Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We elicit human conditional punishment types by conducting experiments. We find that their punishment decisions to an individual are on average significantly positively proportional to other members’ punishment decisions to that individual.