Indexing provision in HHS proposed marketplace regulations is not just bad policy, but could be vulnerable to legal challenge

Dr. Benjamin Hoffman (L) speaks with Nancy Minoui about 9 month old Marion Burgess, who suffers from a chronic heart condition, at an appointment at the Dornbecher Children's hospital in Portland, Oregon, U.S. December 6, 2017. Picture taken December 6, 2017. REUTERS/Natalie Behring - RC14A7754950
Editor's note:

This analysis is part of the USC-Brookings Schaeffer Initiative for Health Policy, which is a partnership between the Center for Health Policy 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.


In January, the Trump Administration issued a proposed rule that would, among other changes, revise the methodology the Administration uses to index parameters of several provisions of the Affordable Care Act (ACA). Other analysts have raised concerns about the policy effects of this change, notably that it would reduce the value of the premium tax credit and raise the statutory cap on out-of-pocket spending for people with private insurance. This piece summarizes these policy concerns and then examines whether this change may be vulnerable to legal challenge, especially as it applies to the premium tax credit.

In brief, the proposal reinterprets a statutory provision that adjusts individual contributions for those receiving the premium tax credit based on health insurance “premium growth” since 2013. The statute does not define “premium growth,” which gives the Administration some flexibility in deciding how it should be measured.  Nonetheless, the Administration’s discretion is limited by the statutory language and rules of administrative law.

The proposal suffers from at least three potential legal deficiencies: the statutory analysis, the justification for the change proffered by the proposed rule, and the procedural approach.

First, the proposal is arguably a worse interpretation of the statutory language than is the existing rule.  While Congress does not specify a measure for premium growth, its approach elsewhere provides an indication of what is likely intended.  Indexing adjustments in the Tax Code generally measure price growth by observing the price over time of a fixed “market basket” of goods. This ensures that any apparent price inflation does not instead reflect changes in what is being sold.

Consistent with this approach, premium growth for purposes of these ACA provisions has historically been measured by examining premiums for employer coverage, which has not experienced major changes since 2013. The proposed rule would change course and consider individual market premiums as well.  Because individual market coverage changed radically between 2013 and 2014, with coverage more comprehensive and a very different population covered, comparing prices before and after those changes does not provide meaningful information about the change in the price of similar coverage – it is like comparing the price of apples in 2013 to the price of oranges today.

Such an approach is arguably not a permissible interpretation of the statutory instruction to measure premium growth.  The proposed rule includes language that seems to concede this point, noting that basing the indexing adjustment on individual market coverage in 2013 would likely “skew” the results.

Second, the proposed rule fails to articulate a clear and consistent justification for the change.  As noted above, it concedes that comparisons of premiums in 2013 and subsequent years that include individual market premiums are likely to be “skewed” by these market changes, even as it proposes to do exactly that. The proposed rule also identifies as a “benefit” of the change that it would reduce health insurance coverage and the affordability of health care. These effects are in clear conflict with the purposes of the statute and raise questions about the Administration’s rationale.

Finally, the Administration’s proposal has a potential procedural vulnerability. The proposed rule is under the jurisdiction of the Department of Health and Human Services (HHS), but the premium tax credit is under the jurisdiction of the Treasury Department. The proposed rule suggests that Treasury and the Internal Revenue Service (IRS) will apply this change to the premium tax credit using guidance that does not go through the normal notice-and-comment rulemaking process. Guidance promulgated in this fashion does not receive the judicial deference afforded to notice-and-comment rulemaking and therefore would face a higher bar if challenged in court.

Background on the Proposed Changes

The premium tax credit is calculated so that an individual would pay an “applicable percentage” of her income towards benchmark coverage, with the remainder of the premium for benchmark coverage covered by the tax credit. Thus, larger applicable percentages mean lower tax credits and higher net premiums for individuals. The applicable percentages vary by income so that lower-income individuals pay a smaller share of their income towards coverage. In 2014, they ranged on a sliding scale from 2 percent up to 9.5 percent.

For years after 2014, the applicable percentages are adjusted based on “premium growth,” as determined by the Treasury Department.[1] For example, the 2015 applicable percentages are adjusted based on the growth in premiums between 2013 and 2014.  The same indexing rule governs the affordability threshold for employer-sponsored coverage. A similar provision, administered by HHS, governs how the level at which insurers are required to cap enrollees’ out-of-pocket costs changes over time.

In 2014, HHS and Treasury issued regulations and guidance interpreting these indexing provisions. They defined premium change based on per-enrollee premiums for employer-sponsored coverage.

The proposal revises the definition of “premium growth” to also include individual market coverage and proposes to apply that change all the way back to 2013.  Revising the indexing methodology in this way would increase applicable percentages and the affordability percentage for 2020 and later years.

Policy Effects

As detailed in a recent piece by the Center on Budget and Policy Priorities, the proposal would reduce tax credits for millions of consumers. For example, a family of four with annual income of $90,000 would pay $220 more for their coverage. (The effect would be smaller for tax credit recipients with lower incomes.) These changes would also mean more people would be considered to have an “affordable” offer of employer coverage and therefore would be ineligible for the premium tax credit.

These changes would reduce the overall affordability of coverage and the number of people covered. By HHS’s estimates, the changes would reduce spending on the premium tax credit by about $1 billion in 2020 and result in about 100,000 fewer people having coverage through the Marketplace, with most of those ending up uninsured. As HHS notes, the proposal could also modestly increase premiums for those not receiving financial assistance since healthier individuals would be more likely to drop coverage as the premium tax credit shrinks.

Legal Concerns

Reactions to the proposed rule have focused on the policy impact.  But there are also potential legal concerns related to the statutory analysis, the rationale provided, and the procedural approach.

The Proposal Appears to Provide a Worse Interpretation of the Statute than Current Policy

HHS claims that the change would “improve accuracy” in the indexing calculation, but it is arguably a less accurate measure of the quantity described in the statute.

Most indexing of tax parameters uses the consumer price index (CPI), a broad measure of price growth. The CPI is developed by looking at the price change for a fixed set of goods, referred to as the “market basket.”  This ensures that any price changes reflect actual inflation rather than just changes in what is being sold. It seems likely that Congress intended premium growth to be measured in a similar way, by comparing the price of a relatively stable good across time.

Including individual market premiums in a measure of premium growth fails this standard.  The individual market changed significantly between 2013 and 2014 as major ACA market rules took effect. Pre-2014 individual market plans frequently excluded coverage for many services, did not cover pre-existing conditions, and charged higher prices or denied coverage to sicker individuals. Post-2014, individual market plans can do none of those things: they are required to cover essential health benefits without any exclusions for pre-existing conditions, to charge the same premiums regardless of health status (community rating), and to offer coverage to all (guaranteed issue). These policy changes drove substantial increases in premiums, changes which continued to unfold through at least 2017 as the market adjusted to the new rules.

Later changes to individual market rules created further discontinuity.  For example, the phasedown of the ACA’s transitional reinsurance program and the expiration (and prior funding shortfall) of its risk corridor program put upward pressure on premium growth in 2015, 2016, and 2017. And in 2018 individual market premiums began to incorporate the cost of cost-sharing reductions.

The result is that comparing the price of individual market coverage between 2013 and later years does not provide a meaningful measure of price growth – it is like comparing the price of apples in 2013 to oranges this year. As explained in greater detail below, HHS seems to admit this weakness in the proposed rule, noting that including individual market data back to 2013 would provide “skewed” results.

By contrast, employer-sponsored health insurance has not experienced major market changes since 2013. The most impactful rules affecting the individual market, like the essential health benefit requirement, community rating, and guaranteed issue, apply only to small group plans and do not appear to have had nearly as significant effects on that market. Many of the ACA regulations applicable to all employer plans went into effect before 2014, including the restrictions on lifetime and annual limits, the requirement to cover preventive care without cost-sharing, the ability for children under age 26 to enroll in their parents’ plans, and medical loss ratio rules. And the changes affecting all employer plans that did take effect in 2014 are far less significant than those affecting the individual market. For example, most plans already met the standards required to avoid penalties under the ACA’s employer mandate. Perhaps the most impactful 2014 provision was the requirement that all employer plans include an out-of-pocket maximum – a far less transformative change than changes like implementing guaranteed issue and community rating.

In keeping with this relative regulatory continuity, the salient characteristics of employer-sponsored coverage stayed generally stable after 2013.  Actuarial values were little changed. Premiums continued to grow at about their pre-2014 rate. And the share of individuals with employer coverage remained steady, meaning that the risk pool likely was stable. To be sure, employer coverage has not been literally unchanged over this period. Regulatory changes surely had some effect, and the market has continued to evolve over time based on labor market trends, changes in medical care, and other factors. But employer coverage did not experience the major discontinuities experienced by the individual market after 2013 and therefore provide meaningful information about price changes.

Rebasing the measure of premium growth to look at individual market coverage, then, does not “improve accuracy.” It clouds a reasonable measure with one that is less meaningful and less well suited to the statutory purpose.

Deficiencies in the Administration’s Rationale for the Proposed Change

Another important legal question about the proposed rule is whether it has put forward an adequate rationale for the change. The Administration’s rationale has important weaknesses.

Proposed Rule Concedes that Including Individual Market Premiums Could Create “Skewed” Comparisons. As noted above, the proposed rule includes language conceding that including the individual market when comparing premiums in 2013 to later years could distort the results. In summarizing the existing regulations, HHS writes,

We [previously] chose employer-sponsored insurance premiums because they reflected trends in health care costs without being skewed by individual market premium fluctuations resulting from the early years of implementation of the PPACA market reforms. We adopted this methodology in subsequent Payment Notices for 2016 through 2019, but noted in the 2015 Payment Notice that we may propose to change our methodology after the initial years of implementation of the market reforms, once the premium trend is more stable.

This line of reasoning suggests a possible justification for using a measure of premium growth that includes individual market data in future years once the individual market has fully adjusted to the new policy regime.  But this is not what the proposal does.  Rather, it would use a measure including individual market premiums all the way back to 2013.  As a result, it would fully incorporate the effects of “premium fluctuations resulting from the early years of implementation of the PPACA market reforms.”  HHS does not address this disconnect or the apparent indictment of its own proposal.

Proposal’s Consequences Conflict with Clear Purposes of the Affected Provisions. The proposed rule includes language explaining that the indexing change would make coverage less generous and less widely available and lead to less utilization of health care services. HHS describes these effects as a “benefit” of the change in its “regulatory impact analysis,” which lists pros and cons of the proposed rule.

In particular, the proposed rule claims that a benefit of the indexing proposal is the “potential reduction in economic distortions, and improvement in economic efficiency as a result of the reduction in Exchange enrollment due to the change in the method of calculating the premium adjustment percentage.” HHS later elaborates that the:

transition [to being without Marketplace coverage] may result in greater exposure to health care costs, which previous research suggests reduces utilization of health care services.  Economic distortions may be reduced, and economic efficiency and social benefits improved, because these individuals will be bearing a larger share of the costs of their own health care consumption…

In other words, HHS claims that a benefit of the change is that more people will be uninsured (or underinsured) and so will utilize less health care services.

These consequences are in clear conflict with the clear purposes of the premium tax credit, the out-of-pocket maximum requirement, and the ACA more broadly: to increase health insurance coverage and the affordability of health care. The fact that the Administration describes these effects as a benefit suggests that the Administration is intending to achieve these effects. The fact that the effects of this change and the Administration’s apparent intent in making this change conflict with the purposes of the statute could complicate efforts to justify the Administration’s purported rationale for the change.

Procedural Shortcomings of the Administration’s Approach

The proposal would be especially vulnerable to challenge if the Administration continues past practice of applying its indexing methodology to the premium tax credit using guidance that does not go through the normal notice-and-comment rulemaking process.

Analyses of the proposed regulations – including HHS’s own regulatory impact analysis accompanying the proposed regulations – have generally treated the proposal as changing the indexing for the premium tax credit.

But this is not the case.  Rather, as noted in the preamble, the proposed rule technically changes the indexing rules on the out-of-pocket maximum requirement and other provisions under HHS authority but has no impact on the premium tax credit.  That is because Treasury has authority over the indexing of the applicable percentages.  Historically, Treasury has followed HHS’s indexing methodology. The proposed rule notes that Treasury and IRS are “expected to” promulgate their own guidance doing that here.

To date, Treasury and IRS have adopted the HHS indexing methodology only in subregulatory guidancethe regulations on premium tax credit indexing merely restate the statutory language.  Subregulatory guidance does not receive the judicial deference afforded to notice-and-comment rulemaking.  Accordingly, if Treasury continues its long practice of specifying the measure used only in subregulatory guidance, a legal challenge would face a lower bar in overturning the policy.[2]


Analysts have raised important policy concerns about the indexing proposal and its impact on insurance coverage and the affordability of health care. But the proposal is also legally suspect.  The proposal would have effects that are hard to reconcile with the statutory purpose, and the justifications HHS offers do not withstand scrutiny.  If HHS does finalize the proposal as it applies to provisions under HHS jurisdiction, Treasury and IRS should think twice about pursuing a questionable policy that could be difficult to defend in court.

[1] Specifically, the applicable percentages are “adjusted to reflect the excess of the rate of premium growth for the preceding calendar year over the rate of income growth for the preceding calendar year.”  The measure of income growth used is not at issue in the proposed rule.[2] Treasury might conceivably argue that HHS’s discussion of this proposal as it applies to the premium tax credit (PTC) is sufficient to comply with the requirement for notice and comment and so confer substantial deference.  This argument would probably not prevail for several reasons: (1) The proposed rule does not actually request comments on the indexing rule’s application to the PTC, even as it explicitly requests comments on the indexing proposal in other contexts; (2) Even if the proposed rule did request comments on the indexing proposal as it applies to the PTC, that would not constitute sufficient notice of the need to comment on the PTC proposal given that the PTC is not among the statutory provisions the proposed rule purports to implement; and (3) It is not clear that subregulatory guidance would receive judicial deference even if the two previous conditions were met.Jason Levitis is a Nonresident Senior Fellow at Yale Law School’s Solomon Center for Health Law and Policy. He is also a principal at Levitis Strategies LLC, where he provides technical assistance and strategic advice on the ACA’s tax provisions, state innovation waivers, tax administration, and regulatory process to state officials nationwide through his work at Princeton University’s State Health and Value Strategies project.

Other than the aforementioned, the author did not receive financial support from any firm or person for this article or from any firm or person with a financial or political interest in this article. They are currently not an officer, director, or board member of any organization with an interest in this article.