Post is at it again, this time with Foreign Policy’s help. The outlet ran a commentary from him on March 15 that assessed Libya’s “mercurial leader” and argued “Qaddafi is indeed prepared to go down in flames.” About a year before that, in May 2010, it ran a piece headlined “Profiles in Phobia—The irrational fears that keep the world’s most powerful leaders up at night,” which also looked at Qaddafi, among four others. Unfortunately, neither of these pieces exhibited the context or skepticism of Omestad’s article from 1994, which laid bare numerous problems related to profiling. It even quoted our current secretary of defense, then recently retired from the CIA, reporting that:

Outside a small circle of practitioners, the methods of profiling are dimly understood. Even former and current officials who praise particular profiles tend to be deeply skeptical about the field. Typical is the comment of Robert Gates, a former CIA director and deputy national security adviser: “Trying to diagnose somebody for 5,000 mile away who you’ve never seen does not fill me with confidence.”

Later, the article expands on this criticism, pointing out that, “The pitfalls of psychological inference aside, the CIA profiling suffers from a more prosaic, but serious problem: factual errors … The CIA would stumble over such basic facts as where a president or minister was educated, whether he was married, and how many children he had.” Omestad actually deems these errors to be profiling’s “biggest flaw,” and concludes with some suggestions about how the technique could be improved:

The CIA needs to renews a rigorous effort to check facts, principally through open sources. More consistent and intensive review of profiles by regional specialists and psychologists outside the agency - and when possible by people who have known the subjects - would root out some mistakes and help deter the CIA from hewing to a rigid institutional view of certain leaders…

The profilers should take care to distinguish conjecture from fact—a kind of truth-in-labeling practice. Those warning labels might remind readers that a profile is but one analytical tool among others. They might also caution that explaining a leader in purely rational terms inevitably overlooks the unknowable influences of irrationality, chance, and even serendipity on how that leader will really act. Policymakers, for their part, should stop pushing the CIA to make bold predictions about foreign leaders when the necessary data is simply unavailable.

It’s spot-on advice for journalists as well, and reporters should look up Omestad’s article before embarking on any stories about profiling or the mental state of dictators. They should also bear in mind the “Goldwater rule” when attempting to cover the antics of alleged madmen—whether they be murderers, celebrities, or anybody that has not undergone a formal psychiatric evaluation.

If you'd like to get email from CJR writers and editors, add your email address to our newsletter roll and we'll be in touch.

Curtis Brainard is the editor of The Observatory, CJR's online critique of science and environment reporting. Follow him on Twitter @cbrainard.