Tweaking the Marketplace Enrollment Process Could Magnify Effects of Larger Premium Tax Credits

Editor’s Note: This analysis is part of the USC-Brookings Schaeffer Initiative for Health Policy, which is a partnership between Economic Studies at Brookings and the University of Southern California Schaeffer Center for Health Policy & Economics. The Initiative aims to inform the national health care debate with rigorous, evidence-based analysis leading to practical recommendations using the collaborative strengths of USC and Brookings. This piece was edited on September 14, 2022 to correct the effective date of a regulation and an associated hyperlink.

The Inflation Reduction Act included a three-year extension of enhancements to the premium tax credit that were originally enacted in 2021. These enhancements reduced the cost of Marketplace coverage by making the tax credit more generous for people who were already eligible and by extending it to people with incomes that were previously too high to qualify. Consistent with model-based projections that lower premiums would increase coverage, Marketplace enrollment is now around 3 million people (29%) above its 2019 level, which has helped push the national uninsured rate to an all-time low.

These tax credit enhancements create big opportunities to improve the Marketplace enrollment process in ways that would further increase insurance coverage. In particular, the expanded tax credits have made many more people eligible for Marketplace coverage at a zero or near-zero premium. Most importantly, they ensure that almost everyone with an income below 150% of the federal poverty level (FPL)—a group that currently accounts for about one-third of all Marketplace enrollees—now has at least two silver plan options with zero- or near-zero premiums. Previously, premiums this low were typically only found on the less generous bronze tier. In this piece, we describe three concrete changes to the Marketplace enrollment process that would leverage this changed premium landscape to increase insurance coverage.

Change #1: Ensure small residual premiums are not a barrier to coverage

One consequence of the recent enhancements to the premium tax credit is that many more enrollees now face premiums that are almost, but not quite, zero. Recent work has estimated that 404,000 Marketplace enrollees in the states served by HealthCare.gov currently owe a small positive premium (defined as a non-zero premium of less than 0.5% of the gross premium of the enrollee’s plan, which translates to around $3 per month on average). Without the recent enhancements to the premium tax credit, this number would have been considerably smaller, at just 77,000.

These small positive premiums can arise in a few different ways. One common scenario occurs when enrollees opt for plans that include limited coverage for services that are not considered essential health benefits (e.g., vision or hearing coverage). Because the premium tax credit can only be used to pay for the EHB portion of a plan’s premium, these enrollees end up responsible for a small premium that covers the non-EHB benefits even if their tax credit is large enough to cover the plan’s full premium. In other cases, enrollees may simply opt for a plan with a premium slightly larger than their tax credit. All of these scenarios become more common when tax credits are larger and net premiums are smaller.

There is ample empirical evidence that even small net premiums can meaningfully reduce insurance enrollment. This is most likely not because small premiums pose a substantial financial burden, but rather because remitting even a small premium requires enrollees to take additional action, which imposes hassle costs and cognitive burdens. For instance, unless enrollees set up automated payments, they must remember to take action on time, every month. For enrollees who lack a bank account, paying premiums can require purchasing a money order each month; about 10% of households with incomes between $15,000 and $30,000 were unbanked in 2019. As a result of these non-financial hurdles, many enrollees facing small premiums either fail to initiate premium payments or stop paying premiums unintentionally.

Drawing on this empirical literature, the research cited above concluded that waiving small positive Marketplace premiums could increase coverage by 48,000 people in the HealthCare.gov states. The Centers for Medicare and Medicaid Services (CMS) could likely achieve this outcome administratively. Under rules finalized in 2016, insurers may treat enrollees who have paid a high enough percentage of the enrollee’s net-of-tax-credit premium as having paid in full for the purposes of effectuating or continuing coverage. To address the scenario of interest here, CMS could extend this policy to cases where the outstanding premium balance is small relative to a plan’s gross premium (e.g., less than 0.5% of the gross premium), not just cases where the outstanding balance is small relative to the net premium.

CMS might want to go further than simply allowing insurers to treat these types of enrollees as having paid in full and require them to do so (if the legal authority exists).[1] While it is plausible that adopting this type of policy would be profitable for insurers since the people who would gain coverage would likely be considerably healthier than average, suggesting that many insurers might take up this option even if not required to do so, this is not guaranteed. Notably, many of the affected enrollees would be eligible for the most generous tier of cost-sharing reductions, and a portion of the benefit of gaining healthier-than-average enrollees would be shared with other insurers via the risk adjustment system, which might make these enrollees less profitable (from the perspective of an individual insurer) than their health status would suggest. Moreover, even if these enrollees were indeed profitable for insurers, the absolute amount of additional profit available would likely be relatively modest, raising the risk that insurers would overlook this opportunity, at least for some period of time after the option became available.

Change #2: Automatically transition people who stop paying premiums to $0 plans when feasible

Currently, when enrollees do not pay their monthly Marketplace premiums, they enter a three-month “grace period.” If the enrollee has not made all outstanding payments by the end of the third month, the enrollee’s coverage is terminated. Payment lapses may be one reason that Marketplace enrollment falls over the course of the plan year. As discussed above, lapses can occur because enrollees forget to pay premiums or because paying involves significant hassle costs. In other cases, enrollees’ life circumstances may change in ways that make it difficult to pay their monthly premiums.

The expanded availability of zero-premium plans may raise the incidence of lapses at the start of new plan years due to year-to-year churn in which plans offer $0 premiums. While enrollees with incomes below 150% of the FPL now have access to two (near-)zero-premium silver plans, the identity of those plans often changes from year to year. Prior work has found that 93% of HealthCare.gov enrollees lived in counties where at least one silver plan had no premium in 2021, but a positive premium in 2022; 84% lived in counties where all silver plans that had no premium in 2021 had a positive premium in 2022. In general, Marketplace enrollees are automatically renewed into the same plan unless they actively switch, so many enrollees will find themselves newly facing premiums despite staying in the same plan. Some may not realize that they suddenly owe a premium and, as a result, fail to initiate premium payments.

“Automatic retention” policies could mitigate attrition due to premium lapsing in cases where an enrollee has access to a zero-premium plan. Under an automatic retention policy, rather than terminating enrollees who have lapsed on their monthly premiums, the Marketplace would transition these enrollees to zero-premium plans for which they are eligible after their grace period expires.[2]

Massachusetts had an automatic retention policy in place in its pre-ACA Commonwealth Care program from 2010 through 2013 for enrollees with incomes between 100 and 150% of the FPL. Research examining this policy has estimated that 14% of enrollees in this income group were retained by the policy over the four years the policy was in effect; a large share of this effect is attributable to enrollees whose plans switched from being zero-premium to positive-premium at the start of a new plan year and never initiated premium payments. Enrollees retained by the policy were 4 years younger, on average, and had 9% lower medical spending than other enrollees in the same income group.

CMS could implement this type of policy on HealthCare.gov, and states that run their own Marketplaces could do the same. Some key questions would need to be answered ahead of implementation. First, should an automatic retention policy apply to any enrollee eligible for a zero-premium plan, even if that means transitioning the enrollee to a plan in a different metal tier (e.g., from a silver plan into a bronze plan)?  Or should it apply only to enrollees who have access to a zero-premium plan in the same metal tier as their current plan (or a more generous metal tier)? Applying the policy to all enrollees eligible for zero-premium plans would maximize the number of people who retain coverage since essentially all of these people would otherwise be disenrolled. Thus, the rationale for a narrower approach would need to involve other goals. In principle, a narrower policy could spur more people to actively re-enroll and select plans well-matched to their needs, but evidence from related settings suggests that this effect would be very small and, regardless, enrollees with incomes above 150% of the FPL would generally be unable to re-enter the Marketplace until the next open enrollment. A narrower policy could also conceivably be less operationally burdensome for CMS or easier for insurers to adjust to, although it is questionable whether either of these considerations are important in practice.

Second, policymakers will need to decide how to select the plans that enrollees transition to. One option would be to select the plan with the lowest premium (perhaps subject to metal tier requirements). Another option would be to select the $0 plan with the highest actuarial value (inclusive of cost-sharing reductions, for those who qualify). The choice between these two options would depend on whether policymakers believed that the improved access to care and financial security afforded by enrollment in a more generous plan justified the added federal tax credit costs. Another, more operationally challenging, approach would be to attempt to map enrollees to zero-premium plans offered by insurers with whom enrollees have prior enrollment experience (which might increase the likelihood that enrollees are transitioned to plans that include their providers) before applying criteria like those described above.

Change #3: Automatically enroll $0-premium-eligible people who start enrolling, but do not finish

Enrollees sometimes begin the Marketplace enrollment process but fail to complete it. There are no national estimates of this phenomenon, but Covered California estimates that tens of thousands of state residents start—but never finish—coverage applications during annual open enrollment periods. When zero-premium plans are available, this incomplete take-up likely typically reflects the cognitive demands and hassle costs of completing the enrollment process, not a considered choice to forgo coverage.

Massachusetts can again be instructive here. In early 2022, the state’s Marketplace changed its online application to include a checkbox that allows enrollees to opt in to being automatically enrolled if they are determined eligible for a zero-premium plan and do not complete the enrollment process. This checkbox is presented after the enrollee enters (and attests to) household information needed to conduct an eligibility determination but before applicants see the menu of plans available to them.[3]

This type of selective automatic enrollment process would raise implementation questions similar to those outlined for automatic retention, particularly: whether to apply the policy to all applicants or only those eligible for zero-premium coverage in certain metal tiers; and how to select the plan that applicants were automatically enrolled into. The tradeoffs involved in those choices would also be similar.

Should policymakers make these changes to the marketplace enrollment process?

The evidence discussed above suggests that these changes to the Marketplace enrollment process would meaningfully increase enrollment. In closing, we offer some thoughts on a couple of key tradeoffs that policymakers would need to consider in deciding whether to move ahead.

The first key tradeoff relates to effects on enrollees themselves. While the enrollees gaining (or retaining) coverage under these policy changes would, by design, owe no premiums during the year, they could be required to pay back some of their premium subsidies when they file their taxes if their income turns out to be higher than what was on file with the Marketplace. Thus, the tradeoffs these enrollees face is not completely trivial. In practice, however, expected repayment obligations would likely be dwarfed by the expected benefits of coverage for financial protectionaccess to care, and health outcomes, particularly in light of the caps on repayment obligations that exist for low- and moderate-income enrollees. Thus, almost all enrollees targeted by these policies would likely be made better off, at least in expected value.

Additionally, the tradeoffs these policies present are very similar to tradeoffs that policymakers already accept in related settings. The small premiums policy would make a near-zero premium functionally equivalent to a $0 premium; keeping enrollees with $1 premiums enrolled until they affirmatively cancel their coverage (rather than permitting insurers to terminate enrollees for premium nonpayment) is not meaningfully different from keeping enrollees with $0 premiums enrolled until they affirmatively cancel. Similarly, automatic re-enrollment is already standard practice in the Marketplace; unless Marketplace enrollees take contrary action during open enrollment, they are generally renewed into their insurance plans using the most recent available information on household income, subject to the same tax credit repayment obligations if their income ends up being higher than anticipated.

Nevertheless, it would be important to communicate clearly with enrollees about their status. In the case of the small premiums policy, that would happen automatically, as enrollees would continue to receive bills and other communications from their insurer, per usual. With automatic retention and automatic enrollment, it would be incumbent upon the Marketplace to proactively communicate with enrollees. “Opt-in” approaches to automatic enrollment offer an opportunity to highlight repayment obligations during the consent process and thereby could help ameliorate any concerns.[4]

The second key tradeoff is a fiscal one. All of these policies would increase federal subsidy costs since the increase in enrollment would occur among enrollees eligible for large premium tax credits. Policymakers would need decide whether the improved financial protectionaccess to care, and health outcomes that came with expanded coverage was worth this fiscal cost. Of course, this tradeoff is not unique to the enrollment policies we consider here; the recent premium tax credit expansion and the Affordable Care Act’s coverage provisions writ large present fundamentally similar tradeoffs.

One potential issue that cuts across both of these tradeoffs is whether some of the additional Marketplace enrollment spurred by this policy would duplicate other coverage. As an empirical matter, this may not be an important consideration; the research on auto-retention policies cited above found that coverage duplication was rare in the auto-retained population. Moreover, the fiscal cost of duplicated coverage is likely much smaller than it appears at first blush. To the extent that enrollees do not use their Marketplace coverage because they hold other coverage, the associated reduction in claims spending will ultimately translate into lower premiums, reducing federal tax credit costs. Similarly, in cases where enrollees with other coverage do use their Marketplace coverage, but the enrollee’s other coverage is through a public program, the net fiscal cost of duplication may also be relatively small.

A final consideration is how policy changes like these would be received by industry stakeholders, especially health insurers. We suspect that the reception would be (or, at least, should be) neutral or positive, as these policy changes should make insurers better off. In the long run, when insurers can adjust premiums adjust to match their costs, higher aggregate enrollment should translate into higher aggregate profits. Insurers that offer zero-premium silver plans would presumably see larger enrollment gains than their peers, but none should experience reduced enrollment, and different insurers would likely cycle through this prime position over time. This policy may be even more attractive to insurers in the short run, before premiums can fully adjust, since both economic theory and the empirical evidence discussed above suggest that the enrollees who gained (or retained) coverage would have below-average claims costs (although this effect could be offset in whole or in part by the fact that these enrollees would be disproportionately likely to qualify for generous cost-sharing reductions). Notably, even insurers who did not receive much additional enrollment could benefit since a portion of any reduction in market-wide claims risk would be shared across insurers via the risk adjustment system.


[1] State-based Marketplaces could choose to encourage or enforce similar policies through their carrier agreements, whether or not CMS implements a requirement at the federal level.

[2] We assume here that coverage under the new plan would begin in the first month after the end of the grace period. As a result, these enrollees would be uninsured for the second two months of the grace period, although they would have retained the option to resume premium payments during those months and, thus, retained some ability to protect themselves against major health shocks during those months. As an alternative, policymakers could explore making coverage under the new plan retroactive to the beginning of the second month of the grace period or, alternatively, transitioning enrollees to the new plan before the grace period expires.

[3] While this policy could increase coverage in any state, it may have especially high returns for states with state-based Marketplaces that opt to integrate the Medicaid/CHIP and Marketplace application processes, like Massachusetts. In these instances, applicants can be routed to either Medicaid/CHIP or Marketplace coverage, as appropriate, so a larger fraction of applicants will have an actionable zero-premium option.

[4] For example, the Massachusetts Health Connector application includes the following statement alongside the automatic enrollment opt-in checkbox: “I understand that I may have to repay some or all of those premium tax credits if my income is higher than what I reported to the Health Connector in the application or if I gain access to or enroll in other coverage during the year and do not report it to the Health Connector.”