This analysis is part of USC-Brookings Schaeffer Initiative on 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.
On July 27, the Senate introduced and narrowly defeated a stripped-down health reform bill, entitled the Health Care Freedom Act but widely called the “skinny repeal” bill, that would make a small number of substantial changes to the Affordable Care Act (ACA). Like earlier drafts released by Senate Republicans, the bill includes revisions to state innovation waivers under section 1332 of the ACA.
While the skinny bill was voted down, similar amendments to section 1332 may be discussed in the weeks and months ahead in the context of ongoing bipartisan efforts to improve the individual health insurance market. Changes to the rules for state innovation waivers have been a common element of such proposals to date, including the outline released on August 1 by a group of legislators calling itself the “problem solvers caucus,” and the agreement released on August 8 by an ideologically diverse group of health policy experts.
As I explained in two previous blog posts, the earlier drafts of the Senate’s repeal and replace legislation, each entitled the Better Care Reconciliation Act (BCRA), would undermine the substantive and procedural safeguards built into section 1332, allowing – indeed requiring – the approval of waivers that would undermine coverage and misuse federal taxpayer dollars.
The section 1332 language in the skinny bill makes a somewhat different set of changes, but still raises serious concerns. Like the earlier drafts, it undermines many of the procedural safeguards built into section 1332, most importantly by eliminating the federal government’s ability to revoke a waiver once approved – even if the state willfully violates the terms of the waiver. And while it leaves in place the substantive coverage protections, or “guardrails,” that earlier drafts removed, it introduces poorly drafted language that could be misread to virtually eliminate the guardrails. The new draft also shortens the timeframe for the federal government to consider waiver applications from 180 days to 45 days, weakening federal officials’ ability to ensure that waivers abide by statutory requirements.
This blog post describes the section 1332 language in the skinny bill, describes its shortcomings, and suggests changes that could mitigate these shortcomings. Even with these suggested improvements, however, the skinny bill’s waiver provisions would raise concerns. Section 1332 currently strikes a careful balance between state flexibility and substantive and procedural safeguards to protect individuals, the federal budget, and program integrity; two state waivers have already been approved under this regime and several more are in the pipeline. It is not at all clear that shifting this balance in favor of flexibility is warranted. But to the extent that changes like those in the skinny bill are enacted, they should be carefully crafted to avoid doing unnecessary damage.
Section 1332 under Current Law
Section 1332 provides broad flexibility for states to eliminate or modify many of the ACA’s central coverage provisions, including the health insurance Marketplace and related subsidies and the individual and employer mandates, and certain rules regarding which plans may be offered in the individual market. If a waiver results in a reduction of federal Marketplace subsidies flowing to the state, the state can receive the difference in “pass-through funding” to implement the state’s own plan for providing coverage.
Section 1332 balances this flexibility with substantive and procedural safeguards to prevent a waiver from harming individuals’ coverage or increasing federal deficits. The substantive safeguards, often referred to as “guardrails,” require that a waiver will not:
- Reduce the number of people with health coverage;
- Reduce the affordability of health coverage;
- Reduce the comprehensiveness of health coverage; or
- Increase the federal deficit.
Section 1332 also includes several procedural safeguards. The Secretary of Health and Human Services and Secretary of the Treasury jointly approve or deny waivers, and they have discretion to deny a waiver even if they find that it satisfies the guardrails. A waiver may be approved for no more than five years (with the possibility of renewal, subject to the same standards). Section 1332 does not specifically address waiver revocation, but the regulations finalized in 2012 provide that a waiver may be revoked if the Secretaries determine that it is not satisfying the requirements – for example, that it is costing more than expected and increasing federal deficits. In addition, before a state can apply for a waiver, it must pass legislation enacting the state’s plan to provide coverage under the waiver – an assurance of broad support within the state.
Together, these safeguards ensure that the broad flexibility available under waivers does not come at the cost of health coverage, federal budget discipline, and program integrity.
Concerns about Earlier Senate Drafts
As explained in greater detail in my previous analyses, the June 22, July 13, and July 20 drafts of the Senate bill would upset section 1332’s careful balance by weakening its substantive and procedural safeguards. In particular, these earlier drafts would have:
- Eliminated the three guardrails requiring that a waiver not adversely impact health coverage (while adding a requirement that the application include non-binding language “describing…alternative measures” aimed at enhancing coverage);
- Eliminated the Secretaries’ discretion to reject a waiver for other reasons;
- Forbidden waiver revocation, apparently for any reason;
- Extended the maximum waiver period to eight years;
- Permitted states to apply for a waiver without enacting authorizing legislation; and
- Created a $2 billion fund to incentivize states to apply for waivers.
Together, these changes would open the door to waivers that severely undermine coverage, especially for individuals with significant health care needs, and that divert federal health care dollars to other uses like state tax cuts, deficit reduction, or other programs. Moreover, such harmful or wasteful waivers could be enacted without state legislative consent, could not be rejected for any reason if they didn’t increase the deficit, and then could not be revoked for eight years, even if it quickly became clear that they caused unexpected harm or that the state was not complying with its commitments under the waiver agreement.
Skinny Bill Changes to Section 1332
The section 1332 language in the skinny bill shares many features in common with earlier drafts, but there are a few important differences. In particular, the new language:
- Leaves out the language in earlier drafts eliminating the coverage guardrails;
- Leaves out the provision in earlier drafts allowing a state’s executive branch to apply for a waiver without authorizing legislation;
- Requires the federal government to approve or reject a waiver within 45 days rather than 180 days as under current law; and
- Adds language that could be misinterpreted to provide authority to disregard all of the guardrails, as described in greater detail below.
The draft is otherwise generally the same as earlier drafts: it eliminates the Secretaries’ discretion to deny a waiver for reasons other than failing the guardrails, prohibits the Secretaries from revoking a waiver after approval, extends the waiver period from five to eight years, and provides up to $2 billion in grant funding to states that apply.
Key Concerns Raised by the Section 1332 Changes in the Skinny Bill, and Possible Mitigations
Keeping the coverage guardrails in place mitigates some of the concerns posed by earlier Senate drafts. But it does not address all of those concerns, and two of the modifications in the skinny bill create new concerns.
The most important improvement in the skinny bill is removing the provision in the earlier drafts that eliminates the coverage guardrails. While a drafting problem may undermine these benefits in practice, the apparent intent of this change is to avoid opening the door to waivers that transparently undermine health insurance coverage for residents of the state. If effective, this change would also go a long way towards preventing the diversion of federal health care dollars to other purposes, since it would be difficult if not impossible to maintain coverage and satisfy deficit neutrality while spending the pass-through funds on tax cuts or highways.
The skinny bill also leaves out the provision in the earlier drafts allowing a state’s executive branch to apply for a waiver without the enactment of authorizing legislation. This is an improvement, as it ensures broader buy-in within the state.
But otherwise the skinny bill raises the same concerns as the earlier Senate drafts, as well as some new ones:
Waivers cannot be revoked even if a state violates its promises under a waiver or the waiver manifestly fails the guardrails. Before a waiver is given final approval, the state must agree to a set of “specific terms and conditions,” which includes its plan for providing coverage under the waiver. Under current law, violating these terms and conditions is grounds for waiver revocation. But the skinny bill appears to rule out waiver revocation under any circumstances, given its clear and emphatic command that “a waiver under this section…may not be cancelled by the Secretar[ies] before the expiration of the 8-year period.” Thus, the Secretaries may be unable to revoke a waiver even if a state willfully violates its promises under the waiver (e.g., if a state cuts promised spending for its health care programs).
While it would likely be best to simply preserve the Secretaries’ current broad authority to revoke waivers, if Congress does wish to limit that authority, changes to the current language could reduce the potential for adverse outcomes. First, there is no reason to permit a state to keep a waiver if it breaks the promises it made as a condition for receiving the waiver. In such cases, the Secretaries should be required to revoke the waiver immediately, after giving the state an opportunity to correct the violation.
A second question, with a less clear-cut answer, is what should occur in instances in which the state implements the policies it promised to under the waiver but, once the waiver takes effect, it becomes clear that the guardrails are not satisfied. For example, a waiver may turn out to cost the federal government far more than expected, or a waiver may turn out to substantially increase the number of people who are uninsured.
There are plausible arguments that waivers should not be automatically revoked in such cases. A state must expend substantial resources to apply for and implement waivers and understandably wants to be able to rely on a waiver approval if it keeps its promises. Allowing revocations in these circumstances could undermine the ability of states and private actors within those states to make long-term plans.
On the other hand, prohibiting revocation could keep a waiver in place even after it becomes manifestly clear that the waiver is violating the guardrails in practice. Forecasting five or eight years in the future is inherently difficult. It seems imprudent for the federal government to lock itself into allowing a waiver of federal law and the diversion of federal dollars without any opportunity to take corrective action if the program does not perform as expected.
In light of these offsetting considerations, it may be appropriate to set a higher bar for waiver revocation in cases where the guardrails turn out to be violated than in cases where a state willfully violates the terms of the waiver. For example, Congress might require revocation after two years of material failure to satisfy the guardrails. In addition, keeping the maximum waiver period at five years instead of eight would mitigate concerns about the challenges of long-term forecasting and the risk of long-term harm.
Waivers that are bad policy for other reasons cannot be rejected. Eliminating the Secretaries’ discretion to reject a waiver for other reasons – such as substantial policy concerns unrelated to the guardrails – means that waivers that seem likely to do harm outside the guardrails’ narrow scope must be approved. For example, the Secretaries could not reject a waiver that was expected to maintain health coverage and have no impact on the federal deficit but would substantially reduce employment in the state.
The purpose of this change seems to be to prevent the Secretaries from using their discretion to reject waivers for political or ideological reasons. But this change goes much further than prohibiting rejection with questionable motives. It also prevents the rejection of waivers for legitimate policy reasons outside the narrow scope of the guardrails – raising the prospect of the Secretaries having no choice but to approve waivers that would do foreseeable harm.
Unfortunately, finding a middle-ground in this case seems more difficult. One could imagine allowing the Secretaries to deny a waiver for reasons other than the guardrails, but only if they present a reasonable case that the waiver would cause other substantial harm. But it would be difficult to devise a rule to prevent such a standard from being used to advance ideological or political goals. For example, the Secretaries could reject a waiver to create a state public option on the grounds that it would undermine the private insurance market. It is not clear who would judge whether this risk represented a legitimate case of substantial harm; courts are ill-suited to make such policy judgments.
In light of these challenges, such a middle ground is probably not practical. Given the remaining options, it seems more important to ensure that waivers “do no harm” by retaining discretion to reject harmful waivers.
Forty-five days is not enough time for the federal government to approve or reject waivers, especially given current resource levels. As discussed above, forecasting the impact of a state innovation waiver can be quite difficult and time consuming, particularly for waivers that fundamentally remake a state’s health care system. Given this reality, a 45-day time limit for approving or denying a waiver seems too short. It increases the risk of incorrectly predicting a waiver’s impact– a risk that is compounded by any limitations on revocation.
Such a short timeline also may not benefit states. Evaluating a waiver application often involves an extended back and forth between the federal government and the state to clarify the precise terms of the waiver, firm up analytical methods, and develop the terms and conditions. If there is not enough time for this process, the Secretaries may have little choice but to deny a waiver that might be approvable given sufficient time.
In light of these concerns, Congress should consider leaving the timeframe at 180 days, or perhaps at an intermediate timeframe like 120 days. Congress could also mitigate the risk of a shorter timeline by limiting it to simpler waiver determination. For example, it could require a faster determination if a waiver proposal is materially identically to one that has previously been approved.
Another option for mitigating the risk of a faster determination requirement is to ensure that the Departments have sufficient resources for such determinations. The challenge of modeling state innovation waivers is currently exacerbated by a lack of resources. For example, the Treasury Department, which is responsible for modeling deficit neutrality and pass-through funding for most waivers, has not received any additional resources to handle the section 1332 program. Additional resources would make faster decisions possible, with or without a requirement.
Because of a drafting problem, the language in the skinny bill could be misinterpreted to allow approval of waivers that completely disregard the guardrails. The skinny bill also introduces a technical drafting problem that could allow the language to be misinterpreted to provide authority to disregard all of the guardrails. This does not seem to be the intent of the drafters, as it inconsistent with both the restoration of the guardrails and the CBO score. Nonetheless, if Congress were to move ahead with this type of approach, it would be important to fix the language to eliminate the risk of serious negative consequences.
The change in question appears in the same passage, section 1332(b)(1), that governs the Secretaries’ discretion to reject a waiver that satisfies the guardrails. Under current law, that section provides (referring to each Secretary) that the “Secretary may grant a request for a waiver under only if the Secretary determines that the State plan” satisfies the four guardrails. The “only if” language clearly prohibits the approval of a waiver that does not satisfy the guardrails.
The skinny bill alters this language by changing “may” to “shall” and removing “only.” As changed, the passage provides that the “Secretary shall grant a request for a waiver under if the Secretary determines that the State plan” satisfies the four guardrails. This change seems intended merely to eliminate Secretarial discretion. Nonetheless, it could potentially be misread to mean that a waiver could be approved even if the guardrails are not met since the language is clear that the waiver must be approved if they are met, but it does not specify the outcome if they are not met. The removal of “only” could be read to suggest that Congress intended to expand the conditions for approval beyond those where the guardrails are met.
This reading would open the door to a host of harmful waivers. Notably, under this reading, the Secretaries would have discretion to approve a waiver regardless of its impact on coverage or on federal deficits. The result could be waivers that reduced health coverage or cost the federal government billions of dollars per year, bypassing Congress’s Constitutional role in setting federal budgets.
This reading is by no means required by the language in the skinny bill; indeed, it seems like a misreading of the language. In addition, the legislative history suggests no intent to vitiate the guardrails: the CBO score of the skinny bill ascribes no cost to the section 1332 provisions (beyond the $2 billion grant funding), which would not be the case if deficit neutrality could be ignored.
Nonetheless, there is no reason to accept the risk of an incorrect reading. The language can easily be fixed to accomplish the apparent purpose by revising the passage to read: the “Secretary shall grant a request for a waiver under if and only if the Secretary determines that the State plan” satisfies the guardrails. Such a change should be included if the change from “may” to “shall” is considered.
As the debate over revising the ACA moves into its next phase, there are opportunities for important improvements and also ongoing risks of undermining what the ACA has accomplished. Section 1332 is among the most powerful and far-reaching provisions in the ACA, and any changes to section 1332 should be carefully considered given the delicate balance the current waiver rules strike between flexibility and safeguards. The skinny bill appears to be an improvement over previous Senate drafts on this front, but it still raises serious concerns. If Congress is intent on shifting the balance in favor of more flexibility, it should proceed carefully to minimize the risks.
Jason A. Levitis is a Senior Fellow at the Yale Law School’s Solomon Center for Health Law and Policy. Until January of 2017, he led ACA implementation at the U.S. Treasury Department. In that capacity, he co-chaired an interagency work group charged with implementing the state innovation waiver program.
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. He is currently not an officer, director, or board member of any organization with an interest in this article.