Self-Esteem, Real and Phony

When Tom Brady joined the New England Patriots as a sixth-round draft pick in 2000, he told the team’s owner, Bob Kraft, “I’m the best decision this organization has ever made.” Not surprisingly, Brady was accused of narcissism. But five years and three Super Bowl championships later, the Patriots’ quarterback has proven himself a superlative team player, “the ultimate in being confident without having to be a show-off,” says Seth Rosenthal, Ph.D. ’05.

Contrast that with Kobe Bryant, the Los Angeles Lakers star whom some have accused of intentionally playing poorly for the first three periods so he can lead a miraculous comeback in the fourth. Lakers coach Phil Jackson even hired a narcissism coach to help deal with Bryant’s famous demands and feuds with other players. Rosenthal’s dissertation research—which, he’s quick to add, did not include the two sports stars among its subjects—presents ways to separate the Bryants from the Bradys of our world.

In disentangling the two concepts, Rosenthal aims to clear self-esteem’s good name. The concept’s popularity soared in the 1980s, but more recently it has taken a beating in the popular media, which confuses it with narcissism. A 2002 New York Times article headlined “The Trouble With Self Esteem,” for instance, described research results indicating that violent criminals tend to be narcissists. “Narcissism is not a kind of self-esteem,” Rosenthal says. “Equating confident people with narcissistic people is like equating happy and manic and then saying, ‘Well, maybe happiness isn’t such a good thing after all.’”

To find out how narcissists really feel about themselves, Rosenthal used a computer test based on the Implicit Association Test codeveloped by Mahzarin Banaji, Cabot professor of social ethics in the department of psychology and Pforzheimer professor at Radcliffe (see “Stealthy Attitudes,” July-August 2002, page 18). First, the test asks subjects to associate themselves with positive qualities and dissociate themselves from bad qualities, and measures how quickly the subjects press computer keys to indicate those associations. Then the instructions are reversed (bad qualities are to be marked “me” and good qualities “not me”) and response time is measured again. Those subjects quicker to assign positive qualities to themselves have high self-esteem; those quicker to assign themselves negative qualities have low self-esteem. Despite the sky-high self-regard narcissists project, Rosenthal found their implicit self-esteem was low.

Thus, he suggests, parents needn’t worry that instilling healthy self-confidence in their children will turn them into creeps. In fact, psychologists believe narcissism—a diagnosable personality disorder whose symptoms include an exaggerated sense of self-worth, feelings of superiority over other people, a sense of entitlement, a willingness to exploit others, and a lack of empathy—is associated with parents who are cold and rejecting, not overly attentive and doting. Because narcissists didn’t get enough positive reinforcement as children, they need to prop up their egos with their own boasting and others’ adoration. But because that inflated self-image lacks internal support, it is fragile, like a reflection that dissolves with the pond’s disturbance, as in the Greek myth that gave narcissism its name. Underneath lies self-loathing.

The confusion, Rosenthal says, arises because psychologists have measured self-esteem poorly. The widely used 1965 Rosenberg self-esteem scale, for example, includes statements like, “On the whole, I am satisfied with myself,” and yields similarly high results for both narcissists and people with healthy self-confidence. (We all rank ourselves relatively high; on Rosenberg’s scale of 10 to 40, the mean is 32, well above the midpoint of 25.) The more elaborate 1979 Narcissistic Personality Inventory had statements like, “I find it easy to manipulate people,” but still embraced items that Rosenthal says actually test for healthy self-esteem—for instance, “I am assertive” and “I see myself as a good leader.”

With his adviser, psychology professor Jill Hooley, Rosenthal developed his own questionnaire to measure grandiosity, the tendency toward aggrandized self-description that’s one component of narcissism. Rosenthal tested 100 traits on Harvard undergraduates—who, perhaps surprisingly, are no more narcissistic than the general population—and winnowed out 16 items that narcissists most frequently ranked high, but non-narcissists ranked low. These self-descriptors include perfect, superior, envied, unrivaled, and omnipotent. Using traditional measures that conflate self-esteem with arrogance, Rosenthal found that narcissists scored high—above 33—while non-narcissists fell below 31. (Because respondents cluster at the top of the scale, a point or two represents a major change.) Arrogance seems to make the difference, because when it’s edited out, narcissists’ average score dropped under the non-narcissists’—just over 31, versus nearly 32.

Rosenthal’s findings may even have political implications. During a postdoctoral research fellowship at the Kennedy School of Government, he has found that members of informal groups with no designated leader are likely to name the narcissist among them as a leader—but only at first. “Those are the people who show off the most, make the most noise, have the most ideas, or at least are most likely to keep throwing them out there,” Rosenthal says. “It does seem that, early on, people are attracted to the chest-thumping and showing off. But as the narcissism starts to show through, people start to get sick of that person.”              

Seth Rosenthal e-mail address: seth@wjh.harvard.edu