Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
I. Moral hazard, 693. — II. Adverse selection, 694.