The Truth About Self-Esteem (*)

December 1994

The Truth About Self-Esteem

By Alfie Kohn

The very act of “debating” a controversial issue tends to reduce the number of possible positions to two. This facilitates an adversarial encounter, which is embedded in the traditional notion of a debate.

Consider the question of whether schools should try to help students feel better about themselves. By now this topic has become sufficiently polarized that the vast majority of people who address themselves to it stand in one of two camps: the pro-self-esteemers, mostly educators, who can scarcely believe that anyone would question the importance of trying to improve children’s perceptions of their own worth; and the critics, who dismiss such efforts as ineffective and nonsensical distractions from academics.

With no ax to grind on this matter, I embarked some time ago on a lengthy investigation of it — one that has finally led me to a posture of skepticism about bothpositions and, by extension, about the debate itself. In what follows, I try to show why most discussions of the theory and practice of self- esteem are unsatisfactory. Part I reviews the available research, which contains unwelcome news for anyone who sees self-esteem as a key causal variable. Part II continues the critique of self-esteem by calling into question the values implicit in the concept.

Part III then turns the tables to look at the increasingly strident criticisms of self-esteem that are offered by conservative social critics and others. I argue that the view of education from which such attacks emerge is even more misconceived than the practices being challenged. This analysis leads me, finally, to a discussion of what might be more constructive than the usual arguments for and against self-esteem.


The Measurement of Self-Esteem

There is no getting around the fact that most educators who speak earnestly about the need to boost students’ self-esteem are unfamiliar with the research that has been conducted on this question. At best, they may vaguely assert, as I confess I used to do, that “studies” suggest self-esteem is terribly important. Very few people in the field seem to have any feel for the empirical literature as a whole — what the evidence really says and how meaningful it is.

Even before examining the results of studies cited to show the benefits of high self-esteem — or the problems associated with low self-esteem — it makes sense to ask just what this phrase signifies and how the concept has been measured. Typically, self-esteem is defined in terms of how we evaluate ourselves and our characteristics, the “personal judgment of worthiness that is expressed in the attitudes the individual holds toward himself,” in the words of Stanley Coopersmith, a pioneering researcher in the field.(1)

But what does this mean in practice for people trying to assess differences in self-esteem? How have researchers operationalized the concept? The answer is that they have, in effect, simply asked experimental subjects how favorably they regard themselves. This is done by having them fill out a form in which they must indicate how much they agree with statements such as “I feel that I have a number of good qualities.”

Putting aside for a moment the questions of what statements are included and how they are scored, the point to be emphasized here is that self-esteem ratings are almost always based on what subjects say about themselves, and self- report measures are rather problematic. They may tell us more about how someone wishes to appear than about his or her “true” state (assuming this can ever be known). In fact, some of the most respected researchers in the area have argued that people designated as having high self-esteem are simply those who demonstrate a “willingness to endorse favorable statements about the self” as a result of “an ambitious, aggressive, self-aggrandizing style of presenting themselves.(2)

As if this fact were not disturbing enough, something on the order of 200 instruments for measuring self-esteem are now in use. Many of them haven’t been properly validated (to use a popular self-esteem term in a different way) and are of questionable value. More important, even if every single test was top- notch, there is no reason to think that any two of them are comparable. It’s difficult to generalize about research findings if self-esteem has been measured — and, indeed, conceptualized — differently in the various studies that have been cited.(3)

One result common to almost all measures, though, is that very few people who fill out self-esteem surveys wind up with scores near the bottom of the scale. When a researcher talks about subjects with “low” self-esteem, he or she means this only relative to other subjects; in absolute terms, the responses of these individuals put them somewhere in the middle range of possible scores. In other words, people classified as having low self-esteem are typically not so much down on themselves as simply “neutral in their self-descriptions.”(4) This suggests that it may be necessary to reconsider all those sweeping conclusions about what distinguishes people who love themselves from people who hate themselves. Moreover, the very fact of defining low self-esteem in relative terms means that no intervention can ever make any headway; half the population will, of course, always fall below the median on any scale.

But let us assume for the sake of the argument that we find none of these facts — or any other methodological criticisms that have been offered of the field(5) — particularly troubling. Let us assume that all the self-esteem studies to date, all 10,000 of them, can be taken at face value. Even so, the findings that emerge from this literature are not especially encouraging for those who would like to believe that feeling good about oneself brings about a variety of benefits. (I am ignoring here the vast number of studies that have treated self-esteem as a dependent rather than an independent variable — that is, those that have tried to figure out what causes self-esteem to go up or down rather than investigating whether such fluctuation affects other things.)

While psychologists have been interested for decades in the way individuals think and feel about themselves, the most recent surge of interest in the concept, which has involved the formation of self-esteem councils around the country and the dissemination of classroom curricula, began in 1990 with the much-publicized report of a state-funded task force in California. In addition to issuing its own report, the task force commissioned a group of scholars at the University of California to review the available research; their monographs were published under the title The Social Importance of Self-Esteem. The point, presumably, was to reassure a skeptical public that claims about the value of self-esteem were based not on wishful thinking but on hard data.

There is no question but that many of the researchers who contributed to this volume earnestly hoped to prove exactly that. Thus, it is all the more significant, and perhaps even somewhat poignant, that the data refused to cooperate. In a passage understandably seized on by critics of the self-esteem movement, sociologist Neil Smelser conceded in his introduction to the book that his colleagues had come up virtually empty-handed:

The associations between self-esteem and its expected consequences are mixed, insignificant, or absent. The nonrelationship holds between self-esteem and teenage pregnancy, self-esteem and child abuse, self-esteem and most cases of alcohol and drug abuse. . . . If the association between self-esteem and behavior is so often reported to be weak, even less can be said for the causal relationship between the two.(6)

Over the years, other reviewers have offered fairly similar readings of the available research, pointing to results that are unimpressive or characterized by “massive inconsistencies and contradictions.”(7)

Self-Esteem and Social Behavior

Consider the claim that feeling good about oneself is linked to a variety of constructive life choices — or at least to the absence of destructive behaviors. The metaphor preferred by the members of the California task force is that self-esteem offers a “social vaccine” against crime and violence, substance abuse, and other cultural diseases. Conversely (if we can press the metaphor a bit), low self-esteem is thought to weaken the immunity to these illnesses. It stands to reason that people who don’t value themselves will be neither inclined nor able to value others.

But what do the data say? Some research, to be sure, has turned up an inverse relationship between self-esteem and delinquency or deviance (fighting, stealing, destroying property, and so on). But these studies either have found that relationship to be remarkably slight(8) or else have been designed so poorly as to raise questions about the validity of their findings. For example, one recent survey of teenagers discovered a moderate negative correlation between deviance and self-esteem, but the only teens that had been questioned were upper-middle-class students at a prep school. What’s more, half of those surveyed failed to return the questionnaire, leaving a self-selected and potentially unrepresentative group from which to draw conclusions.(9) That the research would be unconvincing on this point, incidentally, makes perfect sense to many educators who specialize in teaching conflict resolution. Their strong intuition is that violent behavior does not typically result from low self- esteem, nor do people skilled at conflict resolution always seem to have high self-esteem.(10)

And here is a still more disconcerting possibility: even if low self- esteem were initially associated with delinquent behavior, this very behavior might then serve to enhanceself-esteem. Some research has provided qualified support for this conclusion,(11) which, if borne out, would seem to caution against making our primary goal the elevation of self-esteem. If being part of an armed gang can succeed in helping someone to feel better about himself, then feeling better about oneself is not the last word in social vaccines. (The same concerns might be appropriate in response to the phenomenon whereby the “national self-esteem” is bolstered following a military triumph.) On the other hand, a reanalysis of some older data found no support for the idea that delinquency has this effect — but neither did it support the premise that delinquency is caused by an absence of self-esteem in the first place.(12)

What about the positive end of the social spectrum? Are people who feel good about themselves more likely to help, share, and care? Ervin Staub of the University of Massachusetts looked at some of the early research conducted with children and concluded that while “a poor self-concept makes it more difficult to extend the boundaries of the self in benevolent ways,” it also seemed to be true that “an extremely positive self-concept is less related to positive behavior. . .than a. . .more moderate self-concept” is.(13) Nancy Eisenberg of Arizona State University, like Staub a respected researcher of prosocial phenomena, has raised similar cautions about the simplistic assumption that high self-esteem means more helping. Depending on the circumstances, self-doubters may be just as likely as those who are pleased with themselves to come to someone else’s aid, Eisenberg believes, although perhaps for different reasons.(14)

In looking for more recent research relevant to the question, I turned up nine studies (including three dissertations) published since 1980.(15) On prosocial measures that ranged from sharing among children to agreeing to donate one’s organs after death, and from experiencing “feelings of benevolence” to rescuing Jews from the Nazis, the data offer more reason to doubt than to affirm the relevance of self-esteem. Most of the studies found the relationship between how people felt about themselves and how likely they were to reach out to others to be either weak or nonexistent. In other cases, the correlation was positive but dependent on other variables, such as gender or similarity between the helper and the helped. In sum, high self-esteem appears to offer no guarantee of inclining people toward prosocial behavior — or even steering them away from antisocial behavior.

Self-Esteem and Academic Achievement

Altruism aside, the idea that people who regard themselves favorably ought to be able to learn and work more effectively seems not so much a plausible hypothesis as a matter of plain common sense. Someone with high self- esteem would presumably expect to do well, thereby setting into motion an “auspicious circle” — as distinguished from the vicious one that traps individuals who are consumed by self-doubt. Theorists and therapists, to say nothing of teachers and parents, have been saying as much for decades.

Alas, the empirical evidence once again offers meager support for what seems intuitively indisputable. The problem does not stem from the relatively few studies that have found no significant correlation — or even a negative correlation — between self-esteem and academic achievement.(16) These results are exceptional; most research has indeed confirmed a positive association. The problem, rather, is that the qualifications to that association turn out to be more significant than the association itself.

First, as with social behavior, the size of the relationship is simply not very impressive. In a careful review of 128 studies on this topic, two Australian researchers, B. C. Hansford and J. A. Hattie, found that the average correlation was in the range of .21 to .26, which means that differences in self-esteem can account for only about 4 to 7 percent of variation in academic performance, or vice versa.(17)

Second, those same reviewers reported that the correlation was even lower in studies published more recently; in studies conducted with larger, representative national samples of students; and in studies that used standardized indicators of performance (as opposed to even less reliable measures such as grades).(18) The implication is that the better the research, the less significant the connection it will find between self-esteem and achievement.

Third, virtually all researchers seem to agree that, in order to find any meaningful relationship with how well students perform, it is necessary to look not at some “global” measure of self-esteem (how positively they feel about themselves in general) but at something more specific like academic self- esteem(19) or even self-esteem regarding the ability to succeed at a particular subject.(20) After all, “one can hardly expect an individual’s general sense of self (e.g., ‘I am a pretty good person’) to have very much to do with his or her success on a mathematics test.”(21) The concept of self-esteem grows more empirically useful the more narrowly it is conceived. Unfortunately, what most of us have in mind when we use the term — and when we try to devise programs to enhance children’s self-esteem — is the broadest version of the concept, which also appears to be the least valuable.

But take a step back and reflect for a moment on just what all this research really means — or doesn’t mean. Even if good studies did find a significant correlation between global self-esteem and academic performance, this offers absolutely no reason to think that higher self-esteem causes academic performance to go up (or that lower self-esteem causes it to go down). Statistical concepts don’t get much more elementary than this — correlation doesn’t prove causation — but the number of people, including trained researchers, who fail to grasp its relevance here fairly boggles the mind.

Over and over again, one encounters the assertion that (a) self-esteem and academic achievement are positively associated, followed by the conclusion that (b) it therefore makes sense to try to raise students’ self-esteem if we want them to perform better. This is roughly comparable to noticing that absences from school are positively correlated with the appearance of winter coats — and then concluding that warm clothes make children sick.

Obviously, a causal relationship must be demonstrated. Unfortunately, it turns out to be even more difficult to defend the idea that self-esteem produces academic achievement than it is to defend a robust correlation between the two. The late Morris Rosenberg, for many years a leading figure in the self-esteem field, and his colleagues wrote in 1989 that “global self- esteem appears to have little or no effect in enhancing academic performance.”(22) Other researchers have echoed this conclusion: one huge study of high school students found “no significant causal relation,” a review of the literature turned up “overwhelmingly negative evidence. . .for a causal connection,” and an even more recent review concurred that there is “little if any evidence that children’s academic performance is causally determined by their global self-concept.”(23)

If self-esteem doesn’t cause higher achievement, and if there is some kind of connection between the two, what exactly is the nature of the relationship? The answer you get depends on the researchers you ask. Some say that self-esteem and achievement are causally related, all right, but that the latter causes the former more than the other way around. That is, students feel good about themselves because they do well rather than do well because they feel good about themselves.(24) (I will have more to say later about the ideological uses to which this argument is put.)

But are these two possibilities mutually exclusive? It would seem not, and indeed some writers insist that the relationship is reciprocal, with self- esteem and academic achievement each affecting the other. Still others, however, argue that neither is truly an independent variable; something else is really driving both self-esteem and achievement, giving the appearance of an intimate connection between the two (in much the same way that a third variable called winter is both the reason that people dress warmly and the explanation for increased absences from school). One group of researchers at the University of Minnesota has made a strong case that two variables — social class and academic ability — adequately account for most of the variance in self-esteem and in performance. Thus, self-esteem and academic achievement were “related only insofar as they shared the background variables…as common causes.”(25)

In “a somewhat ill-disciplined field,” as the Australian researchers politely put it, “it is possible to find support for all viewpoints”(26) — which is not particularly encouraging for anyone looking for a definitive answer, or even for reassurance that self-esteem matters at all.

Self-Esteem Programs

Again we coast on hypotheticals. Let us assume that self-esteem, despite all the research reviewed here, really is an unequivocal producer of positive attitudes and behavior as well as better academic performance. It remains an entirely separate question whether self-esteem is enhanced by school programs intended for that purpose. What we are looking for here is some indication that such interventions can do more than jiggle somebody’s current frame of mind, that they can make a lasting difference in students’ basic view of themselves.

Hard data to support the efficacy of such interventions are, as best as I can determine, virtually nil. A few unpublished doctoral dissertations have tested children’s self-esteem in the fall, introduced some sort of program intended to raise those levels, and then tested again in the spring, revealing some gains. But the programs may be quite different from what most self-esteem- conscious teachers are actually using in their classrooms, the method for measuring self-esteem may be homegrown and of uncertain validity, and the duration of the effect is in any case unknown.

If classroom interventions haven’t proved successful in raising self- esteem over the long haul, and if self-esteem hasn’t been causally linked to achievement or positive social behavior, then the claim that classroom-based self-esteem programs will bring about these other benefits becomes even more dubious. The little research that addresses this question tends to be “somewhat contradictory and the effects of the manipulation are typically short-lived,” Martin Covington, one of the California scholars, regretfully concluded.(27) On such results does a national movement rest.

It isn’t clear, at least to me, whether these disappointing findings mean that no classroom self-esteem intervention can ever produce lasting results. But even someone who remains determinedly sanguine about the idea itself must concede that a good number of the self-esteem programs currently in use are, to put it kindly, rather silly. Getting students to chant “I’m special!” — or to read a similar perfunctory message on cheerful posters or in prepackaged curricular materials — is pointless at best. Even Robert Reasoner, president of the International Council for Self-Esteem and a long-time champion of self-esteem programs in his days as a school superintendent in California, has remarked that “efforts limited to making students ‘feel good’ are apt to have little lasting effect because they fail to strengthen the internal sources of self-esteem related to integrity, responsibility, and achievement.”(28)

The Meaning of Disappointing Data

It is striking that even those investigators who would like to show that self-esteem is vital to social and academic development and who believe we ought to embark on a major effort to help students feel better about themselves have been largely unsuccessful in their attempts to demonstrate any of this through research. What is the significance of this failure? Consider several possible reactions to the evidence reviewed here.

*  No one has shown that self-esteem doesn’t matter. This is quite true, but it is generally impossible to prove the negative. Moreover, the burden of proof would seem to rest with those arguing that our educational system ought to be attending to a given factor.

*  Self-esteem is related to things other than academic performance and social behavior.This is very likely so. For example, depression and other psychological problems are highly correlated with low self- esteem. Once again, though, there is some disagreement about which causes which.(29)

*  Self-esteem may not be sufficient to produce achievement or to serve as a social vaccine, but it may be a necessary component. True enough. It is entirely possible that children who feel very good about themselves are not necessarily high achievers or caring people — and yet, at the same time, that those who doubt their own worth are even less likely to be so. If high self-esteem failed to guarantee desirable outcomes but low self- esteem actively interfered with them, the overall correlation might be less than impressive (as indeed it is), but self-esteem could nevertheless be considered a relevant and important factor.

*  If the techniques for measurement are so problematic, how can we rely on studies using these measures to challenge the importance of self- esteem? A reasonable question. But if we are going to dismiss the evidence against, we also have to write off the evidence in favor. That leaves us with no substantiation for the idea that self-esteem matters — at least until good data are found to show otherwise. Some researchers are convinced that these data will eventually appear. Neil Smelser admits (in his introduction to the California research anthology) that the available research fails to find much of an association between self-esteem and other variables. Yet he defiantly, or perhaps wistfully, continues to refer to this association as “the link that we all know exists.” He then tries to figure out what methodological flaws have thus far prevented researchers from demonstrating it.(30) By contrast, another writer argues that limits inherent in the experimental method mean tha t traditional research will never produce more encouraging results. But he, too, remains convinced that self-esteem is a meaningful construct and an important reality, regardless of what the research shows.(31)

That, of course, is the central challenge for all of us: how to reconcile the powerful intuition that self-esteem is a significant causal variable with the studies that fail to support this notion. Are we entitled to dismiss inconvenient findings because they somehow fail to document what we strongly believe is true? Or must we — recalling other things that we think we “know” though intuition, personal experience, or clinical impressions, but which turn out to be false(32) — rethink our basic beliefs? What sort of research, and how much of it, should suffice to topple our intuitions? Conversely, what kind of nonexperimental evidence, and how much of it, should we demand before waving away the studies?

These are big, challenging questions that transcend the issue of self- esteem. But a thoughtful consideration of such issues is absolutely necessary when we find ourselves on the horns of an epistemological dilemma. Personally, I see no simple solution to this particularly vexing clash between research and common sense — or, from another perspective, between experimental and nonexperimental evidence. But in the meantime we have to contend with other objections to the notion of raising students’ self-esteem, challenges rooted in basic values rather than in research.


Apart from questions about the level of empirical support for self- esteem’s significance, it is possible to challenge the desirability of focusing on this issue in the classroom. My concern here is not that children are encouraged to feel good about themselves so much as that their attention is trained primarily on themselves.  I’mspecial, I’m important, here’s how I feel about things. The whole enterprise could be said to encourage a self-absorption bordering on narcissism.

Of course, this focus is nothing new in American education — or in American life more generally. If we exhort students to keep their eyes on their own papers; if we strand them in separate desks, as if on their own private islands; if we make the central lesson of school “how to be alone in a crowd,”(33) then this both reflects and contributes to our society’s infatuation with individualism. Any number of commentators have pointed to our absence of commitment to shared values or to the value of what is shared.(34) What matters, we are taught, is individual achievement and self-actualization. Any talk about generosity or caring is prefaced with the assertion that you must love yourself first in order to be able to love others — a belief that is decidedly debatable and also liable to result in people’s failing to get past the first step. One can spend a lifetime trying to love oneself more fully, while other people fade into the background.

The self-esteem movement is hardly unique in fixing each child’s attention on his or her separate self, but it is open to criticism for doing so just the same. Is it by coincidence that the man who helped make “self- esteem” a household phrase back in the 1970s was Nathaniel Branden, who emerged from the inner circle of Ayn Rand’s ultra-individualists? Is it accidental that the rhetoric of self-esteem proponents often continues to recall the human potential movement’s preoccupation with the self?

Lilian G. Katz, a past president of the National Association for the Education of Young Children, has sharply criticized elementary school self- esteem programs for just this reason.(35) She looks at sentence-starter exercises that begin with “I am special because…” and wonders why children’s attention should “so insistently be turned inward.” She notices assignments in which students produce booklets called “All About Me” and wonders why classes are not working on projects called “All About Us.”

The criticism here derives from the value judgment that it is better to foster a sense of community among students than to produce (however unwittingly) a collection of self-absorbed individuals. But Katz goes on to offer an empirical argument as well, pointing out that self-esteem programs may not even succeed on their own terms, because merely being told (or asked to repeat) that one is worthwhile isn’t likely to change anyone’s underlying self-image.

Of course, the latter argument would seem to apply only to questionable versions of self-esteem programs that consist of chanting hollow phrases. But once the issue of individualism has been raised, the implications spill over into a more damning critique — one that calls into question all such programs and the focus on the self itself. That critique suggests that a self-oriented approach may fail to help students believe in themselves because it overlooks the political and economic realities that offer far more meaningful explanations of why some children doubt or even despise themselves. To put it bluntly, if kids come to school hungry or bruised from beatings, no activity designed to promote self-esteem is going to help. These activities are massive exercises in missing the point, not because they have been designed poorly, but because such activities by their very nature are concerned with the psychological states of individuals, which is not where the trouble lies. James Beane said it well:

A debilitating environment is likely to squash fledgling self-confidence no matter how much we exhort the individual to persist. . . .Moreover, suggesting that self-esteem can be preserved by developing ‘coping skills’ endorses the status quo and, in so doing, ignores the fact that having positive self-esteem is almost impossible for many young people, given the deplorable conditions under which they are forced to live by the inequities in our society. . . .[F]ocusing on self-protection mechanisms. . .is, of course, ‘privatizing’ at its extreme. It is a theory of alienation that pits the individual against the world.(36)

That theory, of course, reverberates with political implications, particularly for anyone whose agenda includes social change. As another critic has noticed, even if focusing on individual feelings is “not socially pernicious,” it may, as a practical matter, divert energy and resources “from some socially useful approaches to social problems” that require the transformation of institutions instead of psyches.(37)

Indeed, the self-esteem movement is only the latest version of a familiar American emphasis on positive thinking and self-help — a tradition that has long served the interests of those who benefit from the status quo. Nothing maintains the current arrangement of power more effectively than an approach that ignores the current arrangement of power — and that focuses attention instead on how you feel about yourself.(38)

But if a concern with individual psychology distracts us from structural factors, it bears mentioning that those structures include the organization of our classrooms, not just of our society as a whole. The study of education has surely taken a step forward by recognizing the importance of internal variables (as opposed to mere behavior), but it is possible to become so preoccupied with what is going on inside a child’s head that we lose sight of how such things as self-concept are influenced by how learning is conceived and designed.

A few years ago I spent some time in a California school district that was supposed to be a showcase for “self-esteeming.” Sure enough, each classroom I visited was dedicated to the mission of pumping up every student’s confidence. One teacher exclaimed, “You can do anything, can’t you?”; another wore a button with a diagonal red slash running through the words “I can’t.” Bulletin boards featured such slogans as “You are beautiful!” And classes were required to applaud after each student made a presentation or, in another case, to recite “I am somebody” each morning.

What struck me about these schools, though, was that this relentless cheerleading was grafted on to a profoundly disempowering dynamic. The rules on the walls in these classrooms were devised by the adults and imposed on the children, and they included “When the teacher talks, everyone listens” and “Follow directions.” Students were manipulated constantly with rewards and punishments to enforce compliance with the teachers’ demands: stickers and lavish praise for conformity, punitive “consequences” for disobedience. Only the “Star Spellers” had their work displayed, turning learning into a quest for triumph. Children in the primary grades had to hold their hands over their mouths as they returned from recess in tightly formed lines.

In one sixth-grade classroom, students were being led through a self- esteem exercise in which they were to reflect on their feelings by completing sentences started by the teacher. When I happened to walk in, she was reading, “I feel down and out when…” My attention was drawn at that very moment to a boy sitting miserably apart from the rest of the class — his desk pulled over there by the teacher, I later learned, as a punishment for talking out of turn. The juxtaposition would have been funny if it weren’t so appalling.

The climate and structures I am describing are, with the possible exception of the emphasis on praise,(39) not part of the self-esteem program — nor even, I assume, a direct result of adopting that program. Yet the self- esteem agenda may reflect a kind of myopia such that activities designed to help students attend to their own feelings can co-exist comfortably with just such retrograde educational practices. Whether or not rampant individualism helps to explain the disappointing research results on self-esteem, I believe it reflects a value system that is disturbing in its own right.


If teachers who are concerned about promoting self-esteem looked up from their lesson plans and if researchers put down their experimental protocols, they would find a growing interest in the topic on the part of people who are not professional educators. Over the last few years, quite a few essays, columns, and even comic strips have been devoted to the question of whether schools should help students to feel better about themselves. The tone of this coverage has generally ranged from harshly critical to downright snide.

The social commentators weighing in on the subject — most of them well toward the starboard side of the political spectrum — include Chester Finn in Commentary; the late Christopher Lasch in The New Republic; John Leo in U.S. News & World Report; Charles Krauthammer in Time; a Newsweek staff writer named Jerry Adler in a sneering 1992 cover story; Barbara Lerner in American Educator; Albert Shanker, Mona Charen, and Judith Martin (Miss Manners) in their respective columns; Phyllis Schlafly (to any reporter who asks); any number of lesser-known local pundits; and the creators of “Doonesbury” and “Calvin & Hobbes.”

The extent of overlap among the arguments of these critics is so great that one sometimes suspects they share a single ghostwriter. But their attack on the enterprise of trying to boost children’s self-esteem at school reflects a very different sensibility than the one that has led me to warn of an excess of individualism or a paucity of research. In fact, I believe not only that the critics base their attacks on faulty premises but also that the implications of their attacks are far more troubling than the self-esteem programs themselves.

The conservative case against trying to raise students’ self-esteem relies principally on the creation of a sharp dichotomy in which worrying about how children feel about themselves is contrasted with spending time on academics. The former is depicted as a touchy-feely fad, the latter as old- fashioned honest toil. The former amounts to coddling students by pretending that everything they do is fine, while the latter means facing up to hard truths and insisting that students measure up to tough standards. When teachers join other commentators in parsing the issue this way, they typically do so with a tone of defiant self-congratulation for choosing the latter option.(40)

The obvious response here, beyond questioning the loaded language with which these alternatives are usually described, is to expose the dichotomy itself as false. Concern about students’ social and psychological needs hardly precludes attention to their intellectual development; in fact, the two enterprises may be mutually reinforcing. But this answer does not mollify the critics, many of whom proceed to make the following argument: if students are expected to work hard, if they are graded strictly and rewarded or praised only when they have earned it, they will come to develop a sense of self-respect which is (depending on the critic) either better than self-esteem or a prerequisite for it.

Usually this point is made by emphasizing the harms of the opposite approach: easy A’s, frequent praise, and a general emphasis on unconditional self-esteem (accepting children as they are and encouraging them to accept themselves) will lead to vulgar self-satisfaction. This, in turn, discourages students from making an effort. After all, why work hard to achieve if you are already perfect? Or, to say it a bit differently, if your attention is focused on the value of who you are rather than on what you do, then you probably won’t do very much.

Before responding directly to this position, I want to make some observations about the context in which it is offered. To begin with, the attack on self-esteem must be understood as but one engagement in a much wider war to preserve what might be called the Old School of education. Critics who decry self-esteem programs typically slide into denunciations of any sort of affective education — and, for that matter, any academic instruction that departs in content or method from a “basics” curriculum in which facts and skills in primary subject areas are transmitted from teacher to student. By and large, these are the same critics who dismiss bilingual education, invented spelling, multicultural curricula, and cooperative learning as illegitimate. They reserve special scorn for anyone who challenges such favored traditional practices as tracking, competition, grades, or standardized testing. Sometimes they also speak up for punitive discipline, in which the point is to maintain control of the classroom and get students to obey.

Not every commentary includes all of these elements, of course, but they coincide often enough to warrant thinking of them as forming a sort of ideological package. An attempt to help students feel better about themselves is thus viewed as just one more departure from the way things ought to be. And for rhetorical purposes critics prefer to describe their approach as being under siege — notwithstanding the pervasiveness of the practices they favor. For example, Finn insists with a straight face that the “prevailing wisdom” today is to renounce standardized testing, tracking, and competition,(41) while Lerner refers to the assignment of a dominant role to self-esteem as “the reigning orthodoxy” in education.(42)

It is also worth observing that the preferred method of justifying attacks on self-esteem (and other challenges to the Old School) is the use of highly selective anecdotes: critics are apt to dredge up an example of silly classroom materials or a quotation about feeling good that would redden the cheeks of most self-esteem loyalists. Very rarely are real data provided to substantiate the criticism itself or its premises. (Of course, supporting evidence is also scarce in defenses of self-esteem programs.)

There is one exception to this tendency to ignore the data, however. It has lately become de rigueur for critics to cite, with a note of triumph, a single survey’s finding that American students express more confidence about their knowledge of mathematics than do their counterparts in other countries, despite the fact that their test scores are actually lower. This contrast makes for arresting headlines, such as “Doing Bad and Feeling Good.” But it proves to be a lot less meaningful on careful reflection. First, apart from the dubious validity of standardized test results, some researchers contend that the relative performance of U.S. students is not nearly as poor as is commonly believed.(43) Second, a misplaced optimism about one’s math skills is quite different from high global self-esteem, which is what most classroom programs are intended to bolster. Third, as James Beane has pointed out, it is just as reasonable to summarize the results of that survey by saying that Asian youngsters were reluctant to pronounce themselves pleased with their own mathematical competence. This may say less about Americans’ swelled heads than about the widely recognized Asian aversion to self-commendation.(44)

Beyond the dearth of evidence to buttress their position, critics offer what appears to be a curious conflation of fact and value. They typically insist on the point that self-esteem does not precede achievement but follows it, which suggests that educators should address themselves directly and exclusively to promoting academic performance. This is a plausible hypothesis, and it even enjoys some limited research support. But watch carefully: while the critics may offer a descriptive claim to the effect that trying to boost self-esteem will not work, the core of their argument is really that it should not be done. Lurking just below the surface of their polemics is the fear that somebody is going to get what might be called a “psychological free lunch.” Children may end up being happy with themselves without having earned that right!

Here we have left the world of evidence and entered (through the back door) the realm of moralistic first principles. It is a place of puritanical fervor, where people should not be permitted to eat except by the sweat of their brows and children must not think well of themselves until they can point to some tangible accomplishment. Like all worldviews, this one cannot be proved true or false, but it ought to be brought out into the open so we can decide whether it is reasonable and consistent with our most basic values.

The same could be said of another major premise of the conservative attack on self-esteem programs (and on all other forays into affective education) — namely, that schools ought to be pursuing a narrowly defined academic agenda rather than trying to help children become happier and healthier. This position is typically asserted as if it were self-evident, overlooking vigorous defenses that have been offered in support of a wider mission for our schools. For example, Nel Noddings of Stanford University has gone so far as to challenge “the deadly notion that the schools’ first priority should be intellectual development.” She argues that “the main aim of education should be to produce competent, caring, loving, and lovable people.”(45)

Others would like to see schools vigorously pursuing intellectual and sociomoral development, pointing out that the two are actually difficult to tease apart. (Note that intellectual development is quite different from a collection of discrete academic skills.) To the critics’ complaint that activities intended to help children become good people just take time away from helping them to become good learners, it might be noted that any number of classroom practices can do double duty. When the members of a class meet to make decisions and solve problems, they get the self-esteem-building message that their voices count, they experience a sense of belonging to a community, and they hone their ability to reason and analyze. When students work together in well-designed cooperative learning groups, they are likely to feel more positively disposed toward themselves and one another — and to learn more effectively. When children are helped to develop the capacity (and inclination) to understand how other people see the world, they are more likely to act generously — and, given the focused use of imagination required, they also develop intellectual skills. Finally, conversations about carefully chosen works of literature can spark reflection on important values while they build a facility with and appreciation for the language.(46)

Next let us consider the criticism of self-esteem programs that assumes children who are fundamentally happy with themselves will have no motivation to achieve. This proposition rests on several faulty premises, the first of which is a confusion between positive self-regard and arrogant self-satisfaction. Someone who has a core of faith in his or her own efficacy and an underlying conviction that he or she is a good person is no more likely than other people – – and possibly a good deal less likely — to opt for stagnation. Slippery definitions have made empirical investigations of this point difficult, but if there is little evidence that self-esteem promotes better learning, there is none to show that self-esteem promotes laziness. The assumption that it does may be the product of careless thinking or it may reflect the crypto-religious beliefs that no one deserves to be satisfied and that nothing will get done without the anxious energy of perpetual self-doubt.

Closely aligned with the latter belief is the assumption that “human nature is to do as little as necessary,” as one columnist remarked in the course of ridiculing self-esteem programs.(47) A thorough refutation of this prejudice would require a review of virtually the entire literature on personality theory and motivation.(48) But we can say this much: the desire to do as little as possible is an aberration, a sign that something is amiss. It may suggest that the individual feels threatened and has fallen back on a strategy of damage control, or that extrinsic motivators have undermined interest in the task by reframing it as a tedious prerequisite to obtaining a reward, or that the task itself is perceived as pointless and dull.

This last possibility suggests another hidden premise of the critics: that learning is bitter medicine, an inherently unpleasant process that will naturally be avoided by happy, satisfied children. (Thus, we had better make sure they aren’t too happy or satisfied). In the abstract, this is demonstrably false: it is almost impossible to stophappy, satisfied children from learning. But if learning is taken to mean multiplying naked numbers, or reading the sodden prose of a textbook and answering the even-numbered questions at the end of the chapter, or memorizing disconnected facts and definitions, then yes, there is some truth to the charge. But the problem does not lie with the students or with “human nature” or with high self-esteem; it lies with a drill-and-skill curriculum. Ironically, this is the very pedagogical approach championed by many critics of affective education.

That is why we must be cautious about assenting to the reasonable- sounding thesis that what we need are tougher standards so students can achieve something of which they can be proud. Whose standards are we talking about? Achievement at what? “Real self-esteem is released when a child learns something,” one columnist insists, to which we must immediately add “… if it’s something worth learning.” You will sift through those anti- self-esteem editorials and essays in vain looking for any attention to the differences between basals and real literature, between worksheets and discovery-based projects, between covering the curriculum and discovering ideas, between test-driven achievement and real learning. No “I Believe in Me!” self-esteem unit could possibly be more foolish than the expectation that students will feel good about themselves because they successfully filled in the blanks on a ditto.

The flip side of the fear that too much self-esteem will kill the desire to learn is the faith that the disappointment attendant on losing a contest or flunking a test motivates children to do better in the future. The logic here is that it is precisely the feeling that one is a failure that creates an incentive to improve, a redoubling of one’s efforts to learn. (Premature or unearned self-esteem would therefore short-circuit that natural process.) Usually this dismal premise lurks in the shadows of diatribes about self-esteem, but occasionally it is spelled out explicitly:

Once upon a time. . . you passed or you failed. You made the team or you didn’t. If you fell short, if your ego was bruised by getting a D or by seeing your name on the cut list, then you buckled down and you made it next time and felt good about yourself. . . .Failure can be a terrific motivator.(49)

Again, it is difficult to imagine a point of view more at variance with everything we know about motivation and learning. We may want children to rebound from failure, but this doesn’t mean that they will. While fortifying students’ self-esteem may not guarantee academic success, the experience of screwing up is a poorer bet by orders of magnitude.

When children fail at a task, the most likely result, all things being equal, is that they will expect to do poorly on similar tasks in the future, and this expectation can set in motion a self-fulfilling prophecy. In this highly qualified sense, self-esteem does indeed seem to matter; the chances of success are higher for students who feel competent. On the other hand, perceived competence comes to a large extent from experiencing success. Thus, because failure can engender a feeling of incompetence (if not helplessness), future levels of achievement are compromised.(50)

Beyond its detrimental effect on performance, an experience with failure can produce two other overlapping results: a desire to take the easy way out and a loss of interest in whatever one has been doing. Both of these are presumably troubling to educators across the ideological spectrum. Students who have failed at something are inclined to prefer less challenging tasks than students who have succeeded.(51) (Many commentators on education seem to forget that trying to avoid failure and trying to succeed are two very different orientations.) Similarly, students who have come to feel incompetent are less likely than others to be interested in what they are working on.(52) This means that even those students who really do buckle down and try harder when they fail — the supposed success stories of the Old School methods — may be doing so out of an anxious, compulsive pressure to feel better about themselves rather than because they enjoy learning.(53) Even if they manage to understand what they are reading today, they may not want to read tomorrow.

The destructive impact of failure on intrinsic motivation shows up with startling consistency. Its effect on short-term performance, however, depends on a number of factors. “It is not so much the event of failure, or even its frequency, that disrupts performance as it is the meaning of failure,” as Martin Covington has put it.(54) Did someone deliberately give me too tricky a task in the hope that I would somehow be a better person for failing at it — or did the task unexpectedly prove too challenging? Was failure defined on the basis of someone else’s judgment — or something intrinsic to the task itself? Did I experience that uniquely toxic form of failure in which one person must lose to another in a public competition? Did the failure take place in the context of intense pressure to succeed — or a relaxed climate of exploration? Do I fear ridicule or punishment (such as a poor grade) for having failed — or am I part of a supportive community where setbacks are no big deal?

Once again, the irony creeps up on us: the conditions least likely to make failure a “terrific motivator” (or even to prevent subsequent performance problems) are those such as tough standards, grades, competition, rewards and punishments — in short, the practices associated with Old School, back-to-basics classrooms promoted by many of the same people who deride self-esteem and extol the motivational benefits of failure.

Of course, beyond situational factors, it is also true that individual differences help to account for the effects of failure: one child may redouble her efforts while another throws in the towel. But the most significant of these predictors may just be the extent to which a student is intrinsically or extrinsically motivated — that is, interested in the learning itself as opposed to being inclined to see success as a way of getting grades, approval, or other rewards. Studies indicate that it is children with the latter orientation who “showed performance impairment and less motivation … following failure.”(55) And the latter orientation follows predictably from being in environments that leadstudents to focus on such artificial inducements.(56)

Of all the factors that determine how students will respond to failure, research has shown that the most important one of all is how they have come to explain that failure. If I fall short, do I think it is because I’m stupid (ability), because I didn’t try hard enough (effort), because the questions were too hard (task difficulty), or because it just wasn’t my day (luck)? Some of these things are internal and some are external; some are stable and some are variable; one (effort) is within my control and the rest are not. This framework, proposed a generation ago by Bernard Weiner, has inspired a small library of work by social and educational psychologists. The consensus is that “children who attribute their failures to invariant or uncontrollable factors, such as insufficient ability, tend to be debilitated by failure.”(57)

Interestingly, even children who succeed are less likely to continue learning effectively if they explain their success in terms of how smart they are. Students who are encouraged to focus on their ability — or lack of it — typically become preoccupied with their performance. By contrast, those who explain their success or failure in terms of how hard they tried are more likely to become absorbed in the task itself. This, in turn, means that they are less likely to be thrown by failure and more likely to be intrinsically motivated and to keep working at something until they get it.(58) The key distinction here is not hard to grasp: success, persistence, and interest tend to follow when children are helped to think about what they are doing; the absence of these results suggests that children have been led to think about how they are doing and, by extension, how able they are. But what classroom practices create the latter, dysfunctional orientation?

Generally, the available evidence suggests that a focus on ability is likely to develop when students are provided little choice concerning tasks, competition and social comparison are emphasized, ability grouping and tracking are used, public evaluations of performance and conduct are common, grading is based on relative ability, and cooperation and interaction among students are discouraged.(59)

To reflect on failure and success in the classroom is to return to the place we began — that is, a discussion of self-esteem’s relation to achievement. As we saw, the evidence doesn’t support the claim that programs intended to help children feel good about themselves are likely to raise students’ achievement. But now it seems clear that the Old School approach favored by many critics of these activities is even worse. So how can we help students to learn? If there is reason to be skeptical about what we hear from both boosters and bashers of self-esteem, where does that leave us?

The answer, once again, depends on our objectives. If we are genuinely concerned with students’ intellectual development (as opposed to their scores on standardized tests), then it makes sense to do all we can to help them focus on effort rather than ability, to become absorbed with the learning itself rather than being preoccupied with their performance. This, in turn, can be facilitated by what I have elsewhere called the “three C’s of motivation”: collaboration, choice, and content (of the curriculum).(60)

Collaboration involves more than occasional cooperative learning activities; it means that students feel connected to their peers and that they experience the classroom as a safe, supportive community — not a place of isolation and certainly not a place where they must compete against each other. Choice means that students are brought into the process of making decisions about what (and how and why) they are learning — as well as other issues of classroom life. Finally, to raise the question of content is to challenge the assumption that students are indifferent about their schoolwork because they are not sufficiently “motivated” (or, from another point of view, because they simply have low self-esteem). The real problem may be that the work itself is not meaningful, engaging, or relevant.

Each of these considerations shows up again in slightly different form if our goals for children extend beyond academics to issues of psychological health. Given that many of the programs billed as self-esteem enhancers fail to have any appreciable effect on how children feel about themselves, what does make a difference? Edward Deci and Richard Ryan of the University of Rochester, drawing on the work of other psychologists before them, have proposed that human beings have three fundamental needs: first, to feel autonomous or self-determining – “to experience one’s actions as emanating from the self”; second, to have a sense of oneself as competent and effective; and third, to be related to others and to be part of a social world.(61)

It is not enough to meet these needs only when school is in session. But to the extent we as educators want to help children feel good about themselves, we would do better to treat them with respect than to shower them with praise. We should embrace affective education, but in the context of building community rather than attending to each individual separately. We ought to work with students rather than doing things (Old School things or New Age things) to them. Contrary to what some in the self-esteem movement seem to hold, students do not come to believe they are important, valued, and capable just because they are told that this is so, or made to recite it. On the other hand, they are even less likely to feel that way when they are compelled to follow directions all day. Students acquire a sense of significance from doing significant things, from being active participants in their own education.

In short, it is time we challenged the false dichotomy that has defined the debate about self-esteem. Whether our objective is to help children become good (that is, creative, self-directed, lifelong) learners or good (that is, secure, responsible, caring) people — or both — we can do better than to concentrate our efforts on self-esteem. But let us be careful that in criticizing that approach we do not end up doing even more harm to students in the long run.


1. Stanley Coopersmith, The Antecedents of Self-Esteem (San Francisco: W.H. Freeman, 1967), p. 5.

2. Roy F. Baumeister, Dianne M. Tice, and Debra G. Hutton, “Self- Presentational Motivations and Personality Differences in Self-Esteem,” Journal of Personality, vol. 57, 1989, pp. 556, 554.

3. See, for example, Thomas J. Scheff, Suzanne M. Retzinger, and Michael T. Ryan, “Crime, Violence, and Self-Esteem,” in Andrew M. Mecca et al., eds., The Social Importance of Self-Esteem (Berkeley: University of California Press, 1989), pp. 175-77, which includes a mention of Morris Rosenberg’s estimate of the number of instruments in use as of 1988. Ruth Wylie, a leading authority on the measurement of self-esteem, has made the same point about lack of comparability (personal communication, 1991). For corroboration of the sheer number of measures, see B.C. Hansford and J.A. Hattie, “The Relationship Between Self and Achievement/Performance Measures,” Review of Educational Research, vol. 52, 1982, p. 135.

4. Dianne M. Tice, “The Social Motivations of People With Low Self- Esteem,” in Roy F. Baumeister, ed., Self-EsteemThe Puzzle of Low Self-Regard (New York: Plenum, 1993), p. 41.

5. One more methodological criticism might be mentioned here: the tendency to confuse self-description with self-esteem. Researchers sometimes assume that subjects who rate themselves poorly must be suffering from poor self-esteem. The truth is that the items dreamed up by the tester may not be particularly important to the subject; indeed, given the norms of his or her subculture, a student may even attach a positive value to the inability to do something well. See James A. Beane and Richard P. Lipka, Self-Concept, Self-Esteem, and the Curriculum (New York: Teachers College Press, 1986).

6. Neil J. Smelser, “Self-Esteem and Social Problems,” in Mecca et al., eds., pp. 15, 17. A more recent review of studies investigating claims of a connection between low self-esteem and drug use confirmed that “no sizeable relationship…has been found” and the results “strongly suggest no direct association” at all. See Debra S. Schroeder et al., “Is There a Relationship Between Self-Esteem and Drug Use?”  Journal of Drug Issues, vol. 23, 1993, pp. 659, 658.

7. This phrase comes from Michael R. Jackson, Self-Esteem and Meaning (Albany: SUNY Press, 1984), p. 3. He, in turn, mentions a frequently cited, if possibly dated, review of the literature: L. Edward Wells and Gerald Marwell, Self-EsteemIts Conceptualization and Measurement (Beverly Hills: Sage, 1976).

8. Florence R. Rosenberg and Morris Rosenberg, “Self-Esteem and Delinquency,” Journal of Youth and Adolescence, vol. 7, 1978, pp. 279-91. The correlations never exceeded .19.

9. Xiaoru Liu, Howard B. Kaplan, and Will Risser, “Decomposing the Reciprocal Relationships Between Academic Achievement and General Self- Esteem,” Youth and Society, vol. 24, 1992, pp. 123-48.

10. This is the view of Bill Kreidler, who conducts workshops on conflict resolution for Educators for Social Responsibility, and of Ron Slaby, an expert on school violence prevention at the Education Development Center in Newton, Mass.

11. See Howard B. Kaplan, Deviant Behavior in Defense of Self (New York: Academic Press, 1980); and Morris Rosenberg, Carmi Schooler, and Carrie Schoenbach, “Self-Esteem and Adolescent Problems: Modeling Reciprocal Effects,” American Sociological Review, vol. 54, 1989, pp. 1004-18. However, Rosenberg’s own analysis a decade earlier failed to find any increase in subsequent self-esteem as a result of earlier delinquent behavior. See Rosenberg and Rosenberg, pp. 284-87.

12. L. Edward Wells and Joseph H. Rankin, “Self-Concept as a Mediating Factor in Delinquency,” Social Psychology Quarterly, vol. 46, 1983, pp. 11-22.

13. Ervin Staub, “A Conception of the Determinants and Development of Altruism and Aggression,” in Carolyn Zahn-Waxler et al., eds., Altruism and AggressionBiological and Social Origins (Cambridge: Cambridge University Press, 1986), pp. 142, 145. References omitted from quotation.

14. For example, people with low self-esteem may be motivated to help “as a means of avoiding rejection or obtaining social approval,” she speculates. See Nancy Eisenberg, Altruistic Emotion, Cognition, and Behavior (Hillsdale, NJ: Erlbaum, 1986), p. 203.

15. References are omitted here for reasons of space.

16. Reading achievement was found in one study to be completely unrelated to children’s self-concept. See Jean H. Williams, “The Relationship of Self- Concept and Reading Achievement in First Grade Children,” Journal of Educational Research, vol. 66, 1973, pp. 378-80. A more recent study with Norwegian students found a “negative direct effect of global self-esteem on academic achievement over time” for a sample of sixth graders, but not for third graders. See Einar M. Skaalvik and Knut A. Hagtvet, “Academic Achievement and Self-Concept,” Journal of Personality and Social Psychology, vol. 58, 1990, p. 305. As Martin Covington explains in his contribution to the research reviews for the California task force, “At times, lack of a sense of worth is a more powerful stimulant to achievement than self-confidence is, and, depending on its source, a sense of self-esteem may not always promote the continued will to learn.” See “Self-Esteem and Failure in School,” in Mecca et al., eds., p. 98.

17. Hansford and Hattie, op. cit. Also see Covington, p. 79.

18. Hansford and Hattie, pp. 134-37. On the last point, another researcher argues that grades are more “responsive to motivational influences” than are standardized test scores and therefore might be more likely to show a relationship to self-esteem. See Herbert W. Marsh, “Causal Ordering of Academic Self-Concept and Academic Achievement,” Journal of Educational Psychology, vol. 82, 1990, p. 648.

19. For example, see Hansford and Hattie, p. 133; Mzobanzi M. Mboya, “The Relative Importance of Global Self-Concept and Self-Concept of Academic Ability in Predicting Academic Achievement,” Adolescence, vol. 24, 1989, pp. 39-46; Marsh, p. 646; and Skaalvik and Hagtvet, p. 292.

20. See Barbara M. Byrne, “The General/Academic Self-Concept Nomological Network,” Review of Educational Research, vol. 54, 1984, esp. p. 450; and Herbert W. Marsh, Barbara M. Byrne, and Richard J. Shavelson, “A Multifaceted Academic Self-Concept,” Journal of Educational Psychology, vol. 80, 1988, pp. 366-80.

21. Beane and Lipka, p. 173.

22. Rosenberg, Schooler, and Schoenbach, p. 1014.

23. The high school study: Sheila M. Pottebaum, Timothy Z. Keith, and Stewart W. Ehly, “Is There a Causal Relation Between Self-Concept and Academic Achievement?” Journal of Educational Research, vol. 79, 1986, p. 142. The first literature review: Mary Ann Scheirer and Robert E. Kraut, “Increasing Educational Achievement Via Self Concept Change,” Review of Educational Research, vol. 49, 1979, p. 145. The more recent literature review: Thomas G. Moeller, “What Research Says About Self- Esteem and Academic Performance,” Education Digest, January 1994, p. 37.

24. See, for example, Robert J. Calsyn and David A. Kenny, “Self- concept of Ability and Perceived Evaluation of Others,” Journal of Educational Psychology, vol. 69, 1977, pp. 136-45 — but also see Marsh, p. 647, for an argument that this study suffers from “serious conceptual and methodological limitations.”

25. Geoffrey Maruyama, Rosalyn A. Rubin, and G. Gage Kingsbury, “Self- Esteem and Educational Achievement,” Journal of Personality and Social Psychology, vol. 40, 1981, p. 973. See also Pottebaum, Keith, and Ehly, op. cit.

26. Hansford and Hattie, pp. 123, 124.

27. Covington, p. 80. The same conclusion was offered as early as 1979, and with a hint of satisfaction, by Scheirer and Kraut, op. cit.

28. Robert W. Reasoner, “You Can Bring Hope to Failing Students,” School Administrator, April 1992, p. 24. Likewise, “there is an enormous amount of junk in the self-esteem world” is the assessment of Hanoch McCarty, a member of the National Council for Self-Esteem. See Jonathan Weisman, “Though Still a Target of Attacks, Self-Esteem Movement Advances,” Education Week, 6 March 1991, p. 17. And James Beane, a leading proponent of affective education, similarly insists that “direct instruction in separate courses alone is not enough to adequately address self- perceptions …. In a time when so many messages suggest that self-esteem is tied to what one buys (commodification of the self), it is perhaps not surprising that some schools would want to buy packages to improve self- esteem,” in Affect in the Curriculum (New York: Teachers College Press, 1990), p. 103.

29. See, for example, Susan Harter, “Causes and Consequences of Low Self-Esteem in Children and Adolescents,” in Baumeister, ed., esp. pp. 105- 6. However, Howard Tennen and Glenn Affleck (“The Puzzles of Self- Esteem,” in Baumeister) caution that “although low self-esteem…may anticipate later depressed mood…there is little evidence that level of self-esteem predicts the onset of depressive disorders” (p. 245).

30. Smelser, p. 17. Similarly, he told an interviewer that “while you don’t have anything like scientific proof, there’s no question there is some legitimacy.” See Weisman, p. 15.

31. Jackson, pp. 4-7.

32. See, for example, Alfie Kohn, You Know What They Say…The Truth About Popular Beliefs (New York: HarperCollins, 1990); and Thomas Gilovich, How We Know What Isn’t SoThe Fallibility of Human Reason in Everyday Life (New York: Free Press, 1991). Among the roughly 80 beliefs debunked in the former book are the significance of birth order and the effect of the full moon on human behavior.

33. The phrase is from Philip W. Jackson, Life in Classrooms (1968; rpt. New York: Teachers College Press, 1990), p. 16.

34. For example, see Philip Slater, The Pursuit of Loneliness (Boston: Beacon, 1970); Robert N. Bellah et al., Habits of the Heart (Berkeley: University of California Press, 1985); Thomas C. Heller et al., eds., Reconstructing Individualism (Stanford, CA: Stanford University Press, 1986); Alfie Kohn, The Brighter Side of Human NatureAltruism and Empathy in Everyday Life (New York: Basic, 1990), chap. 9.

35. See Lilian G. Katz, “All About Me,” American Educator, Summer 1993, pp. 18-23. An excerpt from this essay appeared in the New York Times on 15 July 1993. A similar point is made by Wesley R. Burr and Clark Christensen, “Undesirable Side Effects of Enhancing Self-Esteem,” Family Relations, October 1992, pp. 460-64. Katz criticizes self-esteem programs for other reasons, too, such as their failure to encourage creative thinking: personal reactions (“What I liked about…”) tend to displace critical reflection.

36. James A. Beane, “Sorting Out the Self-Esteem Controversy,” Educational Leadership, September 1991, pp. 27, 29.

37. Harry Specht, dean of the school of social welfare at the University of California at Berkeley, is quoted in Weisman, p. 15. For a more detailed analysis along similar lines, see Ellen Herman, “Toward a Politics of Self- Esteem?” Z Magazine, July/August 1991. Herman observes that “prophets of self-esteem are putting their faith in a new psychic order, certain that change will come, if at all, from the inside” (p. 46).

38. For an elaboration of this analysis, see Russell Jacoby, Social AmnesiaA Critique of Contemporary Psychology from Adler to Laing (Boston: Beacon, 1975); and some of the essays in Rollo May et al., Politics and InnocenceA Humanistic Debate (Dallas: Saybrook, 1986). While most of the criticism of the self-esteem movement has come from the political right (see below), it is interesting to note that the original California task force was enthusiastically supported by conservative Republicans. See John Vasconcellos, Preface to Mecca et al., pp. xvi-xvii.

39. For a critical discussion of praise, see Alfie Kohn, Punished by Rewards (Boston: Houghton Mifflin, 1993), chap. 6.

40. For example, see Angie Ward Marsello, “Boosting Student Esteem All the Way — to Failure!” in English Journal, vol. 80, no. 1, 1991, pp. 72-73 — as well as the letters generated by this essay in vol. 81, no. 3, 1992, pp. 74-81.

41. Chester E. Finn, Jr., “Narcissus Goes to School,” Commentary, June 1990, p. 42.

42. Barbara Lerner, “Self-Esteem and Excellence: The Choice and the Paradox,” American Educator, Winter 1985, p. 12.

43. See, for example, Gerald W. Bracey, “Why Can’t They Be Like We Were?,” Phi Delta Kappan, October 1991, pp. 104-17.

44. Beane, “Sorting Out the Self-Esteem Controversy,” p. 28.

45. Nel Noddings, The Challenge to Care in SchoolsAn Alternative Approach to Education (New York: Teachers College Press, 1992), pp. 12, 174.

46. The work of the California-based Child Development Project confirms each of these examples. See Catherine C. Lewis et al., “Stopping the Pendulum: Creating Caring and Challenging Schools,” Phi Delta Kappan, in press.

47. John Powers, “Feeling Good (for Nothing),” Boston Globe Magazine, 24 January 1993, p. 8.

48. Numerous studies have confirmed that children are naturally inclined to try to make sense of the world, to push themselves to do things just beyond their current level. More broadly, the idea that it is natural to do as little as possible is a relic of “tension-reduction” or homeostatic models, which hold that organisms always seek a state of rest. Few models have been so thoroughly repudiated in modern psychology. Interested readers might look up the work of Gordon Allport as well as findings concerned with the fundamental human impetus to attain a sense of competence (Robert White), to be self- determining (Richard deCharms, Edward Deci, and others), to satisfy our curiosity (D. E. Berlyne), or to “actualize” our potential in various ways (Abraham Maslow).

49. Powers, pp. 7-8.

50. One of many studies showing that failing at something leads to an expectation of future failure is Jacquelynne E. Parsons and Diane N. Ruble, “The Development of Achievement-Related Expectancies,” Child Development, vol. 48, 1977, pp. 1075-79. For confirmation of the point that performance is undermined by feelings of incompetence, see Donald S. Hiroto and Martin E. P. Seligman, “Generality of Learned Helplessness in Man,” Journal of Personality and Social Psychology, vol. 31, 1975, pp. 311- 27.

51. See, for example, Allan Wigfield, “Children’s Attributions for Success and Failure,” Journal of Educational Psychology, vol. 80, 1988, p. 78; and Susan Harter, “The Relationship Between Perceived Competence, Affect, and Motivational Orientation Within the Classroom,” in Ann K. Boggiano and Thane S. Pittman, eds., Achievement and MotivationA Social-Developmental Perspective (Cambridge: Cambridge University Press, 1992), p. 84.

52. See, for example, Harter, “Relationship Between Perceived Competence,” p. 83; and Edward L. Deci and Richard M. Ryan, Intrinsic Motivation and Self-Determination in Human Behavior (New York: Plenum, 1985), p. 61.

53. For empirical confirmation of this distinction, see Richard M. Ryan, Richard Koestner, and Edward L. Deci, “Ego-Involved Persistence,” Motivation and Emotion, vol. 15, 1991, pp. 185-205.

54. Martin V. Covington, Making the GradeA Self-Worth Perspective on Motivation and School Reform(Cambridge: Cambridge University Press, 1992), p. 62. The frequency of failure does matter, however. The more one’s efforts have proved fruitless in the past, the more one comes to expect failure in the future. See Deborah J. Stipek and Joel M. Hoffman, “Children’s Achievement-Related Expectancies as a Function of Academic Performance Histories and Sex,” Journal of Educational Psychology, vol. 72, 1980, p. 861.

55. See Cheryl Flink et al., “Children’s Achievement-Related Behaviors,” in Boggiano and Pittman, eds., p. 200, along with other work by Ann Boggiano and her colleagues.

56. See Edward L. Deci, John Nezlek, and Louise Sheinman, “Characteristics of the Rewarder and Intrinsic Motivation of the Rewardee,” Journal of Personality and Social Psychology, vol. 40, 1981, pp. 1-10. This point, of course, raises deeply unsettling questions about assessment, per se, not merely the particular assessment strategies being used. See Alfie Kohn, “Grading: The Issue Is Not How But Why,”Educational Leadership, October 1994, pp. 38-41.

57. Barbara G. Licht and Carol S. Dweck, “Determinants of Academic Achievement,” Developmental Psychology, vol. 20, 1984, p. 628.

58. See almost anything by Carol Dweck (for example, “Motivational Processes Affecting Learning,” American Psychologist, vol. 41, 1986, pp. 1040-48) or Carole Ames (for example, “Classrooms: Goals, Structures, and Student Motivation,” Journal of Educational Psychology, vol. 84, 1992, pp. 261-71). One recent study suggests that focusing on effort may not always promote motivation, but focusing on ability unquestionably undermines it. See Ellen A. Skinner et al., “What It Takes to Do Well in School and Whether I’ve Got It,” Journal of Educational Psychology, vol. 82, 1990, pp. 22-32.

59. Martin L. Maehr and Carol Midgley, “Enhancing Student Motivation: A Schoolwide Approach,” Educational Psychologist, vol. 26, 1991, p. 404.

60. See Kohn, Punished by Rewards, esp. chap. 11. For more on the creation of caring communities in the classroom and the importance of allowing students to make decisions about their learning, see, Alfie Kohn, “Caring Kids: The Role of the Schools,” Phi Delta Kappan, March 1991, pp. 496-506; and idem, “Choices for Children: Why and How to Let Students Decide,” Phi Delta Kappan, September 1993, pp. 8-20.

61. See Edward L. Deci and Richard M. Ryan, “A Motivational Approach To Self: Integration in Personality,” Nebraska Symposium on Motivation, vol. 38, edited by R. Dienstbier (Lincoln, Neb.: University of Nebraska Press, 1990). The quotation appears on p. 243.

Copyright © 1994 by Alfie Kohn. This article may be downloaded, reproduced, and distributed without permission as long as each copy includes this notice along with citation information (i.e., name of the periodical in which it originally appeared, date of publication, and author’s name). Permission must be obtained in order to reprint this article in a published work or in order to offer it for sale in any form. Please write to the address indicated on the Contact Us page. — © Alfie Kohn