Risk Selection in ACA Marketplace Exchanges

Background: While maintaining stable risk pools is a key challenge for highly regulated, competitive individual insurance markets, relatively little evidence exists on the extent and sources of risk selection in these markets. Yet, this evidence is important for developing policies to promote market stability. In this paper, we provide empirical evidence of the extent and sources of risk selection in ACA marketplace exchanges.

Data and Methods. We use 2015 to 2016 Optum commercial health insurance claims to compare health care spending patterns between exchange and non-exchange enrollees. We identify cohorts of people age 18 to 63 newly enrolling in either exchange or non-exchange plans at any point in 2015 and calculate monthly spending overall and by month of enrollment through December of 2016. We include all covered spending on medical and inpatient services and prescription drugs. Among people with exchange coverage, we identify those enrolling during a special enrollment period (SEP) based on whether they initially enrolled in a month other than January, February or March. We determine the extent to which: 1) the average and distribution of covered spending differ between exchange and commercial enrollees, 2) exchange and commercial enrollment and disenrollment takes place outside standard enrollment periods, and 3) health spending varies by enrollment and disenrollment timing for exchange and commercial enrollees.

Results. Health care spending was greater for people enrolling in exchange plans relative to those with other types of commercial coverage. Among people enrolling in January of 2015, average monthly expenditures were 36% higher for exchange than for non-exchange enrollees ($590 for exchange relative to $434 for commercial enrollees) with more than 80% of the difference attributable to enrollees the top 20^th percentile of the spending distribution. Among exchange enrollees signing up during regular open enrollment, 60% disenrolled outside the subsequent open enrollment period with 30% disenrolling within 3 months. Among people newly enrolling in exchange coverage in 2015, 25% enrolled during a special enrollment period. Average monthly spending was higher for SEP enrollees than for those signing up during regular open enrollment (OEP) ($715 versus $560) Early disenrollees (<= 3 months) had lower average spending than those with longer enrollment periods and average spending was lower in the last month among those dropping coverage outside open enrollment periods.

Conclusions. Our results indicate that people enrolling in exchange coverage have higher average expenditures than those in commercial, primarily employer-sponsored coverage We also find evidence that exchange coverage is less stable than commercial coverage and that unfavorable selection into exchange plans is facilitated by the ability of people to enroll in coverage outside of standard open enrollment periods as well as by selective disenrollment.