Organisations that present themselves as meritocratic can show more bias in their reward decisions, not less.
Castilla and Benard ran experiments in 2010 in which managers allocated bonuses from identical employee profiles. Where the organisation was described as valuing meritocracy, managers awarded more to men than to equally qualified women. The mechanism proposed is moral credentialling: a person or an institution that has just affirmed its own fairness feels licensed to follow its preferences without examining them.
In practice
- A firm publishes a merit-based pay policy and its pay gap widens, because managers now treat their own judgement as the policy's output.
- A committee that has just completed bias training approves its usual shortlist with unusual confidence.
- Possible explanations
- The proposed mechanism is moral credentialling: a person or institution that has just affirmed its own fairness feels licensed to follow its preferences without examining them.
- In its favour
- The experiments are strong on internal validity, using identical profiles with only the organisational description changed, and the result fits the wider moral licensing literature.
- Against it
- The setting is narrow and field evidence is thin, and direct replications in the moral licensing literature it belongs to have often failed, so the scope of the effect is uncertain.
- Standing
- Contested in scope. The experiments are strong on internal validity and narrow in setting, and field evidence is thin. Compare moral licensing, the general form.
- Subject
- Organizations & Bureaucracy
- Essay
- How organisations behave when nobody is steering