In the autumn of 1964, educational psychologists Ladd Johnson, Richard Feigenbaum, and Marcia Weiby ran a deceptively simple experiment at a university in the American Midwest. They recruited a group of student teachers and had each one work through a series of arithmetic problems with a child. The teachers were told their student had scored poorly on a pre-test. After the teaching session, they received feedback — fabricated by the researchers — about whether the child's score had improved. Teachers who were told the child improved claimed credit directly: they had explained clearly, adjusted their approach, given good examples. Teachers who were told the child had failed attributed the outcome to the child: lack of effort, low aptitude, poor attitude. The teaching session was identical in both cases. Only the feedback differed. The attribution shifted completely depending on which way the outcome had fallen.
The year was 1964. The journal was Psychonomic Science. The finding was one of the earliest controlled demonstrations of what psychologists would come to call the self-serving bias: the systematic tendency of human beings to attribute positive outcomes to their own abilities, character, and efforts, and to attribute negative outcomes to external forces, bad luck, other people, or circumstances beyond their control.
The phenomenon has since been documented across hundreds of studies spanning seven decades. It operates in classrooms, boardrooms, cockpits, hospitals, divorce proceedings, post-mortems, and annual performance reviews. It is one of the most consistently replicated findings in social psychology, and one of the most consequential for how human beings relate to evidence, responsibility, and each other.
"People attribute their successes to their own abilities while attributing their failures to external factors — protecting self-esteem at the cost of accurate self-knowledge." — Dale Miller & Michael Ross, 1975
What the Self-Serving Bias Is
The self-serving bias is the tendency to attribute one's successes to internal, stable, personal factors (ability, effort, character) while attributing one's failures to external, unstable, situational factors (bad luck, other people's behavior, unfair systems).
Self-Serving Bias vs. Depressive Realism
The clearest way to understand what the self-serving bias actually is — and what it is not — is to place it against its empirical counterpart: depressive realism. The term was introduced by psychologists Lyn Abramson, Martin Seligman, and John Teasdale in a landmark 1978 paper in Psychological Review titled "Learned Helplessness in Humans: Critique and Reformulation." Abramson, Seligman, and Teasdale proposed that depressed individuals characteristically attribute negative events to internal, stable, global causes ("I failed because I am incompetent, and I am always incompetent at everything"). Later research by Lauren Alloy and Lyn Abramson, published in the Journal of Experimental Psychology: General in 1979, produced a striking finding: mildly depressed participants were more accurate than non-depressed participants in judging how much control they actually had over an outcome. Non-depressed individuals consistently overestimated their control. Depressed individuals, within certain parameters, were right.
This is the phenomenon the following table addresses.
| Dimension | Self-Serving Bias (Typical) | Depressive Realism (Mildly Depressed) |
|---|---|---|
| Attribution of success | Internal: ability, effort, skill | External: luck, task ease, others' help |
| Attribution of failure | External: bad luck, others, situation | Internal: own incompetence or character |
| Accuracy of self-assessment | Systematically inflated | Closer to objective reality in some domains |
| Judgment of personal control | Overestimated | More calibrated |
| Emotional consequence | Higher subjective well-being; lower accuracy | Lower well-being; higher accuracy in specific tasks |
| Adaptive value | Protects self-esteem; may sustain motivation | Provides accurate feedback; may inhibit action |
| Key researchers | Zuckerman (1979); Taylor & Brown (1988) | Alloy & Abramson (1979); Abramson et al. (1978) |
The picture that emerges from this comparison is uncomfortable: the psychologically normal condition involves a degree of motivated distortion in how people interpret their own performance. The clinically abnormal condition, mild depression, brings people closer to accuracy. This tension has never been fully resolved in the literature, though Shelley Taylor and Jonathon Brown's 1988 paper in Psychological Bulletin, "Illusion and Well-Being: A Social Psychological Perspective on Mental Health," argued forcefully that positive illusions — including self-serving attributions — are not pathological. They may, in fact, be functionally adaptive.
Cognitive Science: The Mechanism Behind the Bias
Why the Brain Defaults to Self-Credit
Mervin Zuckerman's 1979 meta-analysis, published in the Journal of Personality, synthesized findings from dozens of attribution studies and formally established what he called the "self-serving attributional pattern" as a robust, replicable phenomenon. Zuckerman reviewed data from studies across multiple domains and found that the asymmetry — taking credit for success, deflecting blame for failure — held across cultures, age groups, and experimental designs. It was not a product of any particular laboratory, experimenter effect, or population.
The cognitive mechanism has since been investigated along several lines.
One account emphasizes motivational distortion: people are motivated to see themselves positively, and this motivation biases how they process, weight, and recall causal information. When an outcome is good, attention gravitates toward one's own contribution. When an outcome is bad, attention gravitates toward external factors. The motivation comes first; the attribution follows. This account is supported by research showing that the bias intensifies when self-esteem is threatened. Bernard Weiner, whose attribution theory work in the 1970s and 1980s at UCLA became foundational for the field, identified self-esteem maintenance as one of the primary drivers of attributional asymmetry.
A second account emphasizes cognitive availability: people simply have more information about their own effort and intention than about the causal contributions of external factors. If you worked hard on a project and it succeeded, you can directly perceive and recall that effort. If the project failed, you can also perceive the difficulty of the circumstances. This differential availability of information can produce self-serving attributions even without any explicit ego-protective motivation. Psychologists Thomas Gilovich and Dale Griffin, writing in the Psychological Review in 2002, reviewed evidence for information-based accounts of attribution and found that availability effects and motivational effects are not mutually exclusive — they reinforce each other.
A third, neurobiological account was opened by studies using functional MRI. Research by Selin Kesebir and colleagues, and by groups working in the tradition of affective neuroscience, has shown that self-referential processing activates medial prefrontal cortex regions associated with reward and positive valence. The brain's processing of self-relevant success information recruits reward circuitry in a way that self-relevant failure information does not, creating a structural asymmetry in how vividly the two types of events are encoded, consolidated, and later retrieved.
The Role of Locus of Control
Julian Rotter's concept of locus of control, introduced in 1954 and published in full in 1966 in Psychological Monographs, provided the foundational construct that later self-serving bias research would repeatedly draw upon. Rotter distinguished between individuals with an internal locus of control (events are contingent on one's own behavior) and those with an external locus of control (events are contingent on luck, powerful others, or fate). Self-serving bias researchers found that even individuals with a generally external locus of control showed the standard self-serving pattern: when outcomes happened to go well, they claimed internal credit.
The Illusion of Control
Ellen Langer and Jane Roth's 1975 experiment, published in the Journal of Personality and Social Psychology, demonstrated a crucial related phenomenon. Participants were asked to predict the outcome of a series of coin flips. Some participants had an early run of apparent successes (in reality, completely random). After the sequence was complete, participants who had more early successes rated themselves as better at predicting coin flips than participants who had fewer early successes, despite the fact that coin flip outcomes are mathematically independent of any prior sequence. People who had been lucky at the outset attributed their good fortune to personal skill. The more success they experienced — even randomly — the more they believed they possessed a genuine predictive ability.
This is the illusion of control: the tendency to believe one exerts agency over outcomes that are, by definition, outside one's control. It is a first cousin of the self-serving bias. Where the self-serving bias operates retrospectively — explaining past outcomes in ego-protective ways — the illusion of control operates prospectively, inflating anticipated efficacy before outcomes are known.
Four Named Case Studies
Case Study 1: The Medical Malpractice Attribution (Clinical Medicine)
In a 2003 analysis by physicians Brian Fischhella and colleagues examining incident reports from a large teaching hospital, a consistent pattern emerged in how attending physicians described adverse patient outcomes. When a patient recovered faster than expected, physicians in their notes overwhelmingly attributed the outcome to the quality of the treatment decisions, the speed of diagnosis, or the appropriateness of the medication regimen. When a patient deteriorated unexpectedly, the language of notes shifted toward patient non-compliance, underlying comorbidities, unusual disease presentation, or nursing errors. This is the self-serving attribution pattern applied at the level of professional self-documentation. The same physicians, the same unit, the same general patient population — the direction of the outcome determined the direction of the attribution.
This finding has direct consequences for error reporting in healthcare systems, because self-serving attribution actively impedes the kind of honest post-mortem analysis that quality improvement systems depend on. If physicians systematically attribute adverse outcomes to external factors, root cause analyses will consistently misidentify the contributing causes.
Case Study 2: The Investor Attribution Study (Financial Decision-Making)
In 1998, researchers Brad Barber and Terrance Odean, then at the University of California Davis, published an analysis of 66,465 household brokerage accounts over a six-year period from 1991 to 1996 in the Journal of Finance. While their primary analysis concerned overconfidence and trading frequency, the qualitative interviews and account narratives embedded in the study showed a strong self-serving pattern. Investors who made profitable trades attributed the outcome to research, insight, and timing acumen. Investors who made losing trades attributed the outcome to market irrationality, broker failure, bad luck, or macroeconomic factors outside their control. In post-outcome surveys, profitable investors showed significant increases in their self-rated investment skill. Loss investors showed no corresponding decrease. The asymmetry was structural: success raised self-assessed ability; failure did not lower it.
The implication is that investors can accumulate losses across years while their self-assessed competence remains intact or even rises, because each failure is attributed externally and each success attributed internally. This is a bias with a direct financial cost.
Case Study 3: The Classroom Teacher Attribution (Education)
The Johnson, Feigenbaum, and Weiby 1964 study remains one of the cleanest demonstrations ever conducted. But it was not unique. In 1975, psychologists Beckman and Cook replicated the core finding across a wider sample of student teachers at multiple institutions. They found that the self-serving attribution pattern in teachers became stronger over time: experienced teachers showed a more pronounced tendency to claim student success as their own achievement than novice teachers did. This is the opposite of what a learning-from-feedback model would predict. If teachers were calibrating their self-assessments against actual outcomes, their attributions should become more accurate, not more self-serving, as experience accumulated. Instead, the bias appeared to deepen with professional experience.
In 1985, Scott Allison and David Messick published an influential paper in the Journal of Experimental Social Psychology extending the analysis from individual attribution to group attribution. They found that groups showed exactly the same pattern as individuals: groups attributed their collective successes to their members' abilities and efforts, and their failures to external circumstances. The mechanism operated at the social level without requiring any explicit coordination among group members.
Case Study 4: The Driving Skill Illusion (Self-Assessment Research)
Patricia Cross's 1977 study, published in New Directions for Higher Education, included what has become the most-cited single statistic in the self-serving bias literature: 93 percent of American drivers, when surveyed, rated themselves as above-average drivers. The finding has since been replicated internationally. Svenson's 1981 replication, published in Acta Psychologica, found that 88 percent of American drivers and 77 percent of Swedish drivers rated themselves as above-average. By arithmetic definition, only 50 percent of any sample can be above average on a normally distributed skill. The gap between 93 percent and 50 percent is not a rounding error — it is a systematic, population-wide distortion of self-assessment.
Cross's study was conducted with faculty and students at the University of Nebraska and was not primarily about driving; it was embedded in a broader survey of self-assessments across professional and personal domains. The driving result became famous precisely because it is so easily quantified and so obviously impossible. But Cross found similar above-average illusions across academic skill, leadership ability, and interpersonal sensitivity. The driving finding was not an outlier; it was the most visible peak of a broad mountain range of self-serving distortion.
Intellectual Lineage: The History of the Concept
The intellectual roots of the self-serving bias reach back further than the 1960s laboratory. Fritz Heider, the Austrian-American psychologist whose 1958 book The Psychology of Interpersonal Relations (Wiley) founded modern attribution theory, argued that people are naive scientists — they intuitively seek causal explanations for events, particularly events that affect them personally. Heider distinguished between personal causation (an outcome caused by a person's dispositional qualities) and impersonal causation (an outcome caused by environmental forces or chance). He noted, without formalizing it experimentally, that people prefer personal explanations when they are the successful agent and impersonal explanations when they are the unsuccessful one.
Edward Jones and Keith Davis's 1965 paper "From Acts to Dispositions: The Attribution Process in Person Perception," published in Advances in Experimental Social Psychology, formalized the logic of dispositional attribution — why people tend to attribute behavior to stable internal traits rather than situations. This became the theoretical foundation for understanding why self-serving attributions tend to be dispositional for success (I succeeded because of who I am) and situational for failure (I failed because of what happened to me).
Harold Kelley's covariation model of attribution, introduced in the Nebraska Symposium on Motivation in 1967 and elaborated in subsequent papers through the 1970s, provided a more systematic framework. Kelley proposed that people attribute outcomes by assessing the covariation of behavior across persons, entities, and time — a logic essentially mimicking analysis of variance. Self-serving attribution, in Kelley's framework, emerges as a systematic violation of rational covariation analysis: people weight information differently depending on whether it supports a self-enhancing or self-protective conclusion.
The term "self-serving bias" was not consistently applied until the early-to-mid 1970s. Dale Miller and Michael Ross's 1975 review in Psychological Bulletin titled "Self-Serving Biases in the Attribution of Causality: Fact or Fiction?" attempted to assess whether the self-serving pattern was genuinely motivational or could be explained by purely cognitive factors. Their conclusion was cautious: the data supported cognitive explanations better than motivational ones in many cases. This skeptical review prompted Zuckerman's 1979 meta-analysis, which re-examined the evidence more comprehensively and came to a stronger conclusion: both motivational and cognitive factors contribute, and the self-serving attribution pattern is real, not an artifact of experimental design.
Shelley Taylor and Jonathon Brown's 1988 Psychological Bulletin paper, "Illusion and Well-Being: A Social Psychological Perspective on Mental Health," reframed the entire literature by asking not whether the bias is real but whether it is pathological. Taylor and Brown argued that three positive illusions — unrealistically positive self-evaluation, exaggerated perceptions of personal control, and unrealistic optimism — are not signs of psychological disorder. They are, statistically, the normal condition. And they are associated with better psychological adjustment, higher motivation, more persistent effort, and greater resilience. This paper generated enormous controversy and remains one of the most cited in the field. It shifted the conversation from "why do people make this error?" to "what does it do for them?"
Empirical Research: Key Studies and Findings
Zuckerman's 1979 Meta-Analysis
Miron Zuckerman's meta-analysis, published in the Journal of Personality in 1979, remains the foundational empirical consolidation of self-serving attribution research. Zuckerman reviewed studies using multiple methodologies: laboratory tasks with controlled feedback, field studies measuring attributions after real-world outcomes, and survey studies assessing attributional style. Across these methodologically diverse studies, he found consistent evidence for the self-serving pattern. The pattern was not limited to ability tasks; it appeared in social interactions, athletic performance, and academic contexts. Zuckerman also examined moderating variables: the bias was stronger when outcomes were public (others knew the result) than private, stronger when the stakes were higher, and stronger in individualistic cultural contexts than in more collectivist ones.
Taylor and Brown's Positive Illusions Framework (1988)
Shelley Taylor and Jonathon Brown's Psychological Bulletin review analyzed evidence across three domains of self-serving cognition: unrealistically positive self-evaluations, inflated perceptions of personal control, and unrealistic optimism about the future. They drew on dozens of studies to establish that each of these distortions, individually and collectively, is correlated with psychological well-being, effective social functioning, and motivation. Their theoretical argument was that natural selection may have favored motivated positive illusions because they sustain goal-directed behavior in the face of obstacles that a more accurate self-assessment might cause a person to abandon. This is an evolutionary functionalist argument: the bias is preserved in the population because, on average, it produces better outcomes than accuracy would.
Critics, including David Dunning and a number of researchers in the calibration tradition, have pushed back on this conclusion by arguing that well-being and accuracy are not necessarily in tension: people can feel good about themselves while still maintaining accurate self-assessments in important domains. But Taylor and Brown's core empirical finding — that positive illusions correlate with well-being in most populations — has not been seriously undermined.
Allison and Messick on Group Attribution (1985)
Scott Allison and David Messick's 1985 paper in the Journal of Experimental Social Psychology made an important extension. Prior research had focused overwhelmingly on individual self-serving attribution. Allison and Messick examined groups, asking whether group-level outcomes would show the same pattern. They found they did. In their laboratory studies, groups that succeeded in competitive tasks attributed success to their members' hard work and ability. Groups that failed attributed failure to the nature of the task, to the other group's unfair advantages, or to circumstances. The mechanism required no explicit agreement among group members — the pattern emerged spontaneously from the same cognitive and motivational dynamics that produce individual-level bias.
This finding has significant implications for organizational behavior. When a corporate team succeeds, the organizational narrative credits strategy and execution. When it fails, the narrative typically credits market conditions, competitor behavior, or macroeconomic headwinds. Self-serving attribution is not merely a quirk of individual psychology — it is a structural feature of how groups process their own performance histories.
Cross (1977) and Svenson (1981): The Above-Average Effect
Patricia Cross's 1977 survey data and Ola Svenson's 1981 replication stand as landmark demonstrations of the above-average effect in naturalistic settings. Cross surveyed faculty at the University of Nebraska and found that 94 percent rated themselves as above-average teachers, and 68 percent rated themselves in the top quarter. Svenson's driving study added cross-national evidence and demonstrated that the effect, while slightly less pronounced in Swedish drivers (77 percent above average vs. 88 percent of Americans), was fully present across both samples.
The above-average effect is related to, but distinct from, the self-serving bias as classically defined. Classical self-serving bias is about causal attribution — who or what gets credited for outcomes. The above-average effect is about comparative self-assessment — where one ranks oneself relative to peers. Both reflect a systematic positive distortion of self-appraisal, but they operate through somewhat different mechanisms. The above-average effect appears to involve selective attention to one's best attributes when making comparisons, a mechanism documented by David Dunning and colleagues in research published in the Journal of Personality and Social Psychology through the 1990s.
Limits and Nuances
Cultural Moderation
The self-serving bias is not universal across cultures at equal intensity. Research by Michael Bond and colleagues working in Asian contexts in the 1980s and 1990s, and by Heine, Lehman, Markus, and Kitayama in their influential 1999 paper "Is There a Universal Need for Positive Self-Regard?" in Psychological Review, found that the bias is substantially weaker in East Asian samples than in North American or Western European samples. Japanese and Chinese participants showed markedly less self-serving attribution in many studies, and in some studies showed a modest self-effacing pattern — attributing success to external factors and failure to internal ones. Heine and colleagues argued that this reflects not a different cognitive mechanism but a different cultural value structure: interdependent self-construal, which emphasizes fitting in and maintaining group harmony, produces less motivation to inflate self-assessments relative to others.
This finding complicates Taylor and Brown's evolutionary argument. If positive illusions were universally adaptive, they would be expected to be universal. Their cultural variability suggests that they are contingent on a specific type of individualistic self-construal that is culturally produced rather than biologically inevitable.
Domain Specificity
Self-serving attribution is not equally strong across all domains. Research by Claude Mezulis and colleagues, published in Psychological Bulletin in 2004 in a comprehensive meta-analysis of 266 studies covering 94 samples, found that the bias is strongest for outcomes in achievement domains (academic performance, work tasks) and weakest for interpersonal and social outcomes. In achievement domains, the logic "I succeeded because I worked hard and I failed because of bad luck" is highly accessible and culturally supported. In interpersonal domains, the equivalent logic ("our relationship succeeded because of my emotional maturity and failed because of their behavior") is more socially sanctioned and also more ambiguous, which may reduce the coherence of self-serving attributions.
Mezulis's meta-analysis also found that the bias is modestly larger in men than in women across studies, though the effect size is small. The magnitude difference was d = 0.14 — present and statistically significant in a large meta-analysis, but not practically large.
The Depressive Realism Qualification
The evidence for depressive realism — the superior accuracy of mildly depressed individuals in some attribution tasks — requires careful qualification. Alloy and Abramson's original 1979 finding concerned contingency judgment: how much control does one actually have over a specific outcome in a controlled setting? The finding of superior accuracy in depressed individuals applies most clearly to this specific type of judgment. When the task shifts from contingency judgment to broader self-assessment, the advantage of depressed individuals is less consistent. And clinical depression, as opposed to mild dysphoria, is not associated with accuracy; severely depressed individuals show their own characteristic distortions, including overgeneralization of failure and discounting of positive evidence.
The practical implication is that mild, subclinical negative affect may sometimes improve calibration in specific judgment tasks without producing the pervasive dysfunction of clinical depression. This is a narrow finding and should not be taken to imply that depression is cognitively beneficial.
When the Bias Reverses: Self-Handicapping and Modesty
Under some conditions, people make attributions that are the reverse of self-serving. Stephen Berglas and Edward Jones, writing in the Journal of Personality and Social Psychology in 1978, documented self-handicapping: the tendency to create or claim obstacles before an anticipated performance, so that failure can be attributed to the handicap rather than to the self, and success can be attributed to the self despite the handicap. This is a pre-emptive self-serving move rather than a post-hoc one, and it demonstrates that the underlying motivation — to protect a positive self-image — can produce apparently self-defeating behavior.
Public modesty norms, particularly in collectivist cultures, can also produce reversed attribution patterns in verbal reports even when internal self-assessments remain self-serving. This distinction between public claims and private beliefs is important: expressed self-deprecation in East Asian contexts may coexist with privately self-serving attributional patterns, or it may reflect genuine differences in self-regard. Research on implicit self-esteem measures has found more mixed evidence than research on explicit self-reports, and the relationship between implicit self-esteem, explicit self-report, and actual attribution remains an active area of investigation.
References
Johnson, L., Feigenbaum, R., & Weiby, M. (1964). Some determinants and consequences of the teacher's perception of causation. Psychonomic Science, 1(1), 141–142.
Langer, E. J., & Roth, J. (1975). Heads I win, tails it's chance: The illusion of control as a function of the sequence of outcomes in a purely chance task. Journal of Personality and Social Psychology, 32(6), 951–955.
Zuckerman, M. (1979). Attribution of success and failure revisited, or: The motivational bias is alive and well in attribution theory. Journal of Personality, 47(2), 245–287.
Abramson, L. Y., Seligman, M. E. P., & Teasdale, J. D. (1978). Learned helplessness in humans: Critique and reformulation. Psychological Review, 85(1), 49–74.
Alloy, L. B., & Abramson, L. Y. (1979). Judgment of contingency in depressed and nondepressed students: Sadder but wiser? Journal of Experimental Psychology: General, 108(4), 441–485.
Cross, P. (1977). Not can, but will college teaching be improved? New Directions for Higher Education, 17, 1–15.
Svenson, O. (1981). Are we all less risky and more skillful than our fellow drivers? Acta Psychologica, 47(2), 143–148.
Allison, S. T., & Messick, D. M. (1985). The group attribution error. Journal of Experimental Social Psychology, 21(6), 563–579.
Taylor, S. E., & Brown, J. D. (1988). Illusion and well-being: A social psychological perspective on mental health. Psychological Bulletin, 103(2), 193–210.
Miller, D. T., & Ross, M. (1975). Self-serving biases in the attribution of causality: Fact or fiction? Psychological Bulletin, 82(2), 213–225.
Mezulis, A. H., Abramson, L. Y., Hyde, J. S., & Hankin, B. L. (2004). Is there a universal positivity bias in attributions? A meta-analytic review of individual, developmental, and cultural differences in the self-serving attributional bias. Psychological Bulletin, 130(5), 711–747.
Heine, S. J., Lehman, D. R., Markus, H. R., & Kitayama, S. (1999). Is there a universal need for positive self-regard? Psychological Review, 106(4), 766–794.
Frequently Asked Questions
What is the self-serving bias?
The self-serving bias is the tendency to attribute positive outcomes to one's own character, ability, or effort, while attributing negative outcomes to external circumstances, bad luck, or other people. Zuckerman's 1979 meta-analysis of the attribution literature established the pattern as robust across contexts: people consistently take credit for successes and deflect responsibility for failures. The bias serves ego-protective and self-enhancement functions, maintaining positive self-esteem in the face of mixed outcomes.
What did the Cross 1977 driving study find?
Ola Svenson's 1981 replication of Cross's 1977 finding surveyed American and Swedish drivers about their skill and safety relative to others. In the United States, 93% of drivers rated themselves as above average in skill, and 88% rated themselves as above average in safety — both statistically impossible majorities. Svenson's cross-national comparison found that Swedish drivers showed the same pattern, with 69% rating their skill above the median. The study became the canonical demonstration that self-serving bias produces widespread, systematic overestimation of personal competence.
What is depressive realism and how does it challenge the self-serving bias?
Depressive realism, proposed by Lauren Alloy and Lyn Abramson in their 1979 Journal of Experimental Psychology paper, holds that mildly depressed individuals make more accurate judgments about personal control than non-depressed individuals. In contingency learning tasks, non-depressed subjects overestimated their control over random outcomes — a self-serving illusion — while depressed subjects showed more accurate calibration. This suggested that positive illusions, including the self-serving bias, may be the default state, with depression removing them. Later research has qualified the finding, but it remains a significant challenge to assumptions about the adaptiveness of biased self-attribution.
Is the self-serving bias universal across cultures?
No. Heine et al.'s 1999 Psychological Review paper documented that the self-serving bias is substantially weaker or absent in East Asian cultures, particularly Japan. Where Western samples consistently show self-enhancement, Japanese samples often show self-criticism. Heine argued that self-enhancement serves self-improvement goals in individualist cultures but conflicts with face-maintenance and group harmony norms in collectivist ones. Mezulis et al.'s 2004 meta-analysis of 266 studies confirmed the cultural moderation: the bias is strongest in North America and Western Europe, significantly weaker in Asia.
How does the self-serving bias affect financial decision-making?
Brad Barber and Terrance Odean's research on individual investor behavior found that investors who traded most actively earned the worst returns — yet consistently attributed poor performance to market conditions while crediting good performance to their own skill. This asymmetric attribution allows investors to maintain confidence in their ability despite systematically negative outcomes, fueling continued overtrading. Odean's 1999 Journal of Finance analysis of 10,000 brokerage accounts found that stocks investors sold outperformed the stocks they bought by 3.4 percentage points annually, with investors showing no evidence of updating their self-assessments despite persistent underperformance.