For a fifteen-year stretch beginning in the late 1980s, American institutions — schools, sports leagues, parenting books, corporate training rooms, the self-help shelves of every Borders and Barnes & Noble — converged on a single idea with the force of a religious conviction. The idea was that self-esteem was a kind of universal solvent. If you could raise it, especially in children, you would lower crime. You would raise grades. You would prevent teenage pregnancy. You would reduce substance abuse. You would lift people out of poverty. You would manufacture happy, productive, civically responsible citizens at scale. The framing was rarely modest. A senior California legislator — about whom more in a moment — described self-esteem as a “social vaccine,” a phrase that implied not merely a useful intervention but a kind of public-health miracle waiting to be deployed.
The policy and cultural footprint was immense. Elementary schools dropped competitive grading and rewrote curricula around affirmation exercises. Youth sports leagues handed out participation trophies on the theory that exclusion from the podium would damage children’s developing self-concept. Corporate training programs ran self-affirmation modules. Parenting bestsellers urged unconditional praise. Books with titles like The Six Pillars of Self-Esteem and The Self-Esteem Workbook sold in volumes that today would qualify as breakout hits in a much larger publishing market. State governments — California first, others following — funded curriculum redesigns. The Office of National Drug Control Policy and various federal agencies cited self-esteem deficits as causal factors in the social problems they were trying to address.
And then, in May 2003, the journal Psychological Science in the Public Interest — a journal whose explicit charter is to commission comprehensive, monograph-length reviews to inform public policy on contested behavioral-science questions — published a 44-page paper by four researchers, three of them among the most cited social psychologists alive: Roy F. Baumeister, Jennifer D. Campbell, Joachim I. Krueger, and Kathleen D. Vohs. The paper was titled “Does High Self-Esteem Cause Better Performance, Interpersonal Success, Happiness, or Healthier Lifestyles?” The authors had been commissioned to read everything — to look at the cumulative body of self-esteem research, which by that point ran to many thousands of studies — and to deliver a verdict. Their verdict, distilled to its core, was this: high self-esteem does not cause much of anything important. It does not cause academic achievement. It does not cause interpersonal success. It does not cause healthier behavior. It does not cause happiness in any robust, reliable way. To the extent that self-esteem correlates with good outcomes, the causal arrow runs mostly the other direction — self-esteem is a consequence of success, not a cause of it. The intervention industry that had been built on the opposite assumption was, on the empirical evidence as it actually stood in 2003, largely without foundation.
This article walks through what California’s task force actually claimed, what Baumeister and colleagues actually reviewed, what the killing findings were, why the causal-direction question matters more than any single effect-size dispute, and — for the strategists, HR leaders, and L&D heads who still routinely encounter “self-esteem” as a foundational lever in corporate-affirmation and leadership-development programs — what the contemporary evidence actually supports as a substitute.
A note before the substance: this is a review of evidence, not character. The architects of the self-esteem movement were not frauds. They were not cynics. They were public servants and earnest researchers, many of them politically progressive, who believed they had identified a humane and scalable solution to a stack of difficult social problems. The critique that follows is not that they were bad people. It is that the empirical foundation under their movement was much thinner than the movement’s tone implied, and that a comprehensive review eventually said so.
The 1986 California Task Force — Vasconcellos, The Hewlett Funding, And The “Social Vaccine” Framing
The institutional origin point of the modern self-esteem movement is a specific piece of California legislation: Assembly Bill 3659, authored by then-assembly member John Vasconcellos (D–San Jose) and signed into law by Governor George Deukmejian in September 1986. AB 3659 created the California Task Force to Promote Self-Esteem and Personal and Social Responsibility — a three-year, twenty-five-member body charged with investigating, in the bill’s own statutory language, the relationship between self-esteem and a list of social problems that included violent crime, drug and alcohol abuse, teenage pregnancy, child and spousal abuse, academic failure, welfare dependency, and recidivism.
Vasconcellos was not a marginal figure. He chaired the Assembly Ways and Means Committee — the budget committee — for eight years. He was one of the most powerful legislators in California, and his interest in self-esteem was deeply personal and longstanding; he had been a serious student of humanistic psychology since the 1960s and had worked with the encounter-group movement at the Esalen Institute. He approached the task force not as a constituent service or a policy experiment but as something closer to a civic mission. The phrase “social vaccine” — the framing that self-esteem deficits were a kind of upstream pathogen that drove a host of downstream social pathologies, and that boosting self-esteem could inoculate against all of them at once — was characteristic of how he and his task force colleagues talked about the project.
The task force was not nothing. It was funded substantially by the William and Flora Hewlett Foundation and other private donors in addition to its state appropriation, and its academic-advisory side produced a serious-looking edited volume: The Social Importance of Self-Esteem, edited by Andrew M. Mecca, Neil J. Smelser, and John Vasconcellos and published by the University of California Press in 1989. The volume commissioned chapter-length reviews from credentialed academics on each of the social-problem domains the task force was meant to address.
Here is where the first wrinkle appears, and it is important to dwell on it for a moment. The 1989 Social Importance of Self-Esteem volume — the supposedly rigorous academic anchor of the task force’s work — was substantially more skeptical about self-esteem’s causal role than the task force’s eventual public-facing report would be. Several of the academic contributors, reading their domains honestly, concluded that the empirical evidence linking self-esteem to outcomes like crime or academic achievement was weak, mixed, or absent. Neil Smelser himself, in his introduction to the volume, noted that the associations between self-esteem and the social problems of interest were generally “mixed, insignificant, or absent.” This is striking. The academic volume said the empirical foundation was shaky. The task force’s policy report would not.
Toward A State Of Esteem (1990) — The Report And The Cascade Of Adoption
In January 1990, the task force published its final report, Toward a State of Esteem. It would become, by some accounts, the best-selling state document in California history, distributed in tens of thousands of copies and quoted approvingly in education-reform circles for the better part of a decade.
The report’s tone was unmistakable. It did not present self-esteem as one factor among many, of contested empirical standing, that might modestly contribute to good outcomes under specific conditions. It presented self-esteem as foundational — as the missing variable whose neglect was driving California’s most serious social pathologies, and whose cultivation would reverse them. The language of “social vaccine” appeared. The list of problems self-esteem would help solve was, again, sweeping: crime, drug abuse, academic failure, teenage pregnancy, welfare dependency, child abuse. The report’s recommendations called for self-esteem to be integrated into school curricula from kindergarten through twelfth grade, into juvenile-justice programs, into welfare programs, into parental-education efforts, and into workplace training.
The cascade of adoption that followed is the part that, twenty-five years later, is the easiest to underestimate from the inside of 2026. State legislatures from Maryland to Missouri established their own self-esteem task forces or commissions, often citing California as a model. School districts adopted self-esteem curricula — Pumsy in Pursuit of Excellence, DUSO (Developing Understanding of Self and Others), Quest, and dozens of other branded packages — that explicitly tried to operationalize the task force’s recommendations. Youth sports leagues, popular parenting authorities, and corporate training vendors operated downstream of the same framing. The participation-trophy joke that came to define the parenting style of the 1990s and 2000s was not a spontaneous cultural drift. It was, in significant part, the visible surface of a policy intervention that had been launched, with serious funding and serious institutional backing, from Sacramento.
By the late 1990s, “raise self-esteem” had become so naturalized as the default goal of education and developmental work that most of the people implementing it could not, if asked, have cited the original evidence base for it. The framing was simply in the water.
What Baumeister, Campbell, Krueger & Vohs 2003 Actually Reviewed
In the late 1990s, the American Psychological Society (now the Association for Psychological Science) launched Psychological Science in the Public Interest (PSPI), a journal whose explicit purpose was to commission comprehensive expert reviews of scientifically contested questions of public-policy importance. The PSPI charter is unusual. Authors are commissioned, not submitted. They are asked to read the entire literature in their domain and produce a long-form review — closer to a federal-agency white paper than to a normal journal article. The reviews are subjected to extensive peer review and are explicitly intended to be cited in policy debates.
In May 2003, PSPI published Baumeister, Campbell, Krueger, and Vohs’s 44-page review, “Does High Self-Esteem Cause Better Performance, Interpersonal Success, Happiness, or Healthier Lifestyles?” (DOI: 10.1111/1529-1006.01431). The authors had been asked, in effect, to deliver the verdict the California Task Force had not delivered: read everything, and tell us what the cumulative evidence on self-esteem actually shows.
The scope of their review was, by the standards of psychology, enormous. They worked through several thousand studies of the relationship between self-esteem and outcomes in major life domains — academic performance, job performance, interpersonal relationships, aggression, antisocial behavior, sexual behavior, alcohol and drug use, eating disorders, depression, and happiness. (The “~15,000 studies” figure that sometimes circulates in popular summaries is an inflation; the authors describe a literature in the thousands and focus their detailed analysis on the strongest available evidence in each domain. The methodological standard they applied was, however, the more important point.)
The methodological move that distinguishes this review from almost everything that had been written about self-esteem before it is the objective-outcomes filter. Baumeister and colleagues were brutally clear that most self-esteem research is contaminated by a specific structural problem: people with high self-esteem reliably rate themselves more favorably than people with low self-esteem on every variable that admits of self-report. They rate themselves as more intelligent, more attractive, more popular, more competent, healthier, happier, more virtuous. A self-report measure of self-esteem correlated with a self-report measure of anything else is therefore close to uninterpretable; high self-esteem can mean “actually doing better” or it can mean “exaggerating in the same direction across every item on every questionnaire.” To get at the causal question, the authors restricted their close analysis to studies that used objective outcome measures — actual grades, actual job performance ratings by supervisors, actual police records, actual health outcomes — rather than self-reports.
That filter changed almost everything. It cut the apparent self-esteem effect on every outcome by a substantial amount, and on several outcomes it eliminated the effect entirely.
The second move was the causal-direction question. Most of the literature was correlational and cross-sectional. Self-esteem and good grades are correlated; does self-esteem cause grades, or do grades cause self-esteem, or does some third variable cause both? The authors focused on the studies — fewer than one would hope — that used longitudinal designs, that measured self-esteem and outcomes at multiple time points, and that could in principle distinguish the direction of the relationship.
The Killing Findings
The 2003 review’s domain-by-domain verdict, fairly summarized, is the following.
Academic performance. Self-esteem and grades are correlated. The size of the correlation is modest. Longitudinal evidence indicates that the causal arrow runs predominantly from grades to self-esteem, not the other way around — students who do well academically subsequently feel better about themselves, rather than students who feel better about themselves subsequently performing better. Programs designed to raise self-esteem in order to raise academic performance — of which a great many had been deployed in American schools by 2003 — were not shown to produce the expected academic gains.
Job performance. Self-esteem correlates weakly with job performance when performance is measured by self-report, and the correlation shrinks substantially when performance is measured objectively or by supervisors. The case for self-esteem as a meaningful predictor of work performance, on the available evidence, is weak.
Interpersonal relationships. People with high self-esteem report more satisfying relationships and rate themselves as more likable. Independent observers do not consistently rate them as more likable. There is no strong evidence that high self-esteem causes better relationships in any objectively measurable sense.
Aggression and antisocial behavior. This is the finding that, more than any other, embarrassed the self-esteem movement. The folk theory underlying the movement held that violent and antisocial behavior was, fundamentally, a compensation for low self-esteem — that bullies, criminals, and abusers were people lashing out from a deep well of unworthiness. The empirical picture is roughly the opposite. Aggression is, if anything, more strongly associated with inflated, threatened, or unstable high self-esteem — with narcissism — than with low self-esteem. The bully is not, on the evidence, the secretly insecure child. The bully is, more often, the child with an unstable high opinion of himself who responds with disproportionate aggression to anything that threatens it. This finding had been documented in Baumeister’s own earlier work (in the 1996 Psychological Review paper “Relation of Threatened Egotism to Violence and Aggression,” with Smart and Boden) and was reinforced by the 2003 review.
Risky sexual behavior, drug and alcohol use. No strong evidence that high self-esteem prevents these behaviors. Mixed and weak associations in both directions, contingent on context.
Happiness. The most robust positive finding in the entire review. People with high self-esteem do report being happier. The authors take this seriously; they note that this is one of the few outcomes for which the evidence is unambiguous and the effect size is meaningful. They also note that this is, in policy terms, the weakest available justification for the self-esteem movement, because the movement had been sold as a way to solve external social problems (crime, drug abuse, academic failure), not as a way to make people feel better about themselves.
Initiative. A second positive finding. High self-esteem is associated with greater willingness to take action, speak up in groups, persist after setbacks, and try things. The authors are again careful not to overclaim — the effect is real but the mechanism is plausible to either causal direction — but this is the other place where they conclude the construct earns its keep.
The cumulative verdict — and the authors are clear about this in the review’s concluding sections — is that the empirical case for self-esteem as a foundational lever for solving external social problems is not there. Self-esteem is correlated with some outcomes, in expected directions, and modestly. It does not, on careful examination, cause much that the social-vaccine framing of 1990 had promised it would cause. The closing paragraphs of the review explicitly call for the field, and policy-makers downstream of the field, to recalibrate.
Why The Causal-Direction Question Matters More Than Any Single Effect Size
The single most important methodological point in the 2003 review is also the one that is hardest to convey in a magazine summary, which is part of why the review’s broader implications took another decade to penetrate the practitioner world.
Imagine two policy interventions, both targeted at students with low grades. Intervention A spends a year working with the students on academic skills — reading comprehension, study habits, math fundamentals, the deliberate practice of difficult material. Intervention B spends a year working with the students on self-esteem — affirmation exercises, journaling about strengths, group sessions about self-worth. Suppose Intervention A produces, on average, a one-letter-grade improvement and an associated rise in measured self-esteem. Suppose Intervention B produces a rise in measured self-esteem and no change in grades.
A cross-sectional correlation between self-esteem and grades, computed across the population of students who experienced either intervention, would show a positive correlation between self-esteem and academic outcomes. A naive reading of that correlation would support the inference that raising self-esteem is a path to raising grades. The longitudinal evidence — which is the evidence that matters — would tell you the opposite: it is improving the academic outcome that raises self-esteem, not raising self-esteem that improves the academic outcome.
This is, in compressed form, what Baumeister and colleagues concluded had happened across most of the self-esteem literature for two decades. Researchers and policy-makers were reading correlations as if they were causal arrows. They were not. The construct that the self-esteem movement had treated as the cause of good outcomes was, in most of the relevant domains, more accurately a consequence of them.
The implication for intervention design is severe. If self-esteem is a consequence rather than a cause, then any intervention that raises self-esteem without raising the underlying competence is decorative at best and counterproductive at worst — it teaches the recipient to feel good about a state of affairs that has not changed, and to expect external validation in proportion to that feeling, and to react with confusion or aggression when the world fails to ratify the feeling. (This is the connective tissue between the 2003 review and the line of work on narcissism that follows.)
What’s Happened To Self-Esteem In Education Since 2003
The honest answer is: a slow and uneven recalibration, not a clean break.
Within academic educational psychology, the post-2003 trajectory has been a gradual shift away from “self-esteem” as a primary target variable and toward constructs with stronger empirical foundations and clearer mechanisms — most prominently growth mindset (Carol Dweck’s framing of the belief that ability is developable through effort), self-efficacy (Albert Bandura’s framing of task-specific competence beliefs), and character education (the family of programs that focus on dispositions like perseverance, self-regulation, and curiosity). These constructs do not all have the same evidential standing — growth mindset’s effect sizes have been substantially revised downward by post-2018 meta-analyses, and grit has had its own course correction — but they have, in general, displaced “self-esteem” as the headline construct in serious education-reform discourse. (Atticusli’s hub piece on growth mindset and grit oversold walks through the corresponding corrections in those literatures.)
Within practice — the schools, sports leagues, parenting books, and corporate training rooms that actually deploy these ideas at scale — the legacy of the self-esteem movement has been considerably stickier. The participation trophy, the everyone-gets-a-certificate elementary school, the parent who has been trained never to deliver a critical word about a child’s performance, the corporate-wellness module that asks employees to list three things they appreciate about themselves: these survived the 2003 review by a wide margin and persist into 2026 as the default. The reason is not that anyone defends the original empirical claims; the reason is that the practices became culturally and emotionally normative independent of the research that had originally justified them.
Within the popular and journalistic conversation, the story took another and somewhat unfair turn: by the late 2000s, the “self-esteem generation” had become a recurring object of scorn, often associated with a separate empirical literature on rising narcissism scores in American college students (Twenge, Konrath, Foster, Campbell, & Bushman, 2008, Journal of Personality, DOI: 10.1111/j.1467-6494.2008.00507.x, and the trade book The Narcissism Epidemic, Twenge & Campbell, 2009). The Twenge work documented a cross-temporal rise in Narcissistic Personality Inventory scores between 1982 and 2006 of roughly a third of a standard deviation, and connected the rise to the broader self-esteem and self-focus turn in American culture. The connection from movement to narcissism is contested in subsequent literature — a 2025 cross-temporal meta-analysis (Oberleiter et al., Journal of Personality) found the rise has plateaued or partially reversed since the late 2000s — but the line of argument that the movement may have produced not robustly happier or more competent young adults but rather more fragile and entitled ones did, fairly or unfairly, become part of how the movement is now remembered.
What Replaced Self-Esteem In Serious Academic Work
For a strategist building or evaluating a leadership-development, L&D, or coaching program in 2026, the practical question is not whether the self-esteem movement was wrong. The practical question is what to use instead. Two construct families dominate the serious academic literature on intervention design, and both have substantially better empirical foundations than “raise self-esteem.”
Self-Efficacy (Bandura, 1997). Albert Bandura’s framing, developed across decades of work and consolidated in his 1997 book Self-Efficacy: The Exercise of Control (W.H. Freeman), is that the operative construct for behavior change is not a global sense of self-worth but a task-specific belief that one is capable of executing the actions required to produce a given outcome. Self-efficacy beliefs are domain-specific: a person can have high self-efficacy for public speaking and low self-efficacy for parallel parking. They are built primarily through four sources, in roughly descending order of strength: mastery experiences (succeeding at a progressively more difficult version of the task), vicarious experiences (watching credible models succeed at it), verbal persuasion (being told one is capable, by a credible source, in a specific context), and physiological states (interpreting one’s own arousal as readiness rather than anxiety).
This framework is the workhorse of contemporary academic-skills interventions, sports-psychology interventions, rehabilitation-medicine interventions, and many enterprise behavior-change programs. It is not new — Bandura had developed most of it by the late 1970s — and it largely predates the self-esteem movement. It survived the 2003 review because it does not make the global cause-of-everything claim that self-esteem made. It makes a narrower, more disciplined, and considerably better-evidenced claim about how task-specific confidence is built and how it then influences task-specific behavior.
Self-Determination Theory (Deci & Ryan). Edward Deci and Richard Ryan’s framework, developed since the 1970s and consolidated across an enormous research program, holds that human motivation is best understood as the satisfaction of three basic psychological needs: autonomy (the experience that one’s actions are self-endorsed rather than externally controlled), competence (the experience of effective interaction with one’s environment), and relatedness (the experience of being connected to others). Interventions designed within this framework target the satisfaction of those three needs in the relevant domain — classroom, workplace, clinic — rather than targeting a global self-esteem score.
The empirical literature behind Self-Determination Theory is large and, on the whole, more robust than the self-esteem literature it has partially displaced. It has informed evidence-based workplace-motivation work, educational-psychology curricula, and clinical-psychology approaches to behavior change in domains ranging from physical-activity adherence to smoking cessation.
A third strand worth mentioning is Jennifer Crocker and Lora Park’s “costly pursuit of self-esteem” framing (Psychological Bulletin, 2004, DOI: 10.1037/0033-2909.130.3.392), which argues that the problem with self-esteem is not so much its level but the pursuit of it — that organizing one’s life around the goal of feeling good about oneself in contingent domains (academic performance, appearance, others’ approval) is itself the source of much of the dysfunction that the movement had tried to solve. The frame is useful for the L&D context specifically, where the pathology being prevented is not low self-esteem but the chronic, anxious, contingent pursuit of validation.
What This Means For HR, L&D, And Coaching Programs Invoking Self-Esteem
For a Chief People Officer, an L&D director, or a coaching practice owner evaluating programs that invoke “self-esteem” as a foundational lever, the 2003 review does not say stop helping people feel good about themselves. It says something narrower and more useful.
It says: shift the target variable. The construct on which contemporary academic evidence is strongest, and which contemporary intervention practice can actually move, is task-specific self-efficacy — not global self-esteem. The mechanism on which contemporary academic evidence is strongest, and which contemporary intervention practice can actually deliver, is mastery experiences delivered through structured practice and structured feedback — not affirmation exercises delivered through journaling and group sharing.
The pragmatic translation, for a program designer:
Audit programs whose theory of change begins with “raise self-esteem” or “build confidence” as a free-standing goal. Ask, for each such program, what the intended downstream behavior change is. If the answer is “perform better in a specific role” or “execute a specific skill” or “persist through a specific kind of difficulty,” the contemporary evidence base supports redirecting the program toward deliberate practice of that skill with structured feedback — Bandura’s mastery experiences plus credible verbal persuasion — and away from affirmation modules.
Be specific about the domain. Self-efficacy is task-specific. A program that produces self-efficacy in the relevant task — leading a difficult conversation, making a sales call, debugging code — will affect behavior in that task. A program that produces global “I am a worthy person” feelings does not reliably affect behavior in any specific task.
Watch the imposter-syndrome framing. A large secondary industry now sells “imposter syndrome” interventions whose theory of change is essentially the self-esteem movement’s, repackaged for adult professionals. The 2003 review’s conclusions apply: the intervention that reliably reduces the felt impostor experience in a specific role is the accumulation of objective mastery experiences in that role, supported by credible feedback. Affirmation modules — “you deserve to be here, you are enough, you are worthy” — feel meaningful in the room and rarely transfer to the underlying competence question they are nominally addressing.
Take the aggression finding seriously. In hiring, promotion, and team-formation decisions, the folk theory that aggressive, dominant, or contemptuous behavior reflects underlying insecurity is not well supported. It more often reflects an inflated and unstable self-evaluation that is responding aggressively to perceived threat. Coaching such individuals on the theory that “underneath, they don’t feel worthy” is, on the available evidence, a misdiagnosis.
Be candid in marketing copy. A program that promises to “build self-esteem” or “build confidence” as foundational benefits is making claims that the comprehensive review does not support. A program that promises to “build task-specific mastery in [the specific competency]” through structured practice and feedback is making claims that the contemporary evidence base does support. The marketing copy is also a forcing function on the program design: programs that cannot honestly claim task-specific mastery outcomes generally do not produce them.
None of this requires throwing out the humane impulse that motivated the original movement. It requires substituting a sharper construct, a better mechanism, and an honest causal model for the one the movement deployed.
Sources
Primary review:
Baumeister, R. F., Campbell, J. D., Krueger, J. I., & Vohs, K. D. (2003). Does high self-esteem cause better performance, interpersonal success, happiness, or healthier lifestyles? Psychological Science in the Public Interest, 4(1), 1–44. DOI: 10.1111/1529-1006.01431. Open-access PDF: https://assets.csom.umn.edu/assets/71496.pdf.
Baumeister, R. F., Smart, L., & Boden, J. M. (1996). Relation of threatened egotism to violence and aggression: The dark side of high self-esteem. Psychological Review, 103(1), 5–33. DOI: 10.1037/0033-295X.103.1.5.
Baumeister, R. F., & Vohs, K. D. (2018). Revisiting our reappraisal of the (surprisingly few) benefits of high self-esteem. Perspectives on Psychological Science, 13(2), 137–140. DOI: 10.1177/1745691617701185.
The California Task Force and its academic anchor:
California Task Force to Promote Self-Esteem and Personal and Social Responsibility. (1990). Toward a state of esteem: The final report of the California Task Force to Promote Self-Esteem and Personal and Social Responsibility. Sacramento, CA: California State Department of Education. ERIC document: ED321170. Archival records: Task Force to Promote Self-esteem and Personal and Social Responsibility Records, 1986–1990, Online Archive of California.
Mecca, A. M., Smelser, N. J., & Vasconcellos, J. (Eds.). (1989). The social importance of self-esteem. Berkeley, CA: University of California Press.
Modern replacement frameworks:
Bandura, A. (1997). Self-efficacy: The exercise of control. New York, NY: W. H. Freeman.
Deci, E. L., & Ryan, R. M. (2000). The “what” and “why” of goal pursuits: Human needs and the self-determination of behavior. Psychological Inquiry, 11(4), 227–268. DOI: 10.1207/S15327965PLI1104_01.
Crocker, J., & Park, L. E. (2004). The costly pursuit of self-esteem. Psychological Bulletin, 130(3), 392–414. DOI: 10.1037/0033-2909.130.3.392. PDF: https://ubwp.buffalo.edu/selfandmotivationlab/wp-content/uploads/sites/91/2018/05/Crocker-Park-2004a-Psych-Bull.pdf.
The narcissism literature that extends the critique:
Twenge, J. M., Konrath, S., Foster, J. D., Campbell, W. K., & Bushman, B. J. (2008). Egos inflating over time: A cross-temporal meta-analysis of the Narcissistic Personality Inventory. Journal of Personality, 76(4), 875–902. DOI: 10.1111/j.1467-6494.2008.00507.x.
Twenge, J. M., & Campbell, W. K. (2009). The narcissism epidemic: Living in the age of entitlement. New York, NY: Free Press.
Oberleiter, S., et al. (2025). A farewell to the narcissism epidemic? A cross-temporal meta-analysis of global NPI scores (1982–2023). Journal of Personality. DOI: 10.1111/jopy.12982.
Contextual journalism:
CalWatchdog. (2010). Retrospective: A state of esteem? Pacific Research Institute. (2010). 20 years later: Self-esteem movement was utopian hucksterism. Zócalo Public Square. California’s Self-Esteem Commission Was Not a Joke.
Related
- Replication Crisis Hub — the full set of behavioral-science claims that did or did not survive serious scrutiny.
- Growth Mindset: From Sensation To Correction — the construct that, alongside grit, has done much of the work of replacing self-esteem in serious education-reform discourse, and which has had its own substantial post-2018 effect-size correction.
- Grit, Oversold — Angela Duckworth’s perseverance-and-passion construct, sold as a foundational predictor of long-run success, walked back by subsequent meta-analyses.
- The Fredrickson-Losada 3:1 Positivity Ratio — the other major positive-psychology export that did not survive comprehensive scrutiny, and that operated in the same corporate-training market the self-esteem movement had built.
- Maslow’s Hierarchy of Needs — the older humanistic-psychology framework whose popular reception followed a similar pattern: weak empirical foundation, enormous practitioner adoption.
FAQ
Are “feel good” interventions useless?
No. The 2003 review concluded that high self-esteem was reliably associated with greater happiness and greater initiative; these are real benefits and they matter. The narrower claim is that “feel good” interventions do not reliably produce the external outcomes — academic achievement, reduced aggression, reduced substance use, better job performance — that the self-esteem movement had been sold on producing. If the program’s honest goal is “help participants feel better about themselves,” it can deliver that, and that is a humane goal. If the program’s goal is to move external behavior, the construct to target is task-specific self-efficacy through mastery experiences, not global self-esteem through affirmation.
What about children or adults who really do have damagingly low self-esteem?
The 2003 review does not deny that low self-esteem is, at the extreme, a real and clinically significant problem — it is a component of depression, of certain anxiety presentations, of some eating disorders. The therapeutic literature on these conditions is its own large field, and treatments like CBT for depression do reliably improve self-evaluations as part of treating the underlying disorder. What the review challenges is the inverse policy claim: that population-wide interventions to raise self-esteem in non-clinical individuals will reduce the rates of social problems the California Task Force targeted. Treating clinical low self-esteem in clinical populations is a legitimate target. Inoculating a general school population against crime by raising their self-esteem scores is the claim that did not survive review.
What about Imposter Syndrome work?
Imposter syndrome — first described by Clance and Imes in 1978 — is a real felt experience, particularly among high-achievers, of being a fraud whose competence is about to be exposed. The 2003 review’s logic applies: the most reliable intervention is the accumulation of objective mastery experiences in the relevant role, supported by credible specific feedback, rather than affirmation exercises about general worthiness. Programs that combine the two — explicit acknowledgment that the felt experience is common and not diagnostic, plus structured practice that delivers actual mastery — are more defensible than programs built entirely on the affirmation side.
What about Bandura’s self-efficacy — isn’t that just self-esteem with a different name?
No. The distinction is the operative one for program design. Self-esteem is a global evaluation of self-worth. Self-efficacy is a task-specific belief that one can execute the actions required to produce a specific outcome. The two correlate, but they are not the same construct, and crucially, the mechanisms that build them are different. Self-esteem interventions historically tried to operate through reflection, affirmation, and reframing of identity. Self-efficacy interventions operate through mastery experiences, vicarious modeling, and specific verbal persuasion. The empirical track record of self-efficacy interventions is substantially stronger, in part because the construct is operationally specific enough to be reliably moved by the things that move it.
Is the “everyone gets a trophy” critique fair?
Partially. The participation-trophy practice is downstream of the self-esteem movement in a recognizable lineage, and the empirical assumption it was built on — that exclusion from recognition damages developing self-concept in ways that harm long-term outcomes — is not well supported by the post-2003 evidence. At the same time, the practice has become a target for cultural commentary that is wider than the empirical critique justifies; many of the things blamed on participation trophies (millennial entitlement, professional softness, etc.) are contested in their own right and have multiple plausible causes. The defensible version of the critique is narrow: the original empirical rationale for the practice did not hold up, and the practice should be evaluated on its own terms rather than on the back of a folk theory of self-esteem that did not survive review.
Has academic psychology fully moved on from self-esteem?
No, but the discourse has shifted. Self-esteem remains a measured variable in many studies, and the construct itself is not regarded as illegitimate. What has shifted is its role: it is now usually treated as one of several outcomes of interest, occasionally as a moderating variable, but rarely as the foundational lever the 1990 California report had positioned it as. The constructs that occupy the foundational-lever position in contemporary academic intervention design — self-efficacy, self-determination, growth mindset (with caveats), character strengths — are substantially better-evidenced.
What was the William and Flora Hewlett Foundation’s role, and does it bear any reputational responsibility?
The Hewlett Foundation provided significant supplementary funding to the California Task Force’s work in the late 1980s, alongside the state appropriation. This was philanthropic support of a state-government policy initiative that was, at the time, broadly viewed as innovative and humane. The 2003 review post-dates the funding by more than a decade. The reasonable historical reading is that the Hewlett Foundation supported a piece of state-government work whose empirical foundations were less solid than its proponents claimed, in a way that no philanthropic funder of the late 1980s could easily have anticipated. The episode is a case study in the difficulty of due-diligencing the evidence base under sweeping social-policy initiatives — a difficulty that has not gone away in the contemporary philanthropic landscape.
Why did the 2003 review take so long to translate into changed practice?
A combination of factors. PSPI reviews are read by academics first and practitioners later. The participation-trophy culture and the school self-esteem curricula had, by 2003, accumulated nearly two decades of inertia, vendor relationships, and emotional resonance with parents and teachers; reversing them required not just an academic verdict but a long process of cultural recalibration. The replacement constructs — particularly growth mindset and grit — also did not fully enter the educational-reform conversation until the late 2000s and early 2010s. And a meaningful share of the self-esteem movement’s downstream practices had, by then, become culturally normative independent of any research base, which made them resistant to research-driven correction. The 2003 review changed the academic discourse quickly; it changed the practitioner discourse slowly; it has changed some downstream cultural practice only partially even now.