Case wall · trace the string
Social Comparison Bias
The tendency to make recommendations that prevent similar others from surpassing you on the very dimension on which you yourself have high standing, because that dimension underwrites your self-esteem.
Social Comparison Bias
The tendency to make recommendations that prevent similar others from surpassing you on the very dimension on which you yourself have high standing, …
Exhibit · study
Study 1 — Harvard Law hiring (hypothetical)
Participants recommended the candidate who did NOT surpass them on their own high-standing dimension.
Exhibit · study
Study 2 — real team-selection choice
High-verbal participants favored the high-math candidate (only 26% chose high-verbal Hardy);
Exhibit · study
Study 3 — real employees + self-esteem mechanism
Employees recommended the new recruit be higher on the dimension where the employee was only average.
Exhibit · field case · 2010
Nobel laureate allegedly dissuading a promising same-field recruit
Garcia, Song & Tesser report, in their General Discussion, an incident in which a Nobel laureate allegedly discouraged …
Exhibit · field case · 2010
Erosion of departmental strengths as members protect personal standing
The authors cite Owen-Smith (2001) and argue the social comparison bias may help explain why some top-tier departments …
Person of interest
Social Comparison Bias
The tendency to make recommendations that prevent similar others from surpassing you on the very dimension on which you yourself have high standing, because that dimension underwrites your self-esteem.
Strings traced
- study · Study 1 — Harvard Law hiring (hypothetical)
Participants recommended the candidate who did NOT surpass them on their own high-standing dimension.
- study · Study 2 — real team-selection choice
High-verbal participants favored the high-math candidate (only 26% chose high-verbal Hardy);
- study · Study 3 — real employees + self-esteem mechanism
Employees recommended the new recruit be higher on the dimension where the employee was only average.
- field case · Nobel laureate allegedly dissuading a promising same-field recruit
Garcia, Song & Tesser report, in their General Discussion, an incident in which a Nobel laureate allegedly discouraged an exceptionally promising faculty member doing similar work from joining his de…
- field case · Erosion of departmental strengths as members protect personal standing
The authors cite Owen-Smith (2001) and argue the social comparison bias may help explain why some top-tier departments lose prestige over time: individuals 'unwittingly fail to reproduce departmental…
The connective logic
Garcia et al. argue the bias is downstream of self-esteem protection within a social comparison context. People's self-worth is contingent on specific domains (Crocker & Park, 2004); the dimension on which one ranks highly becomes the dimension on which self-esteem is staked. Because social comparison concerns intensify in proximity to a standard (being highly ranked), a candidate who would surpass you on that exact dimension threatens your standing, so you 'shape the comparison context' by recommending against them — favoring instead someone strong on an orthogonal dimension. This recasts a Festinger/Tesser self-evaluation-maintenance dynamic as a hidden conflict of interest in recommendations.
The authors pre-empt a 'complementarity' account (you simply want collaborators with different strengths) by holding complementary expertise constant (Study 1) and making the skills irrelevant to the task (Study 2). The self-esteem mechanism in Study 3 is correlational — self-esteem was neither manipulated nor controlled — so causal claims about the mechanism remain weakly identified, as the authors concede.
Case status — replication
Thin file
The seminal evidence is three small studies (N=29, 40, 55) reported in a single 5-page 2010 paper, with significance at p<.05 and no standardized effect sizes. A 2016 conceptual replication and extension by Jia, Lu, Xie & Huang (three experiments + one field study) supported the effect and identified self-threat as a moderator (supervisors < subordinates). No pre-registered direct replications, meta-analyses, or documented replication failures of this specific named effect were located. Confidence is constrained by small samples and the absence of large-scale or registered tests — not by known failures.
How to break the chain
Decouple the recommendation from your own comparison context: judge candidates against the organization's goals, not your standing, and use structured criteria scored before you see who is strong on 'your' dimension. Solicit recommendations from people who are NOT high on the same dimension as the candidate, and make the self-esteem stake explicit so it can be checked. Jia et al.'s finding that authority reduces self-threat suggests framing yourself as a steward of group quality rather than a competitor.
Garcia et al. show the bias is driven by self-esteem protection in a comparison context (Study 3); Jia et al. (2016) show reduced self-threat (e.g., supervisory role) attenuates the bias — implying interventions that lower felt threat should help.
Catch it in the act
Watch for it when you are highly regarded on one specific dimension and are asked to recommend, hire, refer, or vote for someone strong on that same dimension. If you find yourself preferring a candidate whose strengths lie elsewhere — while telling yourself it's about 'complementarity' or 'fit' — ask whether you're actually protecting your own standing. The risk peaks for high-status individuals and is reduced by formal authority (supervisors feel less threatened than subordinates).