Our spring issue, “Pandemic Politics,” is out now. Get a discounted subscription today!

COVID-19 Is Exposing the United States’ Ragged, Shameful Safety Net

A decent welfare state should provide the basics of life so everyone can flourish. The United States’ patchwork of poorly funded safety net programs is doing the opposite — dropping people through a trapdoor as the pandemic ravages the economy.

Miguel Diaz, who works for the City of Hialeah, hands out unemployment applications to people in their vehicles in front of the John F. Kennedy Library on April 8, 2020 in Hialeah, Florida. Joe Raedle / Getty

Like many viruses, COVID-19 is especially dangerous for vulnerable groups such as the elderly or those whose health is already compromised. We can add to this list the American welfare state itself: its preexisting conditions (a patchwork of categorically targeted programs that rest on job- or work-based provision) have made it a susceptible target for the coronavirus’s deadly reach. Its programmatic antibodies (which tend to weaken when they are needed most) have offered little resistance to the spread of economic insecurity. And its symptomatic failures — evident even when the economy is doing well — are starker still as the pandemic unfolds.

“The magnitude of a crisis,” the New York Times editorial board noted somberly in early April, “is determined not just by the impact of the precipitating events but also by the fragility of the system it attacks.” That fragility, in turn, ensures that this pandemic — as it feeds upon and widens existing disparities — will have especially dire consequences for poor and working-class people.

At a moment when our collective health demands a nimble and coordinated response, we are instead saddled with jurisdictional scrums over essential medical supplies and a bewildering array of “shelter in place” policies based on political whims and idiosyncratic metrics.

The disarray is most apparent in the arenas of provision — health care and paid leave — that are now most urgent. The United States, unlike every other peer nation, lacks universal health coverage, instead relying on a shaky foundation of employment-based coverage that stratifies access by income and occupation. Public coverage targets those left behind, but reluctantly and haphazardly. The Affordable Care Act varies widely in its quality and accessibility across jurisdictions, and Medicaid is only broadly available to the poor in post-ACA “expansion” states. Over 10 percent of the population (about 30 million people) lacked insurance before the COVID crisis, and — by one estimate — another 9.2 million lost their coverage in the last four weeks. Also alone among its peers, the United States lacks paid family and sick leave. About a quarter of US workers are unable to take paid sick leave, and only one in five are granted paid family leave.

So, at a time when early intervention and treatment are urgently needed, we are losing health coverage almost as fast as we are losing jobs. And at a time when our collective health rests on limiting social contact, most front-line service workers are unable to take advantage of either paid leave or stable health coverage.

These failures underscore both the immediate peril and the broader, pervasive weaknesses of the US welfare state. The elevation of employment as the single most important marker of “deservingness” has narrowed the reach and the generosity of our social programs. For those in the middle and upper tiers of the income distribution, this yields a heavily subsidized and more generous (though still lackluster compared to peer countries) “private welfare state.” For those in the lower tier of the income distribution, such benefits rarely flow from jobs. Work is, instead, increasingly a core eligibility standard for public assistance — including cash assistance, in-kind or near cash–assistance, and the Earned Income Tax Credit (EITC).

This makes for an inside-out safety net, no matter where we are in the business cycle. Because good benefits follow good jobs, employment-based provision systematically widens market inequalities. And because the unemployed lose key social supports along with their jobs, employment-based provision magnifies rather than counters downturns: job-based benefits, as Steffie Woolhandler and David Himmelstein are fond of pointing out, are like an umbrella that melts in the rain. The draconian work requirements embedded in public programs only “work” when the economy is booming; in a downturn — let alone in a crisis that rapidly shutters much of the private economy — they are punishingly counterproductive.

None of this is accidental. The American welfare state is unequal by design; a patchwork of assistance that differs — across jurisdictions and across citizens — in its reach and generosity. A few programs offer uniform or standardized benefits based on contributory financing (Social Security and Medicare) or categorical eligibility (Supplemental Security Income). But others are narrowly targeted and means-tested, and defer most of the details — benefits, eligibility, sanctions — to state and local governments.

As a result, social provision is starkly unequal across state lines, and (reflecting the discretion in state and local administration) starkly unequal within state programs. We have relentlessly devolved responsibility from federal to state (and from state to local) control and yet, at any hint of broader economic trouble, state and local programs lack the capacity or willingness to help those in need.

In common parlance, social provision is described as our “safety net.” But taking a look at two programs amid the pandemic — unemployment insurance and Temporary Aid to Needy Families (TANF) — produces a much different picture: often, the US safety net is more like a trapdoor.

Unemployment Insurance

Unemployment insurance is a joint federal-state program. Federal law sets broad requirements and levies a tax to cover program administration, the federal share of extended benefits, and loans to states. State taxes on employers build a trust fund to pay claims, and states — who, as the Department of Labor summarizes, “have developed diverse and complex formulas for determining workers’ benefit rights” — have wide leeway to set eligibility, generosity, and duration.

State discretion is exercised in formal and informal ways. The generosity of benefits — both in terms of the weekly benefit and the number of weeks it is available— varies widely across the states. The maximum weekly benefit ranges from $235 in Mississippi to $1,234 in Massachusetts. The duration of benefits is twenty-six weeks in most states, but runs from twelve weeks in Florida to thirty weeks in Massachusetts. A six-month spell of unemployment, in other words, would yield a net benefit of over $37,000 for an unemployed worker in Boston, and less than one tenth of that (barely $3,000) in Biloxi or Boca Raton.

Meager benefits discourage enrollment, as does the administrative burden of intentionally complex application systems. Florida’s online-only system — which until 2014 featured a mandatory forty-five-question math, reading, and research skills test — dissuades or disqualifies over half of those who start an application. As a result, the share of the unemployed who actually see a check also varies across states — from under 10 percent in North Carolina to almost 60 percent in New Jersey.

This disparity can be seen in the graph below, which plots program inclusion (the share of the unemployed receiving benefits) against program generosity (the average benefit received, taking into account both the average weekly benefit level and average benefit duration in each state). The low-road (mostly Southern) states crowd the lower right corner of the graph, with inclusion rates under 30 percent and an average benefit of under $6,000.

Unemployment Insurance, Generosity and Inclusion by State (2018) Source: Bruch et al, State Safety Net Database

In times of exceptional need or demand — such as during a recession or disaster — the federal government typically steps in to supplement and backfill state unemployment insurance programs. Since 1970, high unemployment has triggered the payment of extended benefits — an additional thirteen weeks funded jointly by state and federal dollars. But because the “triggers” for extended benefits work slowly, the federal government has also offered more immediate extensions (as it did with the Emergency Unemployment Compensation program during the Great Recession).

This time has been no different. The Families First Coronavirus Response Act (passed on March 18) pumped $1 billion into administering state unemployment insurance (UI) programs, in exchange for new state standards and conditions. In order to draw down these funds, states must improve their procedures for notifying workers of their eligibility, provide multiple (not just online) methods of filing, give prompt notice of the receipt of a claim, waive waiting periods for benefits, nix the requirement that recipients must be actively searching for work, and ensure that employers are held blameless for COVOID-19 layoffs (conventionally, UI is “experience-rated,” so employers with histories of layoffs are taxed at higher rates).

The CARES Act (passed March 26) bolsters both benefits and coverage. Pandemic Unemployment Assistance (PUA) extends unemployment assistance to workers who are otherwise left out of state UI programs — including self-employed workers, “gig” workers, independent contractors, freelancers, workers seeking part-time employment, workers who do not have a sufficient employment history to qualify for state UI benefits, and those that have exhausted their benefits. These applicants will have to either demonstrate that they are unemployed or unable to work due to COVID-19-related illness, quarantine, caregiving, or layoff. PUA is available for thirty-nine weeks or until December 31, 2020 and will carry a minimum benefit equal to one-half the state’s average weekly UI benefit.

The Pandemic Unemployment Compensation program adds $600 per week (through the end of July) to all unemployment claims paid under either the regular UI program or the Pandemic Unemployment Assistance program. The Pandemic Emergency Unemployment Compensation (PEUC) provides a thirteen-week extension of state UI benefits. All three of these programs are entirely covered by federal dollars.

Such extended benefits have proven absolutely critical in delivering a modicum of financial security for workers (and state budgets) during prolonged or sudden economic crises. But they also highlight the weakness and inequity of state UI programs. States have made little effort to adapt their programs to changes in the labor market. The inclusion and generosity of state programs vary widely from state-to-state. State programs lack the fiscal capacity to respond to any substantial downturn in the economy. They lack the administrative capacity to interpret and launch the new federal programs in a timely and consistent fashion. And they don’t have the technical capacity (many state application systems rest on hardware and software dating to the 1980s) to process the avalanche of claims.

Our fragile and uneven unemployment insurance system, in other words, is powerful evidence for the importance of both universal and uniform standards on eligibility and benefits, and for the sort of countercyclical fiscal capacity that only the federal government can provide. The alternative is scrambling for temporary fixes and infusions of federal money every time calamity strikes.

Temporary Aid to Needy Families

Temporary Aid to Needy Families (TANF), the United States’ cash assistance program for poor families, grew out of Bill Clinton’s regressive welfare reform bill of the mid-1990s. “Ending welfare as we know it,”  the Personal Responsibility and Work Reconciliation Act of 1996 (PRWORA) replaced the already-meager cash assistance program for poor households with TANF, an even sparer program that was made conditional on work and other behavioral surveillance (drug use, school attendance) and sanctions. No longer would the federal government share program costs with states — instead, it would send lump-sum “block grants” to states, which were granted wide latitude to set benefit levels and eligibility rules.

The results have not been pretty. Left to their own devices, states have pared back direct cash assistance and shifted dollars to in-kind services (childcare, work supports, parenthood classes) in pursuit of TANF’s behavioral expectations. Just one-fifth of TANF funds are dispensed as cash assistance to families in need (a diversion of resources from direct assistance to paternalism that is most pronounced in states with higher shares of black families).

TANF has dramatically reduced both the real value of the benefit and the share of poor families receiving assistance. Nationally, only 21 percent of poor families receive cash assistance — down from 82 percent in 1979 and 68 percent in 1996). And the state-to-state disparity is stark.

The graph below plots program inclusion (the share of poor families with children receiving cash assistance) against the average amount received by a recipient family. As with unemployment insurance, the low-road states are crowded into the lower left quadrant: seventeen states send cash assistance to fewer than one in ten poor families; in twenty-four states recipients get less than $4,000 in a year. The gap on both counts is enormous — from an inclusion rate of under 4 percent in Louisiana to over 67 percent in California; from an average amount received of just $1,500 in Maine to over $10,000 in Wyoming.

TANF, Generosity and Inclusion by State (2018)
Source: Bruch et al, State Safety Net Database

The net result, long before the abrupt collapse of the economy in late March, has been a marked shift in public assistance (within TANF and across a raft of other programs) that favors working parents with earnings at the expense of able-bodied nonelderly adults without children, as well as those with only tenuous ties to the labor market. The United States’ most substantial form of direct assistance, the EITC, is reserved for working families with kids. The national poverty rate has settled in at 12 to 15 percent since 1996 but, with the near evaporation of cash assistance as a social support, the portion of those in deep poverty has grown dramatically.

So where does TANF figure into the state and federal COVID-19 response? The “Families First” Act passed on March 18 makes not a single reference to TANF, the United States’ only direct assistance program for needy families. The more expansive CARES Act, passed on March 26, mentions TANF only once — a single paragraph aside on page 411 that simply extends the current appropriation through November 2020.

The sole concession to a crisis where states anticipate a spike in demand for social assistance is a tepid advisory from the Department of Health and Human Services (DHHS) prefaced by the warning that “there are no additional federal TANF funds available for states to address COVID-19 needs . . . any support states and tribes provide using federal TANF funds must come from their existing allocations and unobligated funds, and must meet the requirements and restrictions that apply to the use of TANF funds.” While the agency acknowledges that “[w]e are facing a national public health and economic emergency of unprecedented proportions,” it then simply proceeds to remind states they are “responsible for ensuring that they are providing benefits only to families eligible for those benefits.” (It is within their discretion, DHHS adds, to relax recertification procedures, ease up on work requirements, or use non-recurrent short-term benefits to meet short-term demands.)

Other elements of the federal government’s COVID-19 response have been less miserly. Funding for the Supplemental Nutrition Assistance Program (SNAP) was buttressed by an additional $15 billion in the CARES Act, and most states have taken advantage of program waivers that allow them to streamline recertification, relax work and eligibility requirements, offer emergency supplementary benefits, and sustain school-based meal programs by delivery or pick-up. The administration has, reluctantly and temporarily, put on hold its push to attach work requirements to SNAP and Medicaid. The CARES act also includes additional funding for childcare providers and Head Start programs.

But nowhere in this sprawling $2 trillion dollar relief and stimulus package was there one more dollar for cash assistance to address the pressing needs of the nation’s most impoverished families.

This contrast — between the expansion of unemployment insurance and the silence on TANF — is stark and telling. And it replicates the policy choices that elites made in the wake of the Great Recession. If that experience is any guide, our safety net will respond inadequately and unevenly to need during the downturn, and might shrink even further as the economy recovers.

As in the past few weeks, the extension of benefits and eligibility during the Great Recession leaned heavily on UI and SNAP. TANF received a supplemental appropriation of $5 billion, but states — leery of making program changes that might increase caseloads in the long run—proved reluctant to expand benefits or eligibility. While the federal government put on hold penalties for not meeting work participation goals as unemployment climbed, no states significantly modified their work requirements.

The net result, as Hillary Hoynes and Marianne Bitler have documented, is a safety net that is relatively unresponsive to either economic downturns or demonstrable need. Unemployment insurance is countercyclical by design (net payments jump as the economy slides into recession), but — given the weakness and unevenness of state programs — only temporary and substantial infusions of federal money (in the Great Recession as now) ensure that this is the case. The EITC, tethered to earnings, is of no benefit to the unemployed. And TANF has no countercyclical mechanism. “Protection through TANF has all but disappeared,” Hoynes and Bittler conclude bluntly. “The program no longer appears to be responding to need.” On balance, the neglect of TANF and the increased reliance on social policies that reward or subsidize work (such as the EITC) pushes the impact of those policies up the income ladder — and increasingly out of reach to those in the most need.

The end of the Great Recession is a cautionary tale in this respect. The infusion of federal money into SNAP and UI in 2007–9 cushioned the blow of the Great Recession and kept millions of Americans from slipping into poverty. In the states, however, revenue shortfalls rekindled the political backlash against social assistance before recovery even took hold. Thirty-six states exhausted their UI trust funds during the Great Recession, and many responded by reducing benefits and benefit duration and erecting new obstacles to application. Two years into the recovery, in December 2011, a third of states had lower TANF caseloads than before the recession.

More broadly, we need to honestly assess the impact and the consequences of these policies. While state and federal social programs do ameliorate market inequalities, their reach and their effectiveness vary widely across states. As a result, and on all-important metrics of economic well-being, security, and opportunity, the state where you live shapes your life and your life chances. In a nation marked by pervasive and durable economic disparities, that patchwork of policy choices has itself become a potent source and form of inequality.

A Better Welfare State

All of this begs the question: what is a welfare state for?

Seventy years ago, the British sociologist TH Marshall famously argued that social policies are not simply meant to provide “a modicum of economic welfare and security” or to “abate the obvious nuisance of destitution.” They should also secure “the right to share to the full in the social heritage and to live the life of a civilized being according to the standards prevailing in society.” Such rights should be extended irrespective of “the market value of the claimant” and with the goal of “modifying the whole pattern of social inequality.”

In the United States, Marshall’s aspirations have been frustrated by the fragmentation of social provision and elites’ reliance on employment as either the primary source of economic security or the core criteria for any claim on public support. Policies designed to protect worthy white widows from work in the 1930s are now largely dedicated to forcing poor women into the low-wage labor market. Concessions to southern segregationists in the formative years of the US welfare state persist in the form of starkly unequal benefits from state-to-state. And the patchwork of job-based and work-based benefits firmly tether social citizenship to “the market value of the claimant.”

All of this is on full display in the government’s response to the COVID-19 crisis — in which fragments of temporary assistance at once underscore and paper over the holes and gaps and pervasive inequities of the US welfare state. The result is not relief for those who need it most, but instead a meager and threadbare patchwork that sustains widespread insecurity — and punishing inequality.