status somewhat analogous to that of the jury in criminal law proceedings in common law jurisdictions.

This new conventional wisdom is terribly dangerous.  First, it makes the United States unnecessarily vulnerable to further intelligence failures and increases the risk that decisions will be made on the basis of flawed information.  Analysts should receive deference on matters of “fact” to the degree that their knowledge of the available data, experience with a particular subject, and careful reasoning make them persuasive. None of these qualities, however, are monopolized by intelligence professionals. Giving them unthinking deference is to make policy unthinkingly, and that is to ask for trouble. The complicated history of pre-war Iraq intelligence analysis makes clear that an appropriately challenging and probing engagement between analyst and consumer enriches bothanalysis andpolicy decision-making.  It is madness to foreclose having such a relationship, for doing so would increase the risk of error and re-create the conditions of uncritical acceptance that helped make analytical failures on Iraqi WMD so damaging.

Second, uncritical deference corrodes our democratic institutions.  The top intelligence consumer of them all, the President himself, has been empowered to make the most important decisions on earth because he is chosen by a free people to serve in that role.  Those senior officials whom he appoints—and the Senate confirms—to help him make these decisions partake of the democratic legitimacy both of his own election and of the national legislature.  No one, however, elects intelligence analysts; they are accountable only to the extent that their chain of command holds them so, and that this chain of command itself reports meaningfully and accountably to elected political authority at the highest levels.

The U.S. constitutional system has chosen to give judicial authority to a corps of appointed officials—life-tenured federal judges, ensconced in their own coordinate branch of government—who are intentionally shielded from accountability to political authority while in office.  The great ideal of equal justice before the law, and its corollary of judicial independence, underpins this brave (and sometimes controversial) choice, and on the whole it has served our country and our liberties very well. No such ideal, however, supports outsourcing the factual predicates for national security decisions—the raw material of the most important policy choices Presidents are elected to make—to a deliberately unaccountable bureaucracy.  Intelligence analysts cannot be allowed the last word about what the U.S. government thinks. The intelligence bureaucracy must not become, in effect, an opaque and unaccountable fourth branch of our constitutional system.

For all of these reasons, it is important to return to a sense of balance in the relationship between intelligence analysts and policymaking intelligence consumers.  Both parties need to understand the complex parameters of this relationship, finding the right sort of constructive tension at appropriate points while still respecting the special strengths and institutional roles each side brings to the table. This is certainly not always easy, but it is nonetheless essential.

This lesson of Iraq will be an important one for President Obama and his national security team to learn if they wish to govern competently and effectively.  The new president needs to set things on a sound footing from the outset of his Administration by dispelling the pernicious mythology that has grown up in the wake of the Iraq analytical imbroglio.To begin with, he must resist calls to worsen this problem—such as by making the Director of National Intelligence a quasi-independent official analogous to the Chairman of the Federal Reserve—and he should rein in the Executive Branch’s unwise propensity to write and distribute unclassified versions of intelligence analyses. (It is always tempting to try to use such work for leverage in public policy debates, or simply to score political points for being “transparent,” but it is seldom worth the damage that public release—and the presumption that some such disclosure will occur—can cause to the process over the long term.)  Most of all, President Obama and his national security team should insist upon integrity and upon fidelity to the proper roles of intelligence analyst and intelligence consumer at all levels of the national security bureaucracy. The only real cure for these problems is solid and principled leadership; that is what we elect presidents to provide.

--  Christopher Ford


NPF Pages

Recent Additions to NPF

NPF Discussion Pages

Calendar of NPF Postings

January 2009
« Nov   Mar »

NPF Archives (by month)