Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
We examine the long-term effects of a 1998-2003 randomized experiment in Tulsa, Oklahoma with Individual Development Accounts