April 3, 2017

Canada Shows Medicaid Block Grants Can Work But Probably Won’t

If the defeat of the American Health Care Act (ACHA) of 2017 proves anything it’s that the politics of healthcare in America have irrevocably changed. As Niskanen Center Senior Fellow Ed Dolan recently argued, the expectation of universal coverage is here to stay. The discussion we ought to be having is how to make universal coverage work while controlling costs and preserving the innovativeness of the U.S. medical system.

The place to start that discussion is Medicaid, our largest public health program by enrollment. The passage of the Affordable Care Act (ACA) gave states the option of expanding Medicaid to anyone below 133% of the poverty line. As of January 2017, 32 states including D.C. have adopted the expansion, bringing total enrollment up to 69 million, or one in five Americans. In the shadow of the failed ACA repeal, red states that once resisted the expansion like Kansas are pushing forward with adoption, as well.

The core of the defeated AHCA bill was a plan to send Medicaid back to the states, either as a block grant or pre-enrollee allotment, effectively capping the program’s growth rate. As it stands now, Medicaid is financed jointly by state and federal spending. The federal government has historically covered roughly 50 to 75% of a state’s Medicaid cost depending on the state’s match rate, and in exchange states administer the program according to federal guidelines. The matching grant under the ACA expansion was higher — 100% for every dollar a state spends declining to 90% in 2020 — in order to induce adoption.

Proposals to block grant Medicaid have two core objectives. First, block granting significantly relaxes federal guidelines, and provides states greater flexibility to design Medicaid to their specific needs. And second, without federal matching, states have more fiscal skin in the game, which in theory helps control program growth and promotes cost-saving innovation.

AHCA’s version of block granting went a step further, and set spending targets that meant many people currently covered by Medicaid would have been forced into the private market, although aided by a modified version of the ACA’s insurance premium tax credits. This latter goal of AHCA was presumably an attempt to reverse Medicaid’s enrollment growth, and revert it back into a more targeted safety-net program for the truly indigent. However, the question of Medicaid’s proper purpose is a separate issue from block granting per se, as Medicaid’s financing method and its scope of coverage are distinct. Indeed, a public insurance system can in theory both provide universal coverage and be effectively financed by block grants to states. This is essentially the system used by Canada.

Lessons from Canada on Block Granting

Universal healthcare in Canada was once a joint federal-provincial program run very much like Medicaid is today, with a 50% subsidy from the federal government to share the cost with provinces. And much like Medicaid, matching contributed to rapid spending growth in the 1970s, until 1977 when the entire federal portion of funding was converted into a per-capita block grant indexed to inflation.

The size of the block grant, today known as the Canada Health Transfer, has ebbed and flowed over the years, including some substantial one-off reductions. But if the goal was to shift the burden of healthcare spending onto provinces, it was successful. By 1990 the federal proportion of healthcare spending in Canada had declined in half to 15%, although it later crept back up to around 23%, where it sits today.

While conservatives in the United States often deride Canada’s single-payer model of health insurance for containing costs at the expense of longer wait times, there is much conservatives can admire from Canada’s successful application of fiscal federalism. Indeed, Canada’s federal government has almost no involvement in how provinces choose to run their healthcare systems, and would run up against constitutional questions if it ever overplayed its hand.

Provinces receive federal transfers by voluntarily conforming to the Canada Health Act, which sets national standards of comprehensiveness, universality, accessibility, and portability. Were Medicaid to be block granted, similar criteria would also be incredibly important for its success. For example, without clearly defining accessibility, states might try to save money by restricting access to Medicaid services in rural regions where care is more expensive. Moreover, because money that flows into a state’s treasury is fungible, it can be challenging to assess how block grants are ultimately spent. How provinces spend their share of the Canada Health Transfer is not tracked at all. Thus without clear, verifiable standards it would be easy for a state to use Medicaid dollars on other priorities.

Sara Rosenbaum of HealthAffairs.org points out that the “1995 Medicaid block grant provisions contained in the Balanced Budget Act of 1995 that was sent to (and vetoed by) President Clinton consumed 100 pages of dense legislative drafting.” In contrast, the block grant provisions in AHCA take up just eight pages. According to Rosenbaum, “The lax nature of the amendment can be seen in the lack of federal integrity controls over hundreds of billions of dollars in federal spending.”

The 1996 welfare reform provides a cautionary tale on this point. The creation of Temporary Assistance for Needy Families (TANF) turned welfare into a block grant that gives states enormous discretion in how it can be used. Its loose provisions allow states to use the funds to, among other things, “encourage the formation and maintenance of two-parent families.” This has resulted in cases where welfare dollars are being spent subsidizing middle-class marriage counselors, as just one example. If TANF were to be reformed today it would be wise to add a clear requirement that it only be spent on individuals or households below or near the poverty-line. Likewise, a Medicaid block grant should retain clear provisions on comprehensiveness, i.e. coverage standards, in order to prevent it from becoming an all-purpose healthcare fund. 

Portability is also important to bear in mind. If Medicaid’s eligibility requirements are not relatively uniform across the country it could create an additional barrier to inter-state mobility for lower income households — those one hopes have a greater willingness to move to opportunity. Relatedly, if what Medicaid covers becomes highly heterogeneous across the country, as opposed to how, it opens the possibility of renewed migration based on a kind of benefit shopping.

Spending Per Full-Benefit Enrollee, FY 2011, via the Kaiser Family Foundation

Spending Per Full-Benefit Enrollee, FY 2011, via the Kaiser Family Foundation

The Medicaid service packages offered by states already vary substantially, and usually do not transfer across state lines. Indeed, geographic disparities in Medicaid spending have long been identified as a major issue with the existing program. A major reform to Medicaid could be an opportunity to enhance reciprocity between states, but it could also make the situation much worse.

50 laboratories of Medicaid?

Block granting Medicaid is potentially an effective way to control cost, but it matters enormously how that cost minimization is achieved. Will states control cost by experimenting with more effective program designs, and early inventions that lower costs down the road, or will program innovation be secondary to reducing and eliminating services at the margin?

Canada once again provides some useful lessons. Writing in the Journal of the American Medical Association, Benjamin Sommers of the Harvard School of Public Health and his coauthors argue that,

there is little evidence that the alleged advantages of block grants have materialized in Canada. Advocates argue that with greater flexibility and proper incentives, states can reduce costs by improving the efficiency of care. In Canada, however, the provinces’ primary means of coping with budget pressures under block grants has been to reduce funding to hospitals and bargain harder with provincial medical associations. Ironically, then, if this scenario plays out in the United States, it would exacerbate one of the chief Republican criticisms of Medicaid—that it pays clinicians such low rates that they have reduced incentives to care for low-income patients.

This once again has parallels with the 1996 welfare reform. Advocates had hoped block grants would spur experimentation within states to discover the best program structure for encouraging self-sufficiency and upward mobility. Instead, states largely met participation requirements by reducing caseloads and, over time, learning to game the poorly enforced work requirement. Yet, when the program is basic health insurance, “caseload reduction” is just another word for millions of people losing coverage, which helps to explain why ACHA was ultimately rejected by moderate Republicans.

Federalism, True and False

Replicating Canada’s relative success with block grants in the United States could be difficult for another reason: the two countries have very different modes of federalism. Canada, like many parliamentary democracies, practices what is known as “executive federalism.” Premiers — the Canadian analog of governors — have extraordinary executive control over provincial policy, particularly when their parties hold legislative majorities. The effect is a legislative-executive fusion. Thus, throughout history, premiers have often converged in negotiation rooms to hammer out important things like reciprocity agreements and common terms for negotiating with federal representatives. This autonomy has helped make healthcare block grants work.

Executive federalism in the United States, in contrast, is constrained by the separation of powers at both the state and federal level. Historically, this has led to a greater reliance on waiver-based approaches to state program experimentation. The lower transaction costs of waivers allows the executive branches of both state and federal governments to effectively collaborate on the terms of program implementation without sacrificing accountability and broader programmatic goals.

In this respect, the Affordable Care Act introduced a welcome dose of executive federalism into the US healthcare system. As Columbia Law School professor Jessica Bulman-Pozen writes in her 2016 overview of executive federalism in the U.S. context,

With the Medicaid expansion rendered truly optional by the Supreme Court, and a Congress that has not responded to this holding, the federal executive has entered into a variety of compromises with state executives to achieve its overall objective of Medicaid expansion. For instance, following consultation between governors and high-up executive branch officials, including White House senior advisor Valerie Jarrett, the federal executive approved waivers for Arkansas, Indiana, Iowa, Michigan, New Hampshire, and Pennsylvania, several of which permit Medicaid expansion through private insurance policies.

In general, the ACA’s implementation generated many novel institutional arrangements across states. Of note was the creation of “1332 state innovation waivers” that allow states the ability to experiment with their market places. Within Medicaid, the section 1115 demonstration waivers allow states the ability to test new policy approaches, as well. Importantly, the ACA established a “Coordinated Waiver Process” that consolidates waiver applications, and streamlines their approval. The Department of Health and Human Services should use this existing infrastructure to accelerate steps towards bottom-up healthcare reform, and encourage states to experiment.

An Even Better Way

It is worth remembering that the innovative welfare-to-work programs that ultimately inspired the 1996 welfare reform began as waiver programs in Connecticut and Wisconsin. Federal waivers for welfare began under President Reagan, and provided a source of exogenous accountability on ideas that may or may not have worked. Had a workfare experiment yielded poor results, or enabled gaming of the system, the waivers would not have been renewed, giving states a strong incentive to get it right. It can even be argued that the decision to block grant welfare ultimately hurt state laboratories of democracy by substituting executive federalism with legislative gridlock. Indeed, despite many obvious shortcomings, TANF has not been significantly changed since its last reauthorization in 2005.

Medicaid waivers have likewise been a genuine source of program innovation. Waivers for home care, for instance, are widely seen as having caused a quiet revolution in how care is provided for individuals with developmental disabilities. Another Reagan-era initiative, states have used the waivers to move the disabled out of institutional care and into home- and community-driven service models, enhancing the autonomy of thousands. Use of the waivers has since spread to 44 states and D.C., in each case adapted to local needs. That is the way federalism is supposed to work.

In many ways, Medicaid waivers have replicated the success of Canadian block grants better than even block grants would. Not only do they import a sorely lacking parliamentary-style of executive decision making, but stronger accountability rules and audits help ensure waivers are enacted according to the standards of comprehensiveness and accessibility discussed above. For example, Medicaid waivers require that states demonstrate their alternative service models do not cost more than the default while providing comparable benefit. States can also apply for waivers based on “statewideness,” to direct Medicaid dollars into parts of the state with the greatest need.

Greater federalism could come to Medicaid through the expanded use of waivers. This could even include broad-based waivers that allow the wholesale restructuring of Medicaid, or the transference of dollars into other anti-poverty programs like the EITC, conditional on maintaining comparable standards of access and comprehensiveness. Addressing the federal match rate will still be important for insuring Medicaid’s long term sustainability and to control program growth. But without greater consensus in the halls of congress, waivers provide a way for states to reduce cost in a transparent and equitable way right now.

Why This Matters

As Medicaid takes on a growing role in the U.S. health system, reforming the program to allow for greater state flexibility and experimentation is a paramount concern. The recent failure of ACHA suggests Republican plans to convert Medicaid into an austere block grant are doomed to fail. However, a close examination of Canada’s experience with healthcare block grants suggests an expanded use of Medicaid waivers provides not only a viable compromise, but perhaps a truly superior model of federalism.

With an opioid crisis ravaging many parts of America, the recent suggestive evidence that Medicaid expansions have contributed to opioid over-prescription should give everyone pause. Finding innovative ways to respond to this epidemic is just one among many ways broader Medicaid waiver options would benefit all Americans.