Stuck in neutral? Federalism, policy instruments, and counter-cyclical responses to COVID-19 in the United States

ABSTRACT

Federalism plays a foundational role in structuring public expectations about how the United States will respond to the COVID-19 pandemic, as both an unprecedented public-health crisis and an economic recession. As in prior crises, state governments are expected to be primary sites of governing authority, especially when it comes to immediate public-health needs, while it is assumed that the federal government will supply critical counter-cyclical measures to stabilize the economy and make up for major revenue shortfalls in the states. Yet there are reasons to believe that these expectations will not be fulfilled, especially when it comes to the critical juncture of the COVID-19 pandemic. Though the federal government has the capacity to engage in counter-cyclical spending to stabilize the economy, existing policy instruments vary in the extent to which they leverage that capacity. This leverage, we argue, depends on how decentralized policy arrangements affect the implementation of both discretionary emergency policies as well as automatic stabilization programs such as Unemployment Insurance, Medicaid, and the Supplemental Nutrition Assistance Program. Evidence on the US response to COVID-19 to date suggests the need for major revisions in the architecture of intergovernmental fiscal policy.

KEYWORDS:

• COVID-19
• federalism
• Medicaid
• unemployment insurance
• policy instruments
• United States

Introduction

Crises, as Gourevitch (1986) reminds us, are to countries what reagents are to compounds in chemistry. However unprecedented or unpredictable, the COVID-19 pandemic has revealed much about how political-economic structures affect the capacity of government to provide for human needs in a moment of peril. This insight no doubt applies to the complex institutions and intergovernmental relationships that define federalism in the United States. The American federal system fractures both the authority and capacity policymakers must rely on to respond to pandemics. State and local officials alone possessed the authority to impose (and lift) emergency ‘stay at home’ orders. Yet not only have resources for public-health departments declined in recent years, the Trump administration failed to respond quickly to coordinate a national response to the pandemic. States in turn pursued a patchwork of public-health measures, often colored by partisan motivations. Indeed, Republican governors and state legislatures were, all else equal, slower to implement social distancing policies (Adolph, Amano, Bang-Jensen, Fullman, & Wilkerson, 2020).

The results of this fractured approach to public health in the United States have been deadly. By late-May 2020, there were approximately 32 COVID-19 deaths per 100,000 people, placing the US among the top ten national case fatality rates in the world at that point. While the majority of cases were initially concentrated in a handful of densely populated states, the virus has also spread nationwide. As of June, the counties with the highest number of confirmed cases per resident were not metropolitan centers but rural jurisdictions – some of which were home to large meatpacking plants (Johns Hopkins Coronavirus Resource Center, 2020).

States’ social distancing orders have had an unprecedented impact on economic activity in the United States. In an economy dominated by services rather than manufacturing, a decline in demand has produced unprecedented levels of unemployment. In April, the unemployment rate hit 14.7%, higher than at any time during the Great Recession. By the end of the second quarter, the Congressional Budget Office projected a 38% decline in Gross Domestic Product compared to 2019 (Congressional Budget Office (CBO), 2020). Between 2020 and 2022, state governments were projected to experience a 650 USD billion revenue shortfall, while city governments were projected to experience a 360 USD billion shortfall (Leachman, 2020; National League of Cities, 2020).

Despite these extraordinary numbers, broadly speaking, US policymakers have not revised their prior beliefs about the appropriate role of government or embraced drastic shifts in economic or health policy. The emphasis from the White House was on getting back to ‘normalcy’. Governors and state legislators – especially in Republican strongholds – faced extreme pressure from business associations, conservative pressure groups, and President Trump himself to ‘reopen’ their economies, even as case-fatality rates climbed. While the federal government is the sole entity capable of counter-cyclical spending to alleviate the pandemic’s economic effects, neither congressional leaders nor the Trump administration have proposed changes in economic policy that are capable of addressing the destructive and unpredictable trajectory of the virus. Yet, as we argue in this paper, available policy instruments and the institutions of federalism have left US policymakers relatively unprepared for addressing the pandemic’s economic fallout in an expedient manner. To an even greater degree than during the Great Recession, congressional reliance on discretionary policy instruments has arrested economic recovery. Counter-cyclical spending is now dependent on decisions taken in a context of fragmented fiscal information, an enervated intergovernmental lobby, and patchy policy implementation networks. Automatic-stabilizer programs such as Medicaid and Unemployment Insurance avoid many of these problems. Nevertheless, their fiscal and administrative dependence on state governments often leaves these programs incapable of a nimble response to economic crises.

This paper proceeds as follows. We first consider how federalism affects counter-cyclical economic policy in the United States, and how intergovernmental dynamics diverge between discretionary programs and automatic stabilizers. Next, we apply this framework to examine the implementation of discretionary emergency relief programs, as well as the operation of three automatic stabilizers (Medicaid, Unemployment Insurance, and the Supplemental Nutrition Assistance Program), during the COVID-19 pandemic. We conclude by considering the implications of our analysis for the future of pandemic economic recovery efforts in the United States as well as theories of counter-cyclical policy instrumentation.

US federalism and counter-cyclical policy instruments

Counter-cyclical policies – namely the expansion of spending and loosening of monetary policies – are a crucial component of national responses to economic crises (Armingeon, 2012; Francis, Jackson, & Owyang, 2017). The United States is no stranger to counter-cyclical policy. During the Great Recession, the United States adopted an approach to fiscal policy that was, relative to its European peers, aggressive: a stimulus package worth 5.6% of GDP spread over a period of three years (2008–10) (Blinder & Zandi, 2015). Nevertheless, the US response to the Great Recession is widely believed to have been too limited, resulting in a belated recovery (Boushey, Nunn, & Shambaugh, 2019). One reason for this is that US federalism fundamentally defines the allocation of counter-cyclical policy capacity (Advisory Commission on Intergovernmental Relations (ACIR), 1987). Most states have balanced-budget requirements (BBRs) and statutory spending limits. And all but a small handful of states have placed similar restrictions on local governments (Peck, 2014).

Hence, economic crises tend to place states in a ‘fiscal vise,’ simultaneously gripped by both increasing demand for services and dramatic revenue shortfalls (Pew Charitable Trusts, 2012). As a result, states experience intense political pressure to engage in austerity policies, which translate into major cuts in essential social services. While state governments received significant assistance from the federal government during the Great Recession, their fiscal responses to the crisis remained pro-cyclical. Relative to sub-national governments in other federations, the procyclical nature of crisis response in American states and cities remains extreme (Blöchliger et al., 2010).

The federal government – in contrast to the states – has two powerful sources of counter-cyclical capacity. First, the Federal Reserve can engage in expansionary monetary policies. Second, Congress does not face the same legal constraints on deficit spending that exist in the states, allowing for strong counter-cyclical fiscal policies. As a result, actions to stabilize the economy during a recession invariably occur at the federal level. Following the Great Recession, the United States pursued mild expansionary policies between 2008 and 2010 that were ultimately essential to economy recovery (Council of Economic Advisers, 2014).

Yet the federal government’s counter-cyclical capacity does not invariably translate into counter-cyclical effects. Instead, these effects hinge on the mix of policy instruments federal officials choose to adopt (Howlett, 2019). As Table 1 suggests, these instruments vary along two dimensions. First, instruments can have one of two general designs. Automatic stabilizers are programs designed to trigger automatic increases in budgetary outlays when the economy weakens and the opposite when it strengthens (Russek & Kowalewski, 2015). As unemployment worsens and poverty deepens, eligibility for programs like Medicaid and Unemployment Insurance is intended to expand. By contrast, discretionary programs are enacted during an economic crisis and are typically temporary in nature.

Table 1. Typology of policy instruments for economic recovery.

		Implementation
		Decentralized	Centralized
DESIGN	Automatic Stabilizers	Unemployment Insurance; Medicaid	Payroll Taxes
	Discretionary Programs	ARRA grants to state and local governments	Stimulus payments to individuals

In the United States, counter-cyclical policies tend to be discretionary (Boushey et al., 2019). During a crisis, policymakers must work on a relatively quick timeline to enact policies they believe will have the magnitude and duration appropriate to match the scale of the economic decline. Even when so-called ‘automatic stabilizers’ such as Unemployment Insurance and Medicaid ‘switch on’ during recessions, they usually require some form of discretionary expansion or recalibration. In response to the Great Recession, Congress adopted a patchwork of emergency programs. The American Recovery and Reinvestment Act (ARRA) devoted a roughly even share of outlays to tax cuts, expansions to programs such as Medicaid and unemployment benefits, and discretionary spending, including direct assistance to individuals and support for infrastructure programs. Between 2009 and 2012, Congress expanded its patchwork of relief programs, with the largest shares of fiscal support devoted to a payroll tax cut and extensions to unemployment insurance (Council of Economic Advisers, 2014). Despite the persistence of high unemployment, federal fiscal policy turned contractionary between 2011 and 2014 as emergency discretionary policies expired, representing a major departure from prior economic recoveries (Cashin, Lenney, Lutz, & Peterman, 2017).

A second dimension of counter-cyclical policies in the United States concerns the role of federalism in policy implementation. In some cases, policy implementation is highly centralized. Congress can take advantage of pre-existing tax information to send stimulus checks to every American. In other cases, policy implementation is highly decentralized. This includes several major ‘automatic stabilizer’ programs such as Medicaid and Unemployment Insurance, which function as federal-state partnerships. Additionally, it includes the vast array of discretionary programs that provide fiscal assistance to state and local governments (Conlan, Posner, & Regan, 2017).

The decentralization of counter-cyclical policy implementation to the American states interacts with discretionary programs and automatic stabilizers in distinctive ways. Where discretionary programs are concerned, federalism accentuates the challenge of coherent, comprehensive decision-making in the midst of a crisis. First, the decentralized and uncertain character of fiscal information makes it difficult for Congress to match policy to the scale of the crisis. Congress crafted ARRA before state revenue forecasts could be made to reflect fourth-quarter GDP data for 2008; after several revisions, the data suggested that the scale of economic collapse was far worse than what Congress had initially assumed (Boushey et al., 2019). Second, while counter-cyclical policy is often necessarily place-based, state and local officials lack a formal venue for intergovernmental negotiation during crises and must instead compete with other organized interests for relief payments. Finally, the implementation of discretionary policies at the state and local level depends crucially on the existence and management of highly complex intergovernmental networks (Conlan et al., 2017).

Federalism also affects the implementation of automatic stabilizers in several important ways. First, the federal financing provisions matter a great deal. Because states cannot engage in counter-cyclical spending, programs with low federal matching rates may require extensive congressional recalibration to realize counter-cyclical effects. Indeed, there currently exists no automatic trigger to increase state Medicaid matching rates in response to a recession. As a result, Congress must take discretionary action to ensure that Medicaid plays a stabilizing role during economic downturns. Second, federal and state eligibility rules create administrative hurdles to enrollment in automatic programs, which result in a limited receipt of benefits. Due to state policies related to earnings calculations and restrictions on eligibility for part-time workers, only a fraction of those eligible for Unemployment Insurance take up the benefit, even during economic downturns (Chodorow-Reich & Coglianese, 2019). The emergence of work requirements in means-tested social programs has further impeded their counter-cyclical effects. Finally, there are major variations across the states in the administrative infrastructure that supports automatic stabilizer programs. Inadequate investment in information technologies and the staffing of state agencies can yield significant wait times, ultimately limiting the take-up of benefits. As the next sections suggest, all of these dynamics have been in play during the COVID-19 pandemic.

Intergovernmental politics and discretionary policy responses to COVID-19

Following the onset of the COVID-19 pandemic and the economic fallout that ensued, the federal government leaned heavily on discretionary policies as a means of crisis response. As Table 2 shows, Congress passed four major pieces of emergency legislation, with varying levels of intergovernmental investment and support for automatic stabilizers. An initial supplemental appropriations bill (signed into law on March 6) directed some 8 USD billion dollars. 36% of the spending here was intergovernmental and included, among other things,immediate public health aid to state and local governments as well as research and development for vaccines. A second bill, the Families First Coronavirus Response Act (enacted on March 18), committed 192 USD billion in spending. Intergovernmental spending accounted for 45% of the bill. Additionally, 40% of Families First spending was directed towards expanding automatic stabilizer programs such as Medicaid and Unemployment Insurance. The largest response, however, was the 2.7 USD-trillion-dollar Coronavirus Aid, Relief, and Economic Security (CARES) Act. While CARES Act spending was targeted largely on business loans, it included additional support for automatic stabilizers (24% of total spending) as well as a 150 USD billion Coronavirus Relief Fund for state and local governments. These population-based payments were tightly restricted to costs incurred due to COVID-19 and targeted largely at the state level, with 45% of state funds set aside for local governments with a population over 500,000, a measure that excluded all but 36 large metropolitan governments across the United States from participation. While Congress extended the CARES Act’s business loan program in a subsequent piece of legislation, both Democratic and Republican leaders pushed the question of additional state and local fiscal relief down the road. In the absence of robust intergovernmental aid, the Federal Reserve created a municipal liquidity facility to stabilize state and local finances. Yet this too was restricted to government entities with investment-grade bond ratings and sufficiently large populations (counties with at least 500,000 residents and cities with at least 250,000 residents) (Federal Reserve Board of Governors, 2020).

Table 2. Summary of intergovernmental spending and automatic-stabilizer expansions in COVID-19 relief legislation.

Legislation	Total Spending ($ Billions)	% Intergovernmental (% total)	% Expansion of Automatic Stabilizers
Supplemental Appropriations Act (March 4, 2020)	$8	37.5%	0%
Families First Coronavirus Response Act (March 18, 2020)	$192	45%	39.58%
CARES Act (March 27, 2020)	$2,700	23.56%	23.89%
Paycheck Protection Program and Health Care Enhancement Act (April 24, 2020)	$733	13.64%	0%

Source: Authors’ analysis of COVIDMoneyTracker.org data.

As this brief overview suggests, the absence of automatic stabilizers meant that Congress had to take decisive action in the midst of a crisis to unleash the counter-cyclical capacity of the federal government. This entailed several major policymaking challenges related to informational uncertainty, ideological polarization, and interest-group politics. Intergovernmental dynamics also accentuated each of these challenges.

The informational context for relief legislation

Congress made decisions about economic relief under considerable uncertainty about future economic outcomes. The Bureau of Economic Analysis does not publish first-quarter GDP data until the last week of April. And while the Bureau of Labor Statistics released weekly reports, the speed of COVID-19’s effects overwhelmed state unemployment systems, creating a lag in timely information about the extent of employment losses. Further, the Congressional Budget Office (CBO) did not publicly release revised projections of the pandemic’s effects on the economy until after the passage of the CARES Act (Swagel, 2020).

The fragmentation of state-revenue forecasting capacities only added to these informational problems. While the CBO could move quickly to register economic projections for 2020, state governments were not able to revise revenue estimates to match the tempo of congressional policymaking. At the time of the CARES Act’s passage, only five states had revised revenue projections for Fiscal Year 2021 in light of the pandemic (see Figure 1). When they did emerge, the forecasts painted a stark picture: in many states, expected revenue losses far exceeded existing rainy-day funds. Illinois alone projected one-year losses amounted to nearly 4.6 USD billion dollars. Nevertheless, this sort of fiscal information was not available as Congress began to debate the contours of the CARES Act and subsequent legislative packages. In fact, only after the CARES Act passed did major intergovernmental organizations like the US Conference of Mayors release membership surveys indicating the scope of potential service cuts that would occur in the absence of a massive injection of intergovernmental revenue (Durr, 2020).

Figure 1. States with revised FY 2021 revenue projections, March 7–15 May 2020.

Source: Tabulated by authors from CBPP (2020)

Thus, as Congress debated the initial round of emergency legislation, the effects of COVID-19 on state revenues were difficult to commensurate. When the CBO did ultimately revise its economic projections on April 2nd, and again on April 24th, the outlook worsened substantially. By the end of the month, one analysis suggested that state revenues alone could face shortfalls totaling 650 USD billion between 2020 and 2022, a swifter and deeper decline than during the Great Recession (Leachman, 2020). The implications of inaction are dire. For each percentage point increase in the national unemployment rate, state governments face a total annual budget shortfall of 45 USD billion, largely due to lower state tax revenue (Fiedler & Powell, 2020).

Intergovernmental lobbying

In the absence of automatic stabilizers, motivating congressional action on these policies depends in large part on the influence of intergovernmental lobbying (Jensen, 2018). Cities and states – especially those experiencing fiscal constraints – have a strong incentive to lobby Congress (Dove, 2019). Yet when it comes to lobbying on comprehensive policy changes, recent developments in legislative politics have created barriers to the influence of cities and states on Capitol Hill.

Perhaps most importantly, the rise of partisan polarization in Congress and the centralization of policy bargaining limits intergovernmental lobbyists’ access to key decision-makers, especially during the agenda-setting stage of politics (Leckrone, 2019). Further, state and local governments––however vital their services and their contributions to the national economy––now lobby on a terrain crowded with an increasing number of powerful stakeholders. As lobbyist registrations shot up in the weeks following the pandemic, the crowded interest-group context endured. In the first quarter of 2020, lobbying expenditures by the pharmaceutical industry were five times higher than those of civil servants and public officials. Public officials did not rank among the top ten industries in terms of lobbying expenditure. Even where the initial vehicle for state and local relief––the CARES Act––was concerned, only seven of 50 states filed lobbying reports. Similarly, only 47 of 3,007 counties and 63 of more than 19,000 municipalities made similar investments (Opensecrets.org, 2020).

For their part, national associations of state and local officials did make a vigorous push for federal relief in the weeks following the emergency declaration. Yet as the list of intergovernmental support requests in the Appendix suggests, they did not always act in concert with one another. While the major intergovernmental organizations did coordinate advocacy, they did so primarily on more generic requests. This included asking for legislation to clarify that tax credits for paid sick leave were available to state and local governments as employers and a more generic call for the federal government to expand their ability to use CARES Act dollars for expenses other than those directly related to COVID-19. Yet when it came to requests for specific funding levels and aid packages, groups such as the National Governors Association, the US Conference of Mayors and the National League of Cities tended to work on their own. Individual associations did not release major requests for general relief until after the first major spike in unemployment claims.

Implementing discretionary relief programs

One of the key lessons of the Great Recession is that economic recovery occurs more quickly when relief legislation builds on existing programs and implementation networks (Conlan et al., 2017). In some respects, Congress took this lesson to heart when designing relief legislation directed at state and local governments. For example, the CARES Act devoted 25 USD billion dollars in relief for local transit systems, distributed through pre-existing formula grant programs for urban and rural areas as well as high-density states. Within several weeks of the CARES Act’s passage, transit systems began to receive these grants with a minimum of friction (Virginia Department of Rail and Public Transportation (DRPT), 2020). Similarly, the law’s Elementary and Secondary School Emergency Relief Fund leveraged existing grant formulas to quickly distribute 13.5 USD billion to state and local education agencies (Ujifusa, 2020).

In other cases, however, federal relief programs required state governments to establish new structures and processes for disbursement. Under the CARES Act, governors were allotted a 3 USD billion dollar emergency education relief fund. While Congress structured this as a flexible block grant, the added flexibility opened up new conflicts over how states should direct the funding. In Oklahoma, which qualified for nearly 40 USD million under the program, Gov. Kevin Stitt (R) advocated pushing the money into a tax-credit scholarship program of the sort proposed by US Secretary of Education Betsy DeVos. Yet the state’s Superintendent of Education, Joy Hofmeister, opposed the idea, suggesting that the block grant should be devoted to public-school students directly impacted by the pandemic (Martinez-Keel, 2020).

Perhaps the most significant implementation challenges stemmed from the size and scope of general relief efforts. The CARES Act’s Coronavirus Relief Fund (CRF) apportioned 150 USD billion to state governments, territories, counties, cities, and tribes. The CRF’s formula, however, severely limited the allocation in two ways. First, as noted above, the use of funds was restricted to necessary expenditures related to COVID-19 rather than to make up revenue shortfalls resulting from the pandemic. Second, the CRF limits local government’s access to all but a handful of large cities and counties. Local governments with populations over 500,000 can opt to receive funding directly from the Treasury Department. All other local units of government were forced to appeal to state governments for a share of funding (Driessen, 2020). As Matthew Chase, executive director of the National Association of Counties put it, this forced each below-threshold local government to go ‘hat in hand to the governor and say, “Can we have part of your allocation?”’ In the 16 rural states with no counties, cities or municipalities with populations above 500,000, there is nothing in the CRF formula to prevent the state from using the entire allocation to pay for its own COVID-related expenditures (Associated Press, 2020).

Adding to the confusion, Treasury Secretary Steve Mnuchin interpreted the legislation in several ways that ignited a conflict with national associations representing local governments. For example, while Congress instructed the Treasury Department to apply a 45% multiplier to determine how much funding a local government could receive from the CRF, the Treasury Department interpreted this as a statewide limitation of 45% of funding on all local governments in the state. Within a month, states and eligible local governments had received their share of the CARES Act funding. Yet in the face of massive revenue shortfalls and federal restrictions on the use of this funding, it was unlikely to provide the counter-cyclical stimulus most states and cities needed (Leachman, 2020).

Automatic stabilizers: the politics of fragmented authority

Many economists and policy analysts see automatic stabilizers as a means of responsibly deploying the federal government’s counter-cyclical capacities during periods of economic crisis and obviating many of the political challenges of both crafting and maintaining discretionary policy responses over time (Boushey et al., 2019). Yet while automatic stabilizers do exist in the United States, their effectiveness hinges on how fiscal and administrative responsibilities are divided between the federal and state governments, the capacity of state governments to administer these programs under crisis conditions, and the ability of both state and federal agencies to cooperate to offer relief from administrative burdens that depress the uptake of benefits. In the sections that follow, we review how these dynamics have affected the response of three major automatic stabilizers to the COVID-19 pandemic: unemployment insurance, Medicaid, and the Supplemental Nutrition Assistance Program.

Unemployment Insurance

Most rich democracies, including federal countries such as Canada, operate national unemployment insurance (UI) systems, a situation that increases their counter-cyclical capacity to stabilize the economy in periods of crisis. In the United States, however, there is no national UI system. In 1935, the Social Security Act established a federal payroll tax that created incentives for the states to implement their own UI systems while stopping short of adopting a centralized, uniform framework. As a result, today UI ‘is unique among American social programs: each of the 50 states (plus the District of Columbia, Puerto Rico, and the Virgin Islands) finances and administers its own program, but under federal guidelines and oversight.’ (Woodbury, 2014) Despite these federal guidelines, the existence of discrete UI systems in the states weakens the capacity of the United States to rapidly transform UI into a coherent and effective counter-cyclical tool in times of economic crisis.

To understand how UI in the United States deals with crises, we can break it into three tiers. Tier 1 is always in effect in each state, financed through payroll taxes, and available for up to 26 weeks. Tier 2, known as the Extended Benefits program, is designed to activate automatically in a recession and lengthen the duration of benefits by 13 weeks (or 20 weeks by state option). Finally, Tier 3 is constituted by emergency benefit extensions that Congress routinely passes during recessions (Woodbury, 2014).

The US response to COVID-19 thus depended crucially on the second and third tiers, but was also shaped by the fragmentation of eligibility criteria and benefit levels characterizing the first tier. This is true because the emergency provisions under tiers 2 and 3 reproduce existing cross-state policy disparities. For example, under EB, states must extend the duration of benefits by 50% when insured unemployment reaches specified levels (US Department of Labor (DOL), 2019) Consequently, benefit durations continue to vary across states after EB kicks in. At the same time, the large discrepancies in benefit levels that existed before the crisis remain (for an overview of these initial differences see US Department of Labor (DOL), 2020).

Regarding the last tier, the Families First Coronavirus Response Act (H.R. 6201) provided crucial support. On the UI front, this legislation provided ‘$1 billion in state funding to help fairly process and expand unemployment benefits’ in the context of the already rapid increase in the number of claims across the country (Emsellem & Evermore, 2020). The context of this emergency support was the 30 percent decline in federal UI administrative grants to the states between 1999 and 2019. According to Emsellem and Evermore (2020), these pre-COVID-19 grants were ‘barely enough for states to manage the basic staff they need to operate the program, leaving little for the upgrade and maintenance of unemployment insurance technology.’ While addressing looming administrative challenges magnified and exacerbated by the COVID-19 crisis, the Families First Coronavirus Response Act also provided extra federal funding for UI.

The more comprehensive federal legislation enacted in the early spring 2020 to address the COVID-19 crisis was the CARES Act, which featured three key temporary UI expansions. First, the Emergency Unemployment Compensation (PEUC) extended regular unemployment benefits for a period of 13 weeks. Second, the Pandemic Unemployment Assistance (PUA) offered unemployment benefits for up to 39 weeks to people such as the self-employed who were not entitled to regular UI benefits. Third, the Federal Pandemic Unemployment Compensation (FPUC) boosted by 600 USD per week both regular UI benefits as well as PUA and PEUC benefits (Gigante, Mufson, & Martin, ).

Although these emergency programs addressed many of the flaws of existing state UI systems, their implementation posed major challenges stemming in part from the limited administrative capacity of these systems, which remained a serious concern despite the additional federal support provided as part of federal emergency legislation (Goger, Loh, & George, 2020). In some states, glaring administrative challenges were compounded by political struggles over UI benefits. In Wisconsin, the legislature’s delay in waiving the one-week waiting period for UI caused the state to forgo an additional 25 USD million in support under the CARES Act (Beck, 2020). In Florida, where only five percent of UI applicants received benefits during the first month of the COVID-19 crisis due largely to administrative problems, Democrats accused Republican Governor Ron DeSantis of not using the full range of his powers to increase access to UI benefits (Fineout, 2020). Beyond Florida, in April 2020, millions of Americans struggled to access UI benefits and some states like Ohio postponed the implementation of emergency measures such as PUA until mid-May (Nielson, 2020). According to a national poll conducted in mid-April, 2020, for the Economic Policy Institute, ‘For every 10 people who said they successfully filed for unemployment benefits during the previous four weeks: Three to four additional people tried to apply but could not get through the system to make a claim. Two additional people did not try to apply because it was too difficult to do so.’ (Zipperer & Gould, 2020)

Difficulties in accessing state UI systems contributed significantly to a lag in the receipt of benefits. If the 23 million who the Bureau of Labor Statistics reported as unemployed in April of 2020 received average UI payments for the full month, the total of UI payments would have been double what was actually paid out (Nunn, Parsons, & Shambaugh, 2020). This situation clearly points to how US federalism and, more specifically, the decentralized nature of UI weakened and delayed key counter-cyclical measures in the early stages of the COVID-19 crisis.

Medicaid

Medicaid, which is a means-tested government run health insurance program, is the largest single source of health coverage in the United States. It is also the largest federal-state government program and is hugely complex. Total program spending in FY 2018 amounted to 592.7 USD billion. Of that spending 62.5% was federal government spending with states making up the remainder. There was, however, considerable variation in that distribution across the states. Reflecting the complexities of the program’s funding formulas the federal share ran from highs of 79.1% in New Mexico and 78.7% in West Virginia to lows of 49.9% in Virginia and 52.6% in Wyoming (Kaiser Family Foundation, 2020a).

Since Medicaid was established in 1965 the program has relied on a formula known as the Federal Medical Assistance Percentage (FMAP) along with a spending multiplier to determined how much states received from the federal government. The FMAP is based on a state’s per capita income. In short, the poorer the state the more generous the federal match. The FMAP floor is 50% of a state’s overall spending, with a potential maximum set at 83%. There is a multiplier effect since Medicaid is an open-ended program, with no aggregate cap on program spending. Hence, the more a state spends, the more it receives in funds from the federal government. This created an incentive structure that saw Medicaid grow well beyond the initial expectations of those who designed the program’s framework (Rose, 2013).

The ACA then added to that expansionary dynamic. In laying down a federally mandated floor for eligibility whereby everyone with an income below 138% of the poverty level would qualify for Medicaid with no further conditions attached, the ACA proposed to expand the program and add millions of Americans to Medicaid. Importantly, the law committed the federal government to covering 100% of medical spending on people newly eligible for the program under the terms of the law for three years followed by 90% of those costs thereafter.

Nevertheless, not all states have been persuaded to spend generously. Most notably, as the COVID-19 pandemic hit the country, 15 states were still not participating in the ACA’s expansion (Nebraska is slated to expand its program in October 2020). Hence, in the expansion states everyone with an income below 138% of the poverty line was covered by Medicaid, but in nearly all the non-expansion states adults over 21 in households without children were ineligible however poor they were unless they were pregnant or had a had a disability. In addition, about a dozen states were actively investigating ways of imposing work requirements on some of their Medicaid enrollees. Further, while 31 states used a presumptive eligibility standard for certain categories of Medicaid applicants and granted coverage while their applications were processed, at the start of 2020 the federal government and several states were looking to tighten their verification processes (Brooks, Roygardner, Artiga, Pham, & Dolan, 2020). States’ responses to this situation were highly variable. While virtually all states took at least one action to expand access to telehealth, for example, less than half took actions to expand eligibility, speed enrollment processes, or extend new health benefits (see Figure 2). Thus, as COVID-19 arrived in the United States, an individual’s eligibility for Medicaid and ease of access to the program varied quite significantly according to where they lived.

Figure 2. State emergency modifications to Medicaid in response to COVID-19.

Source: adapted by authors from Kaiser Family Foundation (2020b). Note: reflects changes made through 13 May 2020.

As part of its response to the crisis, the federal government did quickly increase federal funding to states for Medicaid and temporarily prevented states from imposing any new restrictive measures with regards to accessing the program. The Families First Coronavirus Response Act did increase federal funding to the states for Medicaid by increasing the FMAP payments by 6.2 percentage points. The law also stipulated that states could not remove coverage or introduce tightened eligibility requirements while the country was still enduring the public health emergency (Aron-Dine, 2020). If only temporarily, this meant that states could not introduce new work requirements that might result in individuals losing their Medicaid coverage. The law also determined that Medicaid would pay for COVID-19 tests and it allowed, though did not mandate, states to extend that testing to the uninsured. It did not, however, allow for Medicaid funds to be used to treat uninsured individuals with the COVID-19 infection (Brooks & Schneider, 2020). Overall, while the Act provided extra federal funds for Medicaid to help stabilize state budgets at a time of economic distress and prevented states from cutting Medicaid enrollees, neither it nor the subsequent CARES Act directly expanded access to Medicaid or offered additional financial help to people losing their jobs and seeking to either maintain their existing insurance through COBRA arrangements or get individual insurance through the ACA’s insurance marketplaces (Keith, 2020).

One consequence of the extraordinary job losses through April 2020 was that millions of Americans would lose their employer-sponsored health insurance (ESI). By no means all of these would qualify for Medicaid, but millions likely will. In this sense the program will fulfill its role as a counter-cyclical safety net program. Importantly, in terms of households meeting the requirements of the means test, unlike standard UI benefits, the federally funded 600 USD payments through the emergency UI package would not count as potentially disqualifying income. But the pre-existing condition of a state’s program would significantly impact how many people did qualify. A projection by the Urban Institute of what would happen in a scenario where unemployment hit 15% estimated that just over 50% of people losing their ESI in states that were participating in the Medicaid expansion would be covered by the program. That number dropped to 35% in states that had not expanded their program in response to the ACA (Garrett & Gangopadhyaya, 2020, p. 9–10) In addition to the vagaries of a state’s eligibility rules the sudden influx of new enrollees would test each state’s administrative capacity, perhaps delaying access. Further the fact that many newly eligible individuals and households may not have previously engaged with the state’s welfare and health bureaucracy might make them uncertain of whether they were eligible or unaware of how to process a claim (Garrett & Gangopadhyaya, 2020, p. 7).

Supplemental Nutrition Assistance Program

A final example of an automatic federal-state stabilizer is the Supplemental Nutrition Assistance Program (SNAP), formerly known as the Food Stamp Program. SNAP is a means-tested program; to be eligible, households must have incomes below 130% of the federal poverty line. Thus, the program, which covers about one out of every eight Americans during normal economic times, expands during recessions. Importantly, unlike Medicaid and UI, the federal government pays for 100% of SNAP benefits, while splitting the administrative costs with states. During the Great Recession, SNAP enabled about 15% of American households to purchase food.

SNAP’s allocation of fiscal and administrative authority between the federal and state governments affects the program’s responsiveness to COVID-19 in three ways. First, in contrast to Medicaid and UI, direct funding for SNAP beneficiaries is fully federalized, while administrative costs are split between the federal government and the states. This design eases the pressure on state governments during crisis periods, when revenue shortfalls emerge. Nevertheless, the federal government has not been quick to unleash its counter-cyclical capacity. In the Families First Act, Congress allowed for emergency SNAP allotments, yet the Trump administration interpreted the rule in a narrow way, limiting the relief states were allowed to provide. In the massive CARES Act package, Congress restricted its attention to administrative costs and emergency payments to states.

Second, the Food and Nutrition Service (FNS) imposes fairly stringent criteria on states that can limit access to SNAP. These include work requirements for individuals 18–50, in-person application procedures, onerous reporting requirements, and prohibitions on the use of SNAP funds for grocery delivery and restaurant purchases in most states (Hoynes & Schanzenbach, 2019). In comparison to increasing SNAP funding, Congress worked more quickly to address these issues in the face of COVID-19. In the Families First Act, for example, Congress effectively suspended work requirements in the program and gave the FNS authority to reduce reporting requirements to limit the need for in-person interaction. Nevertheless, the FNS also issued blanket denials of numerous state waivers applications, including attempts to waive restrictions on student eligibility, expedite all applications, and reduce the burden of verification requirements (Food and Nutrition Service (FNS), 2020a).

Third, while states generally demanded flexibility to expand SNAP access, they did so to varying degrees (see Figure 3). Some states, such as California and Arkansas, sought numerous waivers, including extensions of SNAP certification periods, allowance for telephonic signature, and waivers of certain penalties for noncompliance. Other states, such as Tennessee and Texas, sought only the bare minimum flexibilities, including emergency allotments, alterations to in-person interview procedures (Food and Nutrition Service (FNS), 2020b). Thus the FNS’s refusal to grant states flexibility on SNAP was reinforced by at least some states’ reluctance to seek it.

Figure 3. SNAP emergency waivers requests, March 13–13 May 2020.

Source: Tabulated by authors from FNS (2020b).

Conclusion

The COVID-19 pandemic should cause scholars and policymakers alike to carefully evaluate the adequacy of existing institutional arrangements and policy frameworks to manage global health and economic crises. In the first months of the crisis, subnational authority over public-health regulations has enabled governors and mayors to play a powerful agenda-setting role during the pandemic. As the crisis wears on, the ability of state and local governments to fulfill these crucial roles will depend largely on the structure and implementation of intergovernmental policy instruments.

Where discretionary policy is concerned, state and local governments are crucial players in lobbying for counter-cyclical aid. Nevertheless, the absence of formal structures for intergovernmental policy negotiation has meant that Congress treats these governments like one set of interest-group players among others, rather than as the providers of critical social services, significant contributors to GDP, and the stewards of fundamental rights. The crowding out of the intergovernmental lobby in debates over COVID-19 relief is only accentuated by an uncertain and decentralized context for policy information.

State and local governments also play a significant role in shaping the implementation of discretionary relief policies. Yet their effects on implementation depend in part on program design. Relief programs built upon pre-existing policy frameworks and intergovernmental networks have been implemented with relative speed. By contrast, those that require the creation of new programs at the state level have already experienced delays. The implementation of general intergovernmental aid is by its nature a more time-consuming endeavor, especially when state legislatures and governors find it difficult to reach agreement on how to appropriate federal funds.

Without a doubt, the federal government’s reliance on complex, discretionary programs rather than on automatic stabilizers has weakened the US response to the economic crisis. Within existing stabilizer programs, including UI, Medicaid, and SNAP, the intergovernmental allocations of fiscal and administrative responsibilities have created several problems of their own. First, programs that rely extensively on state financing, such as UI and Medicaid, require a sustained infusion of emergency federal spending during economic crises. As revenues plummet, states will be unable to cope with influxes of newly eligible beneficiaries, and in the absence of greater federal support will be forced to limit access to benefits. Second, burdensome application procedures can also impose barriers to benefit uptake. As the Medicaid case shows, states vary widely in terms of the administrative burdens they place on those eligible for the program. This raises the transaction cost of expanding eligibility and easing enrollment burdens during times of economic crisis. Federal program-integrity requirements––as the SNAP case shows––can make it difficult for states to waive burdensome application procedures, even during economic crises. Further, when state officials prefer the retrenchment of social policy, they may be less likely to seek waivers. Finally, as in the UI case, inadequate investment in state administrative infrastructures can greatly delay enrollment in new benefits during crisis periods.

Taken together, the evidence here suggests that federalism as it exists in the United States has inhibited the country’s response to the economic crisis caused by COVID-19. This is not because of any generic characteristics of federalism, however. Rather, it has more to do with several unique aspects of the institutions and policies that constitute the US federal system. First, the inability of states to engage in deficit spending places an outsized responsibility for crisis response on a Congress that is presently gripped by partisan polarization, which inhibits rapid and comprehensive action. Second, the federal government’s reliance on discretionary measures rather than automatic stabilizers means that state and local officials must petition for enhanced funding in a crowded lobbying environment. In the absence of a formal venue for intergovernmental policy negotiation, sub-national governments are treated like one set of special interests among others. Third, the automatic stabilizers that do exist allocate fiscal and administrative responsibilities to the states. Because of state revenue shortfalls, these programs demand emergency federal financing – rendering them less ‘automatic’ than they would otherwise be. The state-level patchwork of administrative requirements and outdated application systems can also inhibit the counter-cyclical effects of intergovernmental programs. Thus, while automatic stabilizers have numerous advantages, the intergovernmental character of these policies has inhibited their effectiveness during COVID-19, especially in programs that depend heavily on state financing and where administrative rules create barriers to access.

If state balanced-budget requirements are likely to endure for the near future, building a pandemic-ready federal system will require broadening the conceptualization of automatic stabilizers, federalizing funding for core social programs, eliminating administrative burdens, updating the infrastructure for implementation, and replacing discretionary general aid to state and local governments with automatic revenue sharing (Advisory Commission on Intergovernmental Relations (ACIR), 1978). There are of course numerous political obstacles to systematizing intergovernmental relations and to 'mechanizing' economic stimulus programs (Weaver, 2010). Yet as the fiscal and economic realities of COVID-19 become clearer, it will be more necessary than ever to treat these obstacles not as constants but as variables that must be acted upon.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Notes on contributors

Philip Rocco

Philip Rocco is assistant professor of political science at Marquette University. He is the coauthor of Obamacare Wars: Federalism, State Politics, and the Affordable Care Act (2016, with Daniel Beland and Alex Waddan) and coeditor of American Political Development and the Trump Presidency (2020, with Zachary Callen). He has published numerous articles in, among other venues, Health Affairs, Journal of Health Politics, Policy and Law, Political Science Quarterly, Publius, Forum, and Journal of Public Policy. He is working on a book manuscript titled Madison’s Engineers: How Policy Science Remade Federalism.

Daniel Béland

Daniel Béland is Director of the McGill Institute for the Study of Canada and James McGill Professor in the Department of Political Science at McGill University (Montreal, Canada). A specialist of comparative fiscal and social policy, he has published 20 books and more than 150 articles in peer-reviewed journals such as Comparative Political Studies, Governance, Journal of Public Policy, Journal of Social Policy, Policy Sciences, and Policy Studies Journal. You can find more about his work on his website: www.danielbeland.org.

Alex Waddan

Alex Waddan is an Associate Professor in Politics at the University of Leicester in the UK. He specializes in US and comparative social policy and has published five books and numerous articles in peer reviewed journals such as Political Studies, Policy and Politics, Journal of Social Policy and Policy Studies Journal.

Appendix

Appendix: Intergovernmental Demands for Federal Relief, March 4–4 May 2020

Table

Date	Organization	Request
3/4	USCM	$2 billion for uncompensated care
3/4	NGA	Swift passage of supplemental
3/10	NCSL	Flexibility in use of federal funds
3/18	USCM	$250 billion to stabilize city services, support public-health efforts
3/20	NGA	$150 billion in direct aid, 12% increase in FMAP
3/21	Big 7	Clarification that paid-sick leave tax credits apply to state/local governments
3/21	NACO	Make counties eligible for state stabilization fund
3/21	NLC	$250 billion to cities in direct funding
3/24	USCM	Lower population threshold in CARES Act Formula
3/30	USCM	Urge President to use full scope of Defense Production Act
4/10	NCSL	Revision of CARES Act language
4/14	USCM	$53.55 billion in emergency fiscal assistance through CDBG
4/16	NACO, NLC, USCM	$250 billion in assistance for local governments
4/21	NGA	$500 billion in additional state stabilization aid
4/21	Big 7	Robust flexible assistance

Source: Authors’ analysis of public websites. Note: USCM = US Conference of Mayors; NCSL = National Conference of State Legislatures; NGA = National Governors Association; NACO = National Association of County Officials; NLC = National League of Cities; Big 7 = USCM, NGA, NACO, NLC, Council of State Governments, International City/County Management Association