Theoretical foundation and empirical assessment of representation and meritocracy in academia

Quantifying meritocracy directly is unfeasible because it requires large research efforts (such as surveys and controlled hiring experiments) that do not benefit the existing power structure. We circumvent this conundrum by proposing the use of openly accessible surname-publication data to quantify intergenerational representation in academia, which captures the socioeconomic aspect of diversity relative to the general population. We then use individual-based models of the intergenerational cycle of academic selection and reproduction to show that representation and merit in academia are entangled. We distinguish merit, or an academic candidate’s potential to produce given opportunities, from produced capital, including accomplishments before graduate school that only imperfectly predicts merit in a complex and changing world. Data from Harvard and US income groups and multiple independent model predictions all suggest that US academics are twice as likely as others to historically be academics by surname and underperform compared to a more representative academia, but individual-based affirmative action consistently raises academics’ mean merit. For academics aiming to tackle global crises, a lack of representation and merit may ultimately prevent actions necessary to avert disasters. This study reveals the magnitude of inequality, supports an individual justice foundation of affirmative action, and calls for recruitment evaluation that values merit over capital.