Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
The effect of health-plan membership on the utilization of health-care services is of interest to both consumers and policymakers. Often estimation of that effect is difficult because data are nonexperimental and the dependent variable exhibits a high proportion of zeros. The authors propose a model of utilization that addresses both problems. After controlling for chronic illness and other observed variables, they find no evidence of further selectivity bias in equations for physician contacts and inpatient days. The authors' estimates of the effect of health-plan membership on utilization of services are similar to those from experimental data. Coauthors are Roger Feldman, Steven Cassou, and Michael Finch. Copyright 1991 by MIT Press.