As I’ve written repeatedly before, commenting on proposed federal rules is an act of active and engaged citizenship. If you have specific knowledge about something that the federal government wants to change rules and regulations on, comment. Meaningful, evidence based and thoughtful comments have to be addressed by the federal government. Most of the time, the comment will be addressed as “We thought about it, don’t think this is a good use of our trade-offs and time but we thought about it as we’re going to do what we said we wanted to do….” but sometimes you can get meaningful policy change as the reaction. And if the federal government completely blows off the issues that a comment letter raises, this places the rule at substantial risk in litigation.

The Centers for Medicare and Medicaid Services (CMS) released a proposed ACA rule with comments due on April 11th. The rule is motivated by a desire to make the markets more attractive for non-subsidized buyers and theoretically crack down on fraud by dramatically increasing administrative burden. There are a lot of assumptions being made. In a joint comment letter with Dr. Coleman Drake, we highlight a few areas of concern. Most notably, most of the policy actions being proposed will make the ACA markets substantially more morbid/increase premiums which goes totally against the declared policy imperative to drive down premiums for the non-subsidized. This is a HUHHHHHHH?

10 pages of thoughts below the fold:

April 8, 2025
Dear Secretary Kennedy,
We write in response to the proposed program integrity rule, CMS-9884-P. Dr. Drake is an associate professor in the Department of Health Policy Management at the University of Pittsburgh, and Dr. Anderson is an assistant professor in the Department of Health Services, Policy and Management at the University of South Carolina. Drs. Drake and Anderson are leading academic experts on the Health Insurance Marketplaces, having published over two dozen peer-reviewed articles on this topic in leading journals such as Journal of Public Economics, Journal of Health Economics, Health Affairs, JAMA Internal Medicine and JAMA Health Forum.

We wish to highlight some recent and important work from the academic, peer-reviewed literature that should inform the trade-offs and assumptions that are being made in this proposed rule. We will discuss several topics listed below, including:

• Enrollment trade-offs as affordability changes
• Ordeals and selection markets
• The incidence of administrative burden in selection markets
• The role of automatic re-enrollment to maintain continuity of coverage
• Open Enrollment Periods
• Automatic re-enrollment hierarchy

In the proposed rule, the Centers for Medicare and Medicaid Services (CMS) has prioritized improving affordability and competition in the non-subsidized segment of the ACA-compliant individual health insurance market. We believe that CMS should fully engage with the strong evidence that increasing administrative burdens for Marketplace enrollees will result in a smaller, sicker pool of enrollees that will necessitate insurers increase premiums, both on and off the Marketplaces. The proposed rule will increase administrative burdens by increasing data verification requirements, reducing the length of the Open Enrollment Period, and making it harder for current enrollees to automatically renew their coverage.

We believe that CMS is not fully considering the disenrollment effects of changes in the minimum cost of coverage for both subsidized and non-subsidized individuals, as the changes in enrollment (750,000-2,000,000) are attributed to improper enrollments.

We believe the majority of improper enrollments would disenroll from coverage as a result of the enhanced subsidies, therefore, we assume a range of approximately 750,000 to 2,000,000 fewer individuals would enroll in QHP coverage in 2026 as a result of the proposals in this rule, if finalized jointly and as proposed. We seek comment on this estimate and assumptions.
Increasing the minimum cost of coverage while also increasing administrative burden will lead to substantially higher disenrollments.

Enrollment trade-offs as affordability changes

The ACA individual health insurance market is a bifurcated market where some individuals are eligible for and receive income-based premium tax credits to reduce the cost of their monthly premiums, and other individuals do not receive these credits. The price-linked nature of Marketplace subsidies produces counter-intuitive dynamics that must be considered. 1
Prior research has shown the minimum cost of coverage is, by far, the single most important determinant of whether people enroll in Marketplace coverage.2,3 Potential enrollees are not sensitive to benefit design in choosing whether to enroll.4 Larger subsidies lead to lower premiums for subsidized individuals, in turn increasing enrollment and reducing the average costs of covering enrollees (i.e., marginal enrollees tend to be healthier than other enrollees).5 Treasure et al6 identified that CSR-94 Silver plan enrollees who faced a $22 monthly premium for the benchmark plan and may have been exposed to zero premium plans during the 2018 plan year had similar per member per month spending to Bronze enrollees who likely selected low premium and high cost sharing plans as a matter of their risk profile. Healthy potential enrollees are only purchasing coverage when the minimum cost of coverage is low.

Federal and state policies can and do change the minimum cost of coverage and thus enrollment levels in the marketplaces. Anderson, Golberstein and Drake provide new evidence that Section 1332 reinsurance waivers have a substantial enrollment trade-off.7 They analyzed both enrollment and the cost of the least expensive plans available to enrollees in Georgia at various income and subsidy eligibility levels both before and after the implementation of the Section 1332 reinsurance waiver relative to bordering states. The reinsurance waiver was successful in reducing the minimal cost of coverage for individuals who were unlikely to receive a premium subsidy as intended. However, the minimum cost of coverage increased for enrollees with incomes between 200-400% FPL leading to disenrollments among this group eight times as large as the state projected non-subsidized enrollee growth would be. Small changes in the minimum cost of coverage for unsubsidized enrollees likely will have large changes in enrollment for subsidized buyers who are most likely to be healthier and lower in costs than the average enrollee.

Actions which increase the minimum cost of coverage for subsidized enrollees, such as increasing the allowable de minimas variation for the benchmark silver plan, will likely affect disproportionately large enrollment losses among subsidized enrollees relative to any plausible enrollment gains among non-subsidized enrollees. Insurers will likely compete on price and lower the actuarial value to the new proposed boundaries:

We propose to change the de minimis ranges at § 156.140(c) beginning in PY 2026 to +2/-4 percentage points for all individual and small group market plans subject to the AV requirements under the EHB package, other than for expanded bronze plans…
We also propose to revise §156.200(b)(3) to remove from the conditions of QHP certification the de minimis range of +2/0 percentage points for individual market silver QHPs. We also propose to amend the definition of “de minimis variation for a silver plan variation” in § 156.400 to specify a de minimis range of +1/-1 percentage points for income-based silver CSR plan variations.

Decreasing the actuarial value of silver plans of the -00, -01, -04, -05, and -06 plan variants will lead to lower benchmark premiums and higher minimum cost of coverage as the premium spreads will be compressed.5 Furthermore, as discussed below, enrollees that drop coverage are likely to have lower than average risks and costs relative to the remaining enrollees, leading to higher premiums and higher per enrollee premium tax credit expenditures.

CMS projects that changing the de minimas variation rule will lead to a 1% decrease in premiums. Assuming this 1% change in premium applies evenly to all metal levels, we anticipate that the minimum cost of coverage for a subsidized enrollee will increase by $6.00 to $7.00 per member per month for individuals who do not have the option to purchase zero-premium Bronze plans. This will lead to a substantial decrease in enrollment that is not being considered by CMS in the proposed rule.

Administrative Burden, Ordeals and Selection Markets
We would like to bring new work by Shepherd and Wagner recently published in the American Economics Journal to the attention of CMS.8 Classic public economics thinking has focused on using ordeals to impose non-cash prices on potential enrollees in order to more effectively target public assistance to individuals who have the highest valuation of that assistance.9 In health insurance markets, however, individuals who likely have lower value on assistance are also likely to have far lower costs than individuals who highly value the assistance. In such a market that is vulnerable to adverse selection, ordeals in the form of increased administrative burdens and compliance costs could plausibly lead to increased average morbidity and higher average claims expenses as the lowest acuity, lowest risk enrollees would be the most likely to drop coverage in response to increases in price and/or administrative burdens.
Shepherd and Wagner examine automatic re-enrollment. They note that there is substantial variation in expected monthly costs as a function of risk:

In our health insurance data, the highest-risk (sickest) 10 percent of enrollees incur 15 times higher medical costs than the healthiest 10 percent (about $1,400 versus $90 per month). Moreover, the healthy are likely to value insurance less, precisely because they have fewer medical needs and use less care. This example illustrates the key correlation in settings with adverse selection: low-value types also tend to be low-cost.

The key point is that increasing burden in a health insurance selection market likely drives out the lowest risk and lowest cost enrollees, leading to higher premiums for the remaining enrollees. As they elaborate:

We use a natural experiment to study descriptively how much ordeals matter for take-up and which types of people they screen out. We find that even minor hassles lead to major reductions in take-up among an otherwise uninsured low-income population. Consistent with adverse selection, the excluded group is differentially younger, healthier, and poorer, suggesting ordeals screen out people with low private value (demand) but also low cost of insurance.1 Using an empirical model estimated with our data, we find that ordeals worsen targeting efficiency, despite successfully screening out low-value types. More generally, we show that adverse selection works alongside behavioral frictions to weaken the (revealed preference) link between demand and efficiency that is key to self-targeting. This makes ordeals relatively poorly suited tools for adverse selection markets.

his insight, which reinforces other recent research by Domurat, Menashe and Yin,10 informs the rest of our comments.

Automatic re-enrollment and zero premium plans
The agency has proposed a substantial change in automatic re-enrollment policy for individuals passively re-enrolling in zero-premium plans.
We propose that, when an enrollee does not contact an Exchange to obtain an updated eligibility determination and select a plan on or before the last day to do so for January 1 coverage, in accordance with the effective dates specified in §§155.410(f) and 155.420(b), as applicable, and the enrollee’s portion of the premium for the entire policy would be zero dollars after application of APTC through the Exchange’s annual redetermination process, all Exchanges must decrease the amount of the APTC applied to the policy such that the remaining monthly premium owed by the enrollee for the entire policy equals $5 for the first month and for every following month.

Automatic re-enrollment is critical for maintaining continuity of coverage. Drake and Anderson showed that individuals who were enrolled in December and who were not passively re-enrolled into plans—that is, they had to actively reenroll in coverage in a manner consistent with what CMS proposes here—were 30 percentage points less likely to reenroll in coverage than individuals who could passively re-enroll.11 Mcintyre and colleagues have shown that small premiums, like the proposed $5 premium penalty, act as substantial barriers to re-enrollment. They found that individuals who previously had zero dollar premiums but were defaulted to small premiums of less than $10 per month had a 14% reduction in re-enrollment.12 In other work, McIntyre and colleagues found that the individuals who were most likely to be negatively impacted by administrative burden on re-enrollment are:13

Switchers are younger (by 4.1 years), less likely to have a chronic illness (by 3.4% points, or 6%), and have lower medical risk scores (by 0.025, or 2.5% lower predicted spending). Their average medical spending per month enrolled is 8.6% lower. Notably, the larger percentage gap in spending than risk score indicates that switchers are differentially profitable even after risk adjustment. Spending for auto-switchers is particularly low in the six months following the auto-switch, consistent with research showing that enrollees lapse at times when they use less health care (Diamond et al., 2020).”

Increasing the administrative burdens of individuals who are enrolled in zero dollar premium plans likely leads to a more morbid and higher cost risk pool because these burdens create strong adverse selection incentives. Increasing adverse selection runs counter to the stated goals of the agency in this proposed rule.

Open Enrollment Period Changes to December 15th
The proposed rule indicates a desire to change the open enrollment period:

We propose to amend § 155.410(e), which provides the dates for the annual individual market Exchange OEP in which qualified individuals and enrollees may apply for or change coverage in a QHP. Specifically, we propose to add § 155.410(e)(5) and (f)(4) to change the OEP for benefit years starting January 1, 2026, and beyond so that it begins on November 1 and runs through December 15 of the calendar year preceding the benefit year and to set an effective date of January 1 for QHP selections received by the Exchange on or before this December 15 OEP end date. The Exchange OEP is extended by cross-reference to non-grandfathered individual health insurance coverage, both inside and outside of an Exchange, under the guaranteed availability regulations at § 147.104(b)(1)(ii). We also are making conforming revisions to § 155.410(e)(4) and (f)(3).

Drake and Anderson identified that counties which used Healthcare.gov from 2015-2018 where low income enrollees (175% FPL) were exposed to zero premium plans had 14% higher enrollment.14 In follow-on work using individual level data from Colorado’s state based marketplace and a rigorous regression discontinuity design (RDD), Drake and colleagues found that administrative burdens of small premiums were a substantial barrier to enrollment.15 Individuals that enrolled in zero premium plans were more likely to effectuate coverage by the January 1st deadline. Individuals who faced even small premiums frequently needed a second chance to correct mistakes and good faith errors that they made during the enrollment process. The availability of an Open Enrollment Period after the auto-re-enrollment process was completed likely allowed for individuals to pay their token premiums and start coverage for February 1st.

Eliminating an opportunity for individuals who desire to enroll in health insurance for January 1st but were tripped up by good faith errors and complexity likely will lead to increased adverse selection. Furthermore, the Notice for Benefit Payment Parameters for Plan Year 2026 (NBPP 2026) on p. 82336 of the Federal Register Volume 89, No 197 published on October 10, 2024, indicates that adverse selection due to partial year enrollment declines substantially for adults as evidenced by the Enrollment Duration Factors that only apply to individuals with between 1 and 6 months of enrollment and at least one HCC. NBPP 2023 removed monthly enrollment duration factors while adding the limited HCC contingent enrollment duration. Under the risk adjustment model that CMS has published, received notice and comment on, and has finalized a rule as of January 2025, 11-month enrollment, especially 11 month enrollment for the February-December span, has no additional predictable cost than 12 month enrollment.
Furthermore, we believe that the analysis offered on the impact of post-January 1st OEP is not a relevant analysis. We excerpt the analysis below:

From 2017 (the year before the end date changed to December 15) to 2021 (the last year of the December 15 end date), we found that Exchanges on the Federal Platform experienced a larger (47 percent) growth in enrollment among people who enrolled in coverage with only APTC compared to 28 percent growth among people enrolled with only APTC through State Exchanges. This suggests the change to the December 15 OEP end date did not compromise access to coverage for people selecting plans through the Exchanges on the Federal platform.

States that had always operated a state based marketplace are markedly different than states that have always used Healthcare.gov as either the FFM or FFM-SP.16 A simplistic comparison of the growth rates between states that have different trajectories is neither meaningful nor informative. SBM states had enrolled a higher proportion of eligible enrollees into either Medicaid or an ACA QHP prior to 2017.17 More simply, it is easier to have high enrollment growth in a state with high uninsurance rates in 2017 like Texas than in states like Massachusetts with low uninsurance rates.
One key difference is that for plan year 2018-2021 Silverloading, or the increasing of premiums of Silver plans to compensate insurers for the cost of providing mandatory Cost-Sharing Reduction (CSR) took effect. The incremental change in silver premiums due to Silverloading likely varies as a function of two critical state policies. First, states that have adapted a Section 1331 Basic Health Plan (Minnesota and New York during this time period) or have exceptionally high general eligibility for Medicaid for 19-64 year old adults (Washington DC) would have removed the overwhelming majority of the potential value of Silverloading from the ACA individual market.18,19 CMS recognized this dynamic in the August 24, 2018 Final Administrative Order.
Changes to Special Enrollment Periods

CMS proposes to eliminate a Special Enrollment Period for individuals with incomes between 100-150% FPL in § 155.420. We would like to highlight recently published work by Chatrath et al that found almost no change in adverse selection after a substantial change in the validation and verification of SEPs.20 We would encourage CMS to reconsider as the selection incentives are weak and this low income population is likely to have highly variant income and access to benefits.

If the goal of CMS is to increase the attractiveness and competitiveness of the non-subsidized individual health insurance markets, steps that increase administrative burden and morbidity in the subsidized portion of the market will work against that goal.

Sincerely,
Dr. David M Anderson PhD, Dr. Coleman D. Drake PhD

References
1. Jaffe S, Shepard M. Price-Linked Subsidies and Imperfect Competition in Health Insurance. Am Econ J Econ Policy. 2020 Aug;12(3):279–311.
2. Saltzman E. Demand for health insurance: Evidence from the California and Washington ACA exchanges. J Health Econ. 2019 Jan 1;63:197–222.
3. Abraham J, Drake C, Sacks DW, Simon K. Demand for health insurance marketplace plans was highly elastic in 2014–2015. Econ Lett. 2017 Oct 1;159:69–73.
4. Marquis MS, Buntin MB, Escarce JJ, Kapur K. The role of product design in consumers’ choices in the individual insurance market. Health Serv Res. 2007 Dec;42(6 Pt 1):2194–223; discussion 2294-2323.
5. Drake C, Abraham JM. Individual market health plan affordability after cost-sharing reduction subsidy cuts. Health Serv Res. 2019;54(4):730–8.
6. Treasure G, Anderson DM, Hatcher L, Makhoul AE, Johnson D, Stefan J, et al. Plan Selection, Enrollee Risk, and Health Spending on the Patient Protection and Affordable Care Act Individual Marketplaces, 2019. JAMA Netw Open. 2023 Mar 30;6(3):e234529.
7. Anderson DM, Golberstein E, Drake C. Georgia’s Reinsurance Waiver Associated With Decreased Premium Affordability And Enrollment. Health Aff (Millwood). 2024 Mar;43(3):398–407.
8. Shepard M, Wagner M. Do Ordeals Work for Selection Markets? Evidence from Health Insurance Auto-Enrollment. Am Econ Rev. 2025;115(3):772–822.
9. Madsen JK, Mikkelsen KS, Moynihan DP. Burdens, Sludge, Ordeals, Red tape, Oh My!: A User’s Guide to the Study of Frictions. Public Adm. 2022;100(2):375–93.
10. Domurat R, Menashe I, Yin W. The Role of Behavioral Frictions in Health Insurance Marketplace Enrollment and Risk: Evidence from a Field Experiment. Am Econ Rev. 2021 May;111(5):1549–74.
11. Drake C, Anderson DM. Association Between Having an Automatic Reenrollment Option and Reenrollment in the Health Insurance Marketplaces. JAMA Intern Med. 2019 Dec 1;179(12):1725–6.
12. McIntyre A, Shepard M, Layton TJ. Small Marketplace Premiums Pose Financial And Administrative Burdens: Evidence From Massachusetts, 2016–17. Health Aff (Millwood). 2024 Jan;43(1):80–90.
13. McIntyre AL, Shepard M, Wagner M. Can Automatic Retention Improve Health Insurance Market Outcomes? [Internet]. National Bureau of Economic Research; 2021 Apr [cited 2025 Mar 29]. Report No.: w28630. Available from: https://www.nber.org/papers/w28630
14. Drake C, Anderson DM. Terminating Cost-Sharing Reduction Subsidy Payments: The Impact Of Marketplace Zero-Dollar Premium Plans On Enrollment. Health Aff (Millwood). 2020 Jan 1;39(1):41–9.
15. Drake C, Anderson D, Cai ST, Sacks DW. Financial transaction costs reduce benefit take-up evidence from zero-premium health insurance plans in Colorado. J Health Econ. 2023 May 1;89:102752.
16. Krinn K, Karaca-Mandic P, Blewett LA. State-Based Marketplaces Using ‘Clearinghouse’ Plan Management Models Are Associated With Lower Premiums. Health Aff (Millwood). 2015 Jan 1;34(1):161–9.
17. Aboulafia G, Gruber J, Sommers BD. The Rise and Fall (and Rise) of the Affordable Care Act: Varying Impacts on Coverage Over Time and Place [Internet]. National Bureau of Economic Research; 2025 [cited 2025 Apr 3]. (Working Paper Series). Available from: https://www.nber.org/papers/w33615
18. Blewett LA, Anderson DM. Examining The New Basic Health Plan Financing Rule. Health Aff Forefr [Internet]. 2018 Oct 3 [cited 2025 Apr 8]; Available from: https://www.healthaffairs.org/do/10.1377/forefront.20180927.980559/full/
19. Anderson D, McIntyre A. Policy Implications Of Utah’s Proposed Limited Medicaid Expansion [Internet]. Health Affairs Forefront. Bethesda: Health Affairs Blog; 2019 [cited 2023 Dec 20]. Available from: https://www.healthaffairs.org/do/10.1377/hblog20190212.230279/full/
20. Chatrath S, Galbraith AA, Garabedian LF. Examining selection in Affordable Care Act (ACA) Marketplaces: special enrollment periods. Health Aff Sch. 2025 Apr 1;3(4):qxaf048.