The Limits of Intelligence

It is conventional wisdom that whether because of President George W. Bush’s aversion to complexity, Vice President Dick Cheney’s obsession with Saddam Hussein, or something else entirely, somehow Washington simply ignored the U.S. intelligence community’s doubts that Saddam was collaborating with Al Qaeda and that a stable Iraq could emerge after an invasion. But the Bush administration’s mistakes in Iraq are only the most recent illustration of the challenges policymakers and intelligence analysts face when attempting to communicate—challenges that presidents of every political stripe encounter as they struggle to lead with confidence in an ambiguous world, writes Robert Jervis, a professor of international politics at Columbia University.

Presidents don't usually want to hear an intelligence analyst's doubts, preferring confidence (even when unwarranted) in one policy option.

For both political and psychological reasons, presidents often expect the intelligence community to be able to provide them with clear answers. Politically, presidents need intelligence backing to sell their policies to the public. Psychologically, they need to sleep at night, and that requires seeing a world in which one policy is clearly preferable to another and its costs, often measured in lives, are less than those of any alternative. Even when the news is good, it may not be greeted favorably. This was the case when the Central Intelligence Agency told Lyndon B. Johnson that other countries would not fall to communism even if South Vietnam did. Since Johnson’s Vietnam policy was based on the domino theory, he did not welcome the information.

Presidents don’t usually want to hear an intelligence analyst’s doubts. Policymakers will try to convince both themselves and the public that one policy measure is better than an alternative on every dimension, even when, as Jervis writes, “there [is] no reason to expect the world to be arranged so neatly.” The confidence (even when unwarranted) that comes from believing one policy option is clearly superior is not simply for a president’s personal benefit but also necessary to a successful policy. If a leader is plagued with doubts, the uncertainty can filter down to the rank and file and doom a policy before it is even launched.

When uncertainty exists, intelligence analysts can be susceptible to pressures from policymakers to change their conclusions. But the charge of “politicization” is too easily lobbed about, Jervis argues. How can you tell the difference between a politician making sure that due diligence has been done and one simply demanding a different answer? “In many of these cases, I suspect that one’s judgment will depend on which side of the substantive debate one is on,” he remarks.

The president’s need to have the backing of the intelligence community in order to sell his policies stems from the public’s faith in the quality of the intelligence community’s judgments. But when the president presses intelligence analysts to support his policies, the quality of the information is likely to suffer. And even in the absence of political pressure, reliable intelligence is difficult to come by. When the United States failed to anticipate the 1974 coup in Portugal, then secretary of state Henry Kissinger resented congressional complaints about intelligence failure: “Anytime there’s a coup you start with the assumption that the home government missed it. . . . Why the hell should we know better than the government that’s being overthrown?"

THE SOURCE: “Why Intelligence and Policymakers Clash” by Robert Jervis, in Political Science Quarterly, Summer 2010.

Photo courtesy of The U.S. National Archives