Will health plan outreach increase take-up by enrollees that are healthier or sicker?

  • by

Why don’t people eligible for health insurance exchange plans (i.e., Obamacare plans) enroll? One reason could be that they value the health insurance benefit at less than the cost. Another reason could be behavioral frictions; informational search costs and psychological frictions are costly and may preclude enrollment even when benefits are more than cost. For instance, many individuals may do not know what their subsidy would be (if any) in the Obamacare exchanges (see Greenberg 2017)

We all know that filling out employee benefit selection each year is a pain, but does it affect take-up? A paper by Wright et al. (2017) uses a randomized controlled design and find that randomized “nudges” substantially increased enrollment in Oregon’s state Medicaid plan.

If these nudges are effective, a key question is whether the marginal enrollee impacted by advertising, nudges, reminders (or whatever you want to call them) are more or less expensive than the average patient. One may think that sick individuals (i.e., i.e., those with more medical costs) may have more frictions since searching for insurance is more costly; on the other hand, the benefit of insurance to sicker individuals is likely higher so perhaps the nudges/reminders are more likely to enrollee healthier individuals.

To examine this question empirically. a paper by Domurat, Menashe, and Yin (2021) examines potential enrollees with California’s ACA exchange, known as Covered California. The potential enrollees include two groups of individuals: those who applied directly to Covered California, were determined eligible to enroll (and, if applicable, eligible for subsidies), but never selected a plan; and individuals previously enrolled in Medicaid but who were now no longer eligible. The authors deem these groups the “open enrollment” and “county referral” group respective.

These individuals were then randomized into 1 of 5 outreach arms:

  • Control (Arm 1): Received no direct communication beyond the generic outreach and state-wide marketing activities used by Covered California for all consumers
  • Basic Letter (Arm 2): Reported the open enrollment deadline, general benefits of insurance, and the Covered California website and telephone number where they could shop for plans.
  • Subsidy and Penalty (Arm 3): Reported Basic Letter information (Arm 2) plus the household’s estimated monthly subsidy and tax penalty, based on their reported income and household size.
  • Price Compare (Arm 4): Reported the content of Arms  2 and  3, plus a table listing the Silver and Bronze plans offered in their market, with their net-of-subsidy premium.13
  • Price and Quality Compare (Arm 5): Reported the content of Arm 4, but the table also included plans’ quality rating under the ACA’s five-star quality rating system (QRS).

The authors used three data sources: (i) Covered California application and enrollment information, (ii) risk scores from Office of Statewide Health Planning and Development (OSHPD) measuring prospective risk and (iii) Verisk risk scores reported by IBM. The authors measure differences in take-up across groups controlling for family size, demographics, and family income. The authors also measure the relationship between individuals’ health risk score and to the group they were assigned to conditional on take-up.

Unsurprisingly, the authors do find that all the intervention arms (Arm 2-5) lead to an increase in take-up. The increase was 1.3 percentage points or a 16% increase over baseline. The impact of these “nudges” was more effective for the low income groups than high incomes groups.

Finally, to answer the question posed in the title of this post, the marginal enrollee impacted by the outreach was healthier than average.

Overall, the letter interventions led to a 5.1  percent decrease in average risk…This implies that the average risk of the marginal respondent to the letter interventions was 37 percent lower than the average risk of inframarginal enrollees in the study sample. This finding is consistent with larger treatment effects of letters on take-up among those with lower baseline CDPS [i.e. health] risk.

However, this effect was much more pronounced among the open enrollment group (i.e., those who enrolled but did not pick a plan) than the county referral group (i.e., individuals who moved from Medicaid eligibility to ACA plan eligibility)

In short, if you are running a health plan, more direct outreach to ACA-eligible individuals is likely to increase enrollment by relatively healthy people. If health insurance premiums are not risk-adjusted, health plans could gain significant return on investment from additional outreach.