Bridging the Research-Practice Gap: The Paradox of Meritocracy

Most people in capitalist countries such as the United States probably think that the economic system should be and is a meritocracy. In addition, an increasing number of organizations emphasize meritocracy as a core value and establish meritocratic systems (e.g., pay for performance), in part to eliminate bias and increase the perception of fairness in the workplace.

Unfortunately, Emilio J. Castilla of the Massachusetts Institute of Technology and Stephen Benard of Indiana University argue that meritocratic values and practices — if not implemented carefully — may actually increase bias rather than decreasing it. They call this “the paradox of meritocracy.”

Their underlying logic draws from psychological research on bias. First, organizational emphasis on performance may increase the tendency to apply stereotypes about competence. Second, because most people try to avoid appearing biased — both to others and to themselves — they behave in a biased manner only when they are confident that the behavior will not be seen as biased.

When working in an organization that stresses meritocracy, decision-makers feel confident that their actions are unbiased and will be seen as unbiased. This is precisely the situation in which their biases are most likely to affect their behavior. Thus, organizational emphasis on meritocracy may both stimulate the tendency to apply competence-related stereotypes and provide the cover needed to act on the stereotypes.

Past research has revealed gender and race pay disparities in organizations that emphasize meritocratic procedures, even when controlling for human capital (e.g., experience). However, because the meritocratic practices already existed in the studied organizations, Castilla and Benard point out that we cannot know whether the disparities occurred in spite of the meritocratic procedures or because of them. The only way to distinguish between those two possibilities is through experimental research, which they report.

The participants in their experiments were business graduate students, most of whom had and or liked jobs with supervisory responsibilities. They were given information about three [imaginary] employees who worked at the same job in the same unit and had been evaluated by the same supervisor. Based on the supervisor’s evaluations (ratings and comments), the participants allocated $1,000 in bonuses among the three employees. This two-stage evaluation system was used because it is becoming increasingly popular, in part because it is believed to limit discretion. The participants were told either that the organization’s core values emphasized meritocracy or that the values emphasized use of regular evaluation procedures.

Two of the employees (one man and one woman) had been given equivalent high performance evaluations and the third (a man in Study 1 and a woman in Study 2) had been given a lukewarm evaluation.

The participants gave the high performing man larger bonuses than the equally high performing woman when and only when they were told that the organization’s core values emphasized meritocracy. In addition, in this situation the high-performing man received a larger bonus than did the lower-performing man, but performance had no effect on the bonuses received by the female employees.

Consistent with their assumptions about the underlying mechanisms, the paradox of meritocracy was more evident in allocation of bonuses (which are normally private) than in decisions that usually are more visible (e.g., promotions). It is important to note that these results were not affected by the research participants’ gender or other characteristics.

In summary, the authors found that an emphasis on meritocracy and performance-based pay actually led to gender discrimination in the allocation of bonuses.

The favor with which we view meritocracy and performance-based systems gives this research significant implications for diversity practitioners.

The findings touch on the most difficult aspects of diversity work — stimulation of stereotypes, moral credentialing that provides cover to act on those stereotypes, bias masquerading as objectivity, the nature and workings of privilege, and the potential devastation to self-confidence and performance of the women and minorities who may be deemed, through bias, as poor performers.

Because organizations are unlikely to abandon their cultures of perceived meritocracy, practical implications for mitigation of the paradox should be considered. Monitoring the organization through skilled use of statistical assessments can provide insight into the differential impact of management practices on different employee groups.

All leadership training should include material on the links between diversity, personal leadership and performance.

Finally, we urge practitioners to invite a scholar to replicate this research in their own organization, if their organization plans to adopt a merit-pay or other meritocratic policy. Such research might evaluate the success of procedures designed to mitigate problems.

— Dr. Renee Yuengling is a workplace diversity consultant in Washington, D.C. Dr. David A. Kravitz is a professor of management at George Mason University.