Self serving bias
Self serving bias is the tendency to explain your successes by what you are and your failures by what happened to you, so that credit is internal and blame is external.
What self serving bias is
Self serving bias is an asymmetry in attribution. Attribution is the everyday act of deciding why an outcome occurred, and the two options are broadly internal, meaning something about the person, and external, meaning something about the situation. Nobody uses one exclusively. The bias is that the same person switches between them depending on whether the outcome was good, and switches in a direction that protects their own standing.
What distinguishes it from the neighboring members of the cognitive bias family is the presence of a self to protect. Confirmation bias will run on a belief about the weather. Self serving bias needs the outcome to reflect on you, which is why it is strongest in exactly the settings where accurate diagnosis matters most: performance reviews, project post-mortems, exam results and any account of a decision you personally made.
The one sentence version, and the two halves of the asymmetry
In one sentence, success gets attributed to ability and effort, and failure gets attributed to luck, timing, the tools or other people. Split into its two halves, the bias has a self-enhancing side and a self-protective side, and they are not always present together.
| Outcome | Typical explanation offered | What it does for the speaker |
|---|---|---|
| Good, and mine | Skill, preparation, judgment, hard work | Enhances standing and self-image |
| Bad, and mine | Bad luck, a difficult brief, unreliable data, other people | Protects standing and self-image |
| Good, and someone else's | Favorable conditions, an easy problem, timing | Preserves relative position |
| Bad, and someone else's | Carelessness, poor judgment, lack of skill | Preserves relative position |
Read the four rows together and the shape becomes obvious: the explanation is chosen to keep the speaker looking competent, whichever way the result went and whoever produced it. Each individual explanation may be perfectly true, and no rule of deductive reasoning is broken by any of them. The pattern across all four is the evidence.
The evidence, and the argument about what causes it
The most quoted early demonstration outside the laboratory is a 1980 study of newspaper coverage of sports by Richard Lau and Dan Russell, published in the Journal of Personality and Social Psychology. They coded the explanations that players and coaches gave in press accounts and found the expected asymmetry: wins were explained by the team's own qualities, losses were explained by conditions, officials and circumstance. The data was not collected in an experiment and no one was primed to produce it.
The interesting part of the literature is the argument about the cause. Dale Miller and Michael Ross, in a 1975 review in Psychological Bulletin, argued that the asymmetry does not require any motive at all. People generally expect to succeed at things they have practiced, and they intend the outcomes they work toward. A result that matches your expectation gets explained by the thing you were doing; a result that surprises you sends you looking outside for a cause. That is an informational account rather than a motivational one, and it predicts much of the same behavior without any ego defense.
Later work, including a large meta-analysis published in 2004, found the attribution asymmetry to be widespread but not uniform, with variation by age and across cultures. That variation is a point in favor of both accounts and settles neither. For a reader trying to think more clearly, the honest position is that the pattern is well established and the mechanism is not fully resolved, which changes nothing about the remedies.
Where self serving bias shows up in an ordinary week
Self serving bias shows up wherever an outcome has to be explained by the person responsible for it, such as annual reviews, project retrospectives, sales debriefs, returned exam papers, driving, cooking and any conversation that begins with the words "what went wrong".
- The retrospective. A software team ships two weeks late. The retrospective identifies unclear requirements, a dependency that broke and a sick colleague. All three are real. The estimate that was optimistic on day one does not appear on the list, and the same team, when it shipped early in the spring, credited its planning.
- The exam paper. A student who scores well says the material suited them. The same student, scoring badly on the next paper, says the questions were unfair. Both statements can be true. Only one of them was checked against the paper.
- The forecast. A manager whose sales forecast came in high blames a competitor's discount. The competitor's discount was real, and the manager's forecast was also 40 percent above the previous three years, which is the part that goes unmentioned.
Notice what this does to learning. A failure explained entirely by external causes contains no lesson, because nothing you control produced it. A team that runs three such retrospectives in a row has documented its own history and learned nothing from it, and the shortfall in insight is invisible from the inside, in much the way described on the Dunning Kruger effect page.
Self serving bias, actor observer bias and the fundamental attribution error
Self serving bias is about good and bad outcomes; actor observer bias is about self and other. They are frequently confused because both concern internal and external attributions, and exam questions exploit the overlap.
- Self serving bias: you explain your own success internally and your own failure externally. The axis is the outcome.
- Actor observer bias: you explain your own behavior by the situation and other people's behavior by their character. The axis is who is being explained.
- Fundamental attribution error: you underweight the situation when explaining another person's behavior. It is the observer half of the pattern above, stated on its own.
- Confirmation bias: you notice and remember evidence that fits what you already believe. The axis is the belief, not the self.
- Hindsight bias: after the outcome, you believe you expected it. It reshapes the memory of the prediction rather than the explanation of the result.
There is also a group version, sometimes called group serving bias, in which the same asymmetry is applied to a team, a department or a school. It behaves identically, and it is harder to challenge, because the person pointing it out is now arguing with a room rather than with an individual. That is where it meets groupthink.
What actually reduces self serving bias
What reduces it is symmetry, enforced in writing before the outcome is known. Five moves that work:
- Record your prediction and your reasoning before the result arrives, and reread it afterward.
- Explain a success and a failure using the same categories, in the same format, on the same page.
- Ask what a stranger with no stake in your reputation would list as the top three causes.
- Apply your explanation of the failure to the earlier success and see whether it still fits.
- Have someone else draft the account of what happened, and restrict yourself to correcting facts.
The first is the strongest, because the bias operates on memory as well as on explanation, and a written prediction is the one thing memory cannot revise. Be honest about the ceiling: none of this makes you neutral about your own reputation, and it is not clear that being fully neutral would be good for you. A modest self serving bias appears to be the ordinary condition of a functioning adult, not a defect. The goal is to keep it out of the documents that are supposed to produce learning.
The test to run on your own explanation
Ask this pair of questions whenever you catch yourself explaining a result:
- Would I be offering this same explanation if the outcome had gone the other way?
- If a colleague gave me this exact explanation for their result, would I accept it?
The second question is the sharper one, because it borrows the standard you already apply to other people, which is usually stricter. If you would raise an eyebrow at a colleague's version, the eyebrow belongs to your own.
The exam definition, for AP psychology and MCAT questions
Exam questions on self serving bias test one distinction and one label, so the compact answer is worth having precisely. Self serving bias is the tendency to attribute one's own successes to internal, dispositional factors and one's own failures to external, situational factors. It sits in social psychology, under attribution theory, alongside the fundamental attribution error and actor observer bias.
Two traps recur in test items. First, a question that describes a person explaining someone else's failure by their character is testing the fundamental attribution error, not self serving bias. Second, a question that describes a person recalling that they knew the result all along is testing hindsight bias. Read for the axis: outcome versus outcome is self serving, self versus other is actor observer, and prediction versus memory is hindsight.