Joshua Rovner on Iraq and the Politics of Intelligence

In Fixing the Facts, Joshua Rovner explores the complex interaction between intelligence and policy and shines a spotlight on the problem of politicization. He describes how the Johnson administration dealt with the intelligence community during the Vietnam War; how Presidents Nixon and  Ford politicized estimates on the Soviet Union; and how pressure from the George W. Bush administration contributed to flawed intelligence on Iraq. He also compares the U.S. case with the British experience between 1998 and 2003, and demonstrates that high-profile government inquiries in both countries were fundamentally wrong about what happened before the war.

So just what is “politicization” of intelligence? Some—often those accused of doing the politicizing—tend to wave away its existence or import or even argue “intelligence is politics.” What is it and why is it important to look at?

Politicization is a word we use all the time without defining it. Some observers use it whenever there is overlap between the worlds of intelligence and policy. Others shrug it off because they assume there’s no way to keep those worlds apart.  To some extent this is true: if intelligence agencies are to play any part in the policy process then they must work closely with their policy counterparts.  But calling the normal day-to-day interaction “politicization” isn’t very useful for our understanding of intelligence-policy relations.  One of the things I do in the book is describe routine relations and then explain why politicization is a sharp deviation from the norm.

Politicization happens when leaders make a strong public commitmentthey rely on the PR value of intelligence estimates.  If it turns out that standing intelligence does not actually support their decisions, they may pressure intelligence officials to change it.

I define politicization as the manipulation of intelligence to reflect policy preferences.  Sometimes policymakers pressure intelligence leaders to change their views so they line up with stated policy.  Sometimes intelligence analysts color their findings in ways consistent with their own views. In either case the result is that political bias creeps into estimates.

Politicization is important because it has terrible effects on the quality of intelligence.  In the short term, it causes intelligence analysts to present their findings with unrealistic confidence, even when the underlying information is patchy and unreliable.  Politicization occurs when issues are open to multiple and competing interpretations.  If the answers were obvious there would be no reason to turn to intelligence in the first place.  But those who are interested in using intelligence to win political arguments cannot abide estimates that are hedging or inconclusive.  So they manufacture exaggerated intelligence and pretend that it represents a firm consensus.

In the medium term, politicization inhibits reassessment even after better information emerges.  This happens because intelligence agencies have bureaucratic reasons to avoid reviewing their own work.  Having published firm and unequivocal conclusions, they have no incentive to conduct the kind of reassessment would reveal that their earlier findings were not just wrong but aggressively wrong.  The upshot is that bad analyses linger long after they should be discarded.

In the long term, politicization poisons relationships between intelligence officials and policymakers. Episodes of politicization reinforce negative stereotypes on each side: policymakers view intelligence officials as bureaucratic obstacles, and intelligence officials view policymakers as bullies. Relations can deteriorate so badly that intelligence plays little role in helping policymakers understand events.

You point out that there is no law requiring policymakers to heed intelligence estimates and they have free reign to reject or accept them in whole or in part. That being the case, why would policymakers seek to actually changeintelligence estimates?

Intelligence scholars have offered a number of explanations.  One is that leaders have a psychological need for support when facing difficult choices, so they cajole intelligence to provide analyses that justify their decisions.  Another explanation has to do with organizational design.  When intelligence and policy agencies are co-located then policy preferences will somehow seep into intelligence products, even when policymakers do not deliberately seek to politicize estimates.

The best explanation, however, has to do with domestic politics.  Leaders use intelligence estimates to win over sceptics at home when they make controversial policy decisions.  Intelligence is a powerful advocacy vehicle because intelligence agencies are in the business of secrets, and individuals tend to overrate secret information. Thus when leaders use intelligence in public debates, they claim they have access to unique sources and deserve the benefit of the doubt.

Politicization happens when leaders make a strong public commitment to a controversial new policy.  In these cases they rely on the PR value of intelligence estimates.  If it turns out that standing intelligence does not actually support their decisions, they may pressure intelligence officials to change it. Uncommitted policymakers may remain open to intelligence that contradicts their beliefs, but after going public they are much less tolerant.  They fear that if they accept contrary intelligence and reverse course they will look feckless and weak.  They also fear that their political rivals will embarrass them by leaking estimates that make their decisions look unnecessary.  Under these conditions, they have strong incentives to make sure that intelligence estimates are consistent with their public statements.

Post-war inquiries either ignored policymakers’ role or defined politicization so narrowly that is was unlikely to appearwhen investigators looked at the issue they only asked whether intelligence officials and analysts were the victims of direct pressure.  They did not explore the variety of other ways that policymakers can exert influence.

The U.S. government post-mortems on Iraq—the WMD Commission, the Senate investigation—have roundly laid the blame on the intelligence community for the Iraq WMD intelligence failure. However, as Robert Jervis points out, these reports—by agreement—left out investigation of the role policymakers and politics played in the outcome. Does this mean that only half the story of what led to this intelligence failure is being told?

That’s exactly what it means. Post-war inquiries either ignored policymakers’ role or defined politicization so narrowly that is was unlikely to appear.  As I point out in the book, there are many different ways to politicize intelligence—some crude, some subtle—but when investigators looked at the issue they only asked whether intelligence officials and analysts were the victims of direct pressure.  They did not explore the variety of other ways that policymakers can exert influence.

Those on the other side of the debate argued that intelligence agencies weren’t to blame at all for the Iraq debacle. They suggest that the intelligence community would have gotten it right if policymakers hadn’t twisted their arms.

Neither side is completely right.  A review of the declassified estimates from 1998-2003 shows that intelligence analysts started with plausible but erroneous assumptions about Iraq.  Their views about Saddam Hussein, while understandable given his behaviour in the 1980s, were not consistent with his actions in the 1990s.  Analysts were suspicious about Saddam’s intentions, though they were admirably candid about how little they knew about his unconventional weapons programs.

All this changed in the summer before the war, when the Bush administration came under public pressure to justify the war.  At this point it started leaning on intelligence to exaggerate information on Iraq and to make sure public intelligence supported the president’s claims on Iraq.  Policy pressure caused intelligence officials to transform their suspicions about Saddam into firm conclusions about Iraqi capabilities.

The Iraq story wasn’t simply about intelligence blunders or policy bullying.  What happened was a total collapse in intelligence-policy relations.  Intelligence agencies started with false assumptions, and policy pressure caused them to transform worst-case scenarios into most-likely estimates.

Regarding Iraq, you draw a comparison between the U.S. and UK intelligence communities. Politicization is supposed to be eliminated from U.S. intelligence through physical and systemic “buffers” between policymakers, whereas in the UK system there is mixing between policy and intelligence professionals. Talk a little about this contrast.

Major U.S. intelligence headquarters are geographically and symbolically removed from the Washington policy fray, and they have layers of bureaucratic insulation separating analysts from decision-makers.  This is based on the widely held belief that organizational design is the key to mitigating politicization. No similar distance exists in the United Kingdom, where the intelligence and policy communities are intertwined.

Curiously, however, the difference seems to have had no effect on the pattern of intelligence-policy relations before the war in Iraq.  In both cases politicization occurred at the same time and for the same reason.  The Bush administration and the Blair government both faced opposition to the war in Iraq, and both pressured intelligence officials to exaggerate the case against Saddam in order to overcome domestic sceptics.  If organizational design was a key cause of politicization, we should have seen different outcomes. The fact that we did not suggests that it is not as important as we previously believed.  It also means that there is no organizational fix for the problem of politicization.

Are there any other episodes of the politicization of intelligence that are worth further exploring?

The most important unexplored cases are outside America.  I included the UK case in the book in order to see whether my theory could travel.  I believe it does, but I’m the first to admit a lot more work is needed.  Some exciting new work is emerging on intelligence-policy relations in Israel, Italy, the UK, and elsewhere.  I suspect we’ll know a great deal more about these cases as intelligence archives become available.  I’m also optimistic because of the really smart and ambitious scholars – people like Uri Bar-Joseph, Rose McDermott, Keren Yarhi-Milo, and Matteo Faini—who want to know how intelligence affects strategy and policy, and who are also interested in how it works in different political systems.

Is there any hope of keeping politics out of our intelligence, or is this something we should not really worry about as long as it does not get out of hand?

We are never going to keep politics completely out of intelligence, nor should we.  Effective intelligence officials must be able to maneuver in the political world or they will have little influence on policy.  This is increasingly important as intelligence agencies compete with an array of competitors for attention: traditional media, cable news, think-tanks, and social media.  Finding ways of convincing policymakers that intelligence agencies are still relevant without overselling the quality of intelligence will be extremely difficult.

Dealing with politicization is another matter.  The task here is not keeping politics out of intelligence but making sure that political biases do not determine intelligence conclusions.  Unfortunately there is no simple solution, and the easiest methodsorganizing intelligence to create distance from policymakers, etc.have failed.

The best way to reduce politicization is to bring back the norm of secrecy. The recent move toward transparency led intelligence agencies to publish sanitized versions of estimates on current issues. The goals of this sunshine policy are laudable – democracy requires government openness but the effects on intelligence-policy relations are terrible.  If the public expects the government to declassify intelligence on controversial policies, then policymakers will have very strong incentives to make sure that intelligence agencies stay in line. Perversely, the desire for a more informed public debate actually encourages politicization, which leads to bad intelligence, which leads to a misinformed public.  Restoring secrecy will help put a stop to this vicious cycle.  It will also allow intelligence and policy officials to speak candidly, just as we encourage open discussions through mechanisms like doctor-patient confidentiality and attorney-client privilege.

What are you working on next in this field—or others?

I’m in the early stages of a new book on strategy.  I also have a few articles coming out this year with Caitlin Talmadge on U.S. military power in the Persian Gulf.  Our work uses hegemonic stability theory and the British experience there to make arguments about the appropriate force posture moving forward.  A related piece coming out soon deals with limited war theory and the problems of recognizing victory. A third project I’m working on here at SMU with Tyler Moore involves IR theory and cyberspace, with reference to the issue of whether intelligence agencies should exploit known software vulnerabilities or reveal them to be fixed by vendors.


Joshua Rovner is the John Goodwin Tower Distinguished Chair in International Politics and National Security at Southern Methodist University, where he also serves as Director of Studies at the Tower Center for Political Studies. His book, Fixing the Facts, is winner of the 2011 Edgar S. Furniss Book Award given by the Mershon Center for International Security Studies and the 2012 International Security Studies Book Award from the International Studies Association.


Leave a reply

Your email address will not be published. Required fields are marked *