Skip to main content

Verified by Psychology Today


Everything You Always Wanted to Know About Narcissism

An expert weighs in on some crucial misunderstandings.

Narcissism, and the social consequences of narcissistic personality and behavior, are one of the most (if not the most) popular and debated topics on Psychology Today. Although a subject of popular interest, narcissism is actually a complex personality trait surrounding many myths, misperceptions, and scientifically unsupported claims that the popular media continues to perpetuate.

How do psychologists actually measure “narcissism"? Where do we draw the line between a healthy amount of self-love and narcissism as a clinical disorder? Is Donald Trump really a narcissist or is he just putting on a show? Why are people drawn to narcissists? And is our society inevitably becoming more "narcissistic?"

My close colleague Seth Rosenthal kindly agreed to join me for an intriguing conservation about the psychology of narcissism. Seth received his Ph.D. in experimental psychopathology from Harvard. His doctoral research focused specifically on the measurement of narcissism in the non-clinical population, and how narcissism differs from normal, healthy self-esteem.

Sander: Before we jump in, let's start at the beginning: What is narcissism? I often notice that people have a tendency to “diagnose” someone as a “narcissist." Yet, there seems to be an important difference between the clinical definition of “Narcissistic Personality Disorder” (NPD) and what social-personality psychologists refer to as “trait narcissism." Could you elaborate on the difference between these two concepts?

Seth: The question of whether narcissism is a disorder or a normal personality trait really involves two separate issues.

The first is the one most people think about when they consider this question—is narcissism a binary phenomenon (a severe disorder that someone either has or doesn’t have) or does it occur on a continuum (an individual can be a little narcissistic, extremely narcissistic, and everything in between)? It doesn’t seem contradictory to think that narcissism occurs to varying degrees among healthy, non-disordered individuals, but that there is also a point at which it becomes so extreme that it causes pervasive psychological impairment and distress to the degree that it would be considered a disorder. However, defining that specific point at which an individual’s narcissism becomes disordered and diagnosable is an ongoing and often contentious challenge. In my experience, most psychologists accept some version of the theory that narcissism occurs on a continuum of severity and can manifest at sub-clinical levels. I think we can best understand narcissism as a personality trait that occurs on a continuum. At the high end of that continuum, it can become disordered—so extreme that it interferes with healthy psychological functioning.

But even if we do mostly agree that narcissism occurs on a continuum of severity, there is a second, equally critical issue—determining which traits do and do not belong within a meaningful definition of that continuum. Narcissism is a complex personality construct that, most scholars believe, is comprised of a number of different dimensions. For example, we all pretty much agree that traits such as grandiosity (believing you are better than almost everyone else), entitlement (believing that you deserve special treatment that others don’t deserve), and acting in an accordingly arrogant manner are all central facets of narcissism. This is whether you’re talking about narcissism as a clinical disorder or personality trait.

But do more normative traits such as confidence and assertiveness also belong under that same narcissism umbrella? Is a highly confident and assertive individual, who isn’t particularly grandiose or entitled, still displaying behavior that we should classify as narcissistic? It’s a crucial question, because most social-personality research on narcissism starts with the premise that high levels of confidence and assertiveness are a part of the narcissism continuum (albeit the lower end of that continuum). And this affects the conclusions people reach in their studies, sometimes making narcissism appear less harmful than I think it really is. It’s a question that I’ve focused much of my research on and has led to some spirited debates. My take on it is that a narrower conception of narcissism is better aligned with most historical conceptions of narcissism. It focuses on core traits like grandiosity, entitlement, arrogance, etc., but does not include normative traits like confidence and assertiveness. It provides more clarity to our research and understanding of narcissism. Most importantly, it more accurately characterizes what differentiates people who are narcissistic from people who aren’t.

Sander: This is important because the media throws around the word "narcissist" quite a bit. For example, the idea that leaders, especially political leaders, are often narcissists, has become a “hot” topic. To illustrate, a recent Psychology Today article reads: “Therapists Confirm Donald Trump’s Narcissistic Personality Disorder." Beyond the fact that usually no form of formal measurement takes place (clinical or otherwise) in such “loose” armchair-assessments, there is some academic literature on the topic of leadership and narcissism. In fact, you have written about it yourself. For example, research has found that groups with no formal leader naturally gravitate towards those who make the most noise, appear to have all the ideas, and frequently show off. Are narcissists really natural-born leaders? Do you agree with the popular observation that political leaders often have (many) traits consistent with a narcissistic personality?

Seth: We know two things pretty clearly from the research on narcissism and leadership: first, narcissists tend to be drawn, most likely by their egotism, to seek out positions of leadership. Second, people gravitate toward potential leaders who publicly demonstrate narcissistic behavior. What’s much more inconclusive is whether narcissism is, on the whole, a positive or negative leadership trait (or, most likely, a mixed bag).

As you point out, I can’t say whether Donald Trump is a narcissist. Leaders and other public figures have a public persona, and only people who know them well can determine whether it’s the same as their true personality when the TV cameras are off. But Trump certainly seems to play a narcissist on TV. In fact, much of his public behavior—the extreme grandiosity, self-promotion, entitlement, self-righteousness, contempt for others, taunting and bullying, etc.—gives the appearance of such textbook narcissism that it almost looks like a caricature of narcissism.

Particularly in times of crisis, people are attracted to a leader who is not afraid to stand up and publicly express bravery, strength, cockiness, and decisiveness. So, for leaders, narcissistic behaviors such as these can signal to followers that they have everything under control, will protect them, and will lead them to a better future. Narcissistic behaviors are attractive to others (at least at first) in other spheres as well, such as romantic relationships. It’s not surprising that behavioral signals of dominance are attractive to others, particularly in the short-term, whether in the domain of leadership, romance or anything else. That seems to be true of some other primate species as well.

Some research suggests that simply having the ambition and drive to become a leader is indicative of narcissism in itself. Believing that you can and should serve in a high position with authority over others is inherently a grandiose belief. I don’t entirely buy this. Ambition is another one of those normative personality traits, like confidence and assertiveness, that might (or might not) be driven by an individual’s underlying narcissism.

We often can’t really tell whether a particular leader is truly narcissistic, or is just using the appearance of these narcissistic traits strategically. But for leaders who are truly narcissistic, the main challenge is controlling their narcissism so it remains an asset and doesn’t get them in trouble. If they begin making key leadership decisions as a means of stroking their own egos, rather than based on rational assessments of the situations at hand, it can be very destructive—both to their followers and institutions, and ultimately, to themselves. It’s mostly with 20-20 hindsight, after watching a leader we already suspected as narcissistic actually self-destruct, that we can most confidently assert that they probably were narcissists and not just hard-nosed strategists. A leader whose seemingly ego-centric decisions land him or her in prison or at the gallows was most likely exhibiting narcissistic leadership along the way.

Sander: This is interesting, especially in light of the fact that the American Psychiatric Association is increasingly doubtful as to whether or not the Narcissistic Personality Disorder (NPD) should still be classified as a “mental disorder." Some argue that personality traits such as superficial charm, overconfidence, and well-developed manipulation skills can be beneficial. In other words, it is difficult to view narcissism solely as a disorder, because of the seemingly functional or “adaptive” traits associated with this type of personality. In fact, narcissism is often referred to as the least “dark” of the so-called “dark triad” (narcissism, psychopathy, and Machiavellianism). Yet I suppose the same could be said of psychopathy (superficial charm, intelligence, social manipulation) but few would call “psychopathy” adaptive. In your opinion, is there an upside to narcissism?

Seth: First, the technical answer: The recent “dark triad” scale that has fueled the research you’re referring to is problematic. It does consistently show that among the dark triad traits, narcissism is the most mild. The problem is that it defines and measures narcissism with items that largely touch on the most mild and normative aspects of narcissism like attention and admiration seeking. But that’s not the case for the items used to measure psychopathy and Machiavellianism, which are more severe and adhere more closely to the core aspects of those constructs. So, of course psychopathy and Machiavellianism come out looking worse than narcissism, just by virtue of how the scale defines and measures narcissism.

But even using a more severe definition and measure of narcissism doesn’t preclude the possibility that there’s an upside to narcissism. Important research has shown that narcissism helps people get what they want, get more power, money, sex. Narcissists can also be more uninhibited and take more risks. They’re more likely to go ahead and do the things that are fun but risky. So being a narcissist, or being around a narcissist, can be exhilarating. It’s also more likely to be destructive, ultimately leading to shaky relationships and poor decisions about things like money, drugs, etc.

My pet (and as yet untested) theory is that narcissists are ultimately more likely to either be on top or on the bottom of society—in the boardroom or the homeless shelter. There’s less middle ground for them than for others. That’s true in their use of splitting when they assess their world—they see the things and people around them as either the best or the worst. It wouldn’t be surprising to find that outcomes for narcissists are similarly split.

Sander: Evaluating the pros and cons of narcissistic traits seems especially prudent given the increasingly popular notion that our society is collectively becoming more “narcissistic." A culture that revolves around individualism and self-achievement has made the “younger” generation increasingly more self-absorbed, the theory goes, from self-promotion on Twitter to the infamous selfie stick. What do you make of the idea that we live in a “culture of narcissism"? If a tendency to “show off” and elevated "self-views” are indeed becoming a new social norm, will everyone soon be diagnosed with narcissistic personality disorder?

Seth: There’s pretty clear evidence that average Narcissistic Personality Inventory (NPI) scores have been rising since the 1980s, particularly among college students. There’s also evidence that it’s the NPI’s less extreme, more normative items that have largely driven this rise. So, it may be that young people are more likely now than in the past to openly express a sense of self-confidence and assertiveness, not necessarily that they’re truly more grandiose, entitled, and exploitative. A bit more “look at me…” isn’t necessarily more narcissistic if it’s not also accompanied by more “…and give me everything I want because I’m superior, so I deserve it.

But even so, bottom line, something has led to these increased narcissism scores. And the simplest explanation is that narcissism has, in fact, been on the rise. But it’s also possible that NPI scores have risen because our cultural norms have changed. Perhaps it’s more culturally acceptable now to make oneself the center of attention and talk about how great you are, and to endorse these things when asked about them on a self-report personality questionnaire. In other words, an alternative explanation is that perhaps young people’s underlying narcissism levels haven’t necessarily risen, but their reluctance to advertise their narcissism, their sense of reticence or even shame over it, has decreased as our cultural norms have shifted.

Sander: A lot of this seems to eventually come down to the issue of how narcissism—as a personality trait—is measured, so I want to get back to the issue of what narcissism is and what it is not. You've argued that the dominant personality scale used to measure narcissism (the Narcissistic Personality Inventory or “NPI”) systematically conflates healthy self-esteem with narcissism. This claim has attracted quite some attention in the field. If the NPI indeed conflates narcissism with self-esteem, what are the larger implications for both past and future research on the narcissistic personality?

Seth: I don’t want to be misunderstood as an all-out critic of the NPI. The NPI largely does a good job of measuring narcissism.

But unfortunately, at the same time, the NPI doesn’t do such an effective job of not measuring self-esteem—the researchers who developed the NPI don’t appear to have put an emphasis on this aspect of the scale’s discriminant validity. What I mean by that is that some of the scale’s items ask people to rate themselves using items that aren’t clear indicators of narcissism, like “I am assertive,” and “I see myself as a good leader,” but are more aligned with self-esteem. Our research has shown that, unfortunately, some of these items do a better job of measuring self-esteem than narcissism. And further, these items have an outsized influence on the scale overall, pulling it away from measuring core aspects of narcissism. So, our recommendation at least is that when you encounter NPI-based research showing that narcissism really isn’t that bad, and might even be a positive thing, you should ask if those findings hold up if you leave out the scale’s most “self-esteemy” items. Our research shows that dropping just a few of these items can lead to very different results, and paints a much more negative, and I think accurate, portrait of narcissism. Removing a small number of items from a well-constructed scale should not affect research conclusions to this degree. So, the contention of our research is that the NPI not only measures narcissism, but also captures non-narcissistic self-esteem (or similar traits like confidence and assertiveness).

Sander: Our conversations and your work on this have certainly convinced me that this is an incredibly important point to make. Let me bring in a quote you gave to Harvard Magazine in an interview on this topic some years ago:

“Narcissism is not a kind of self-esteem. 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.’”

Although I agree that this is a powerful analogy, the difference may seem subtle and perhaps even lost on some people. Could you explain the difference between self-esteem and narcissism in a bit more detail, particularly in terms of when and where we should draw the line between a healthy amount of self-love and outright narcissism?

Source: XiXinXing/Shutterstock

Seth: That question is really the main challenge for researchers and lay people alike. For instance, on my way back from a psychology conference a number of years ago, I was talking to the person next to me on the airport shuttle about what we each do. I told her that I study narcissism and self-esteem. Her thoughtful reaction was along the lines of “So, basically the same things. Or actually, kind of opposites … which one is it?” I told her that was exactly what we had been arguing about at the conference.

Historically, high self-esteem was thought of as feeling positively about oneself to a realistic degree, but not going beyond that to think, unrealistically, that you are better than everyone else, perfect, etc. Some aspects of narcissism, like feelings of grandiosity and superiority, can look similar to high self-esteem, particularly when doing research using self-report scales. In fact, some researchers have argued that grandiosity and superiority should be part of the definition of high self-esteem.

Much of this debate stems from how self-esteem is typically measured in social-personality research. The vast majority of research on self-esteem uses the scale that the sociologist, Morris Rosenberg, developed in the 1960s. He defined self-esteem as feeling realistically “good enough” about oneself. Unfortunately, the self-esteem scale he developed doesn’t do a great job of making the distinction between that and grandiosity. Rosenberg’s definition of self-esteem contrasts with the definition proposed in a highly influential review of self-esteem research by Roy Baumeister and colleagues. For them, self-esteem is exactly what self-esteem scales measure—positive self-assessments with no requirement that they be accurate or justifiable. Under their definition, there’s no amount of positive feeling about oneself that goes “beyond” what they would consider to be high self-esteem. So, narcissistic grandiosity is kind of the pinnacle of high self-esteem, rather than representing something that is different from high self-esteem.

As for the mania/happiness analogy, the point I hoped to make was that just as narcissistically grandiose people may appear to have high self-esteem when they fill out standard self-esteem scales, one could imagine that if you gave a standard happiness scale to someone who was in a manic state; that person would probably seem extremely happy. In other words, if you based your sciences of happiness and of mania around scales that didn’t do a good job of differentiating between the two states, you could eventually become very confused about both. If your happiness measure picked up aspects of mania, happiness would begin to look like it was tinged with surprisingly negative correlates—dangerously impulsive behavior, antagonism, depressive episodes, even hospitalizations. And if your mania scale picked up too much happiness, mania would start to look less severe and destructive than it really is. The key here is that even though happiness and mania might seem similar in some respects, they’re really two qualitatively different psychological states, which produce demonstrably different cognitions, behaviors, and outcomes. My assertion is that self-esteem and narcissistic grandiosity are similarly distinct from each other, so we shouldn’t conflate them in our work, even if our measures can’t distinguish well between them.

Sander: Along with other colleagues, including Harvard psychology professor Jill Hooley, you have developed a narcissism scale called the “Narcissistic Grandiosity Scale” (NGS). Tell us a little about the scale, and how it’s different from other personality measures.

Seth: Most narcissism scales measure the construct broadly, combining all of its key facets (grandiosity, entitlement, exploitativeness, lack of empathy, etc.) into a single composite narcissism score. But there is good reason to believe that different facets of narcissism actually operate somewhat independently of each other in predicting what people feel, think, and do. One aim of our scale was to hone in on grandiosity as one of these key facets so it could be considered independently of others. Subsequent research, particularly by Ryan Brown and colleagues, has borne out the independence of different facets of narcissism. They showed, for instance, that grandiosity and entitlement predict different types of unethical behavior. This type of research, comparing different core aspects of narcissism against each other, shows real promise for helping us gain a more in-depth understanding of narcissism.

But my initial goal when first thinking about the scale was actually to develop a way to help boost our understanding of the differences between narcissism and high self-esteem. As you and I have discussed, the correlation between narcissism scales (particularly the NPI) and self-esteem was puzzling to me at first. It seemed that if there were a relationship between the two, there was a good chance that it wasn’t because narcissists really feel good about themselves in a healthy way, but instead was an artifact of narcissists’ highly inflated self-views. In other words, I hypothesized that narcissists were giving themselves high scores on self-esteem scales because they believed they were “extraordinary,” “perfect,” and “superior,” not because they were experiencing what we might think of as normal high self-esteem (for instance, thinking they are “pretty good”).

So, we developed the Grandiosity Scale and found what we predicted. When you control for grandiosity, the link between narcissism and self-esteem goes away entirely. What this means is that the appearance that narcissists have high self-esteem is driven entirely by their tendency to be grandiose, and not because they had higher levels of what we might consider “true” self-esteem. Bottom line, although narcissists may look like they have high self-esteem on self-report scales, it’s really something different—their grandiosity—that makes them look that way. Without our Grandiosity Scale, or something like it, we wouldn’t have had as clear a way to illuminate the differences between narcissists’ grandiosity and non-narcissists’ high self-esteem.

Seth Rosenthal, Ph.D.
Source: Author

Seth Rosenthal is an expert in psychometric measurement and personality assessment. He has published on narcissism, leadership, intergroup attitudes, and survey methodology. Seth received his Ph.D. in Experimental Psychopathology from Harvard University and B.A. in Psychology from Wesleyan University. His academic and professional positions include a research fellowship at the Center for Public Leadership at the Harvard Kennedy School, research methodology advisor to the Global Agenda Councils of the World Economic Forum, and Director of Opinion Research at the Merriman River Group. He is currently a Research and Methodology specialist with the Yale Program on Climate Change Communication at Yale University. You can follow Seth Rosenthal on Twitter: @MethodsMaven.

Copyright (2016). Sander van der Linden, Ph.D.


Gudrais, E. (2005). Self esteem, real and phony. Harvard Magazine.

Rosenthal, S.A., & Pittinsky, T.L. (2006). Narcissistic leadership. The Leadership Quarterly, 17(6), 617-633.

Rosenthal, S.A., & Hooley, J.M. (2010). Narcissism assessment in social–personality research: Does the association between narcissism and psychological health result from a confound with self-esteem?. Journal of Research in Personality, 44(4), 453-465.

Rosenthal, S.A., Montoya, R.M., Ridings, L.E., Rieck, S.M., & Hooley, J.M. (2011). Further evidence of the Narcissistic Personality Inventory’s validity problems: A meta-analytic investigation—Response to Miller, Maples, and Campbell. Journal of Research in Personality, 45(5), 408-416.

Stippler, M., Moore, S., Rosenthal, S., & Doerffer, T. (2011). Leadership: Approaches, Developments, Trends. Verlag Bertelsmannstiftung.

More from Psychology Today

More from Sander van der Linden Ph.D.

More from Psychology Today