Frankel Lecture Introduction

Between 2020 and 2025, the United States experienced the most dramatic expansion and subsequent contraction of its social safety net in modern history. In response to the COVID-19 pandemic, the United States adopted sweeping benefit expansions that dramatically reduced child poverty and brought the nation closer to universal health coverage than ever before.^[1] Yet this moment proved fleeting. Congress in 2025 not only rolled back these gains; it went further—curtailing benefits to levels far below their pre-COVID baselines. This rapid cycle of expansion and contraction defied the conventional wisdom that “new policies make new politics”—that once people receive benefits, they will fiercely resist any subsequent attempt at retrenchment.^[2]

In the Houston Law Review’s Thirtieth Annual Frankel Lecture, COVID and the Great Retrenchment, Professor Samuel R. Bagenstos seeks to answer the question of why the COVID-era benefits expansion was much less sticky than policymakers anticipated.^[3] Drawing on his experience as General Counsel to both the Office of Management and Budget and Department of Health and Human Services from 2021–2024, Bagenstos offers an insider’s account of how government officials designed the COVID-era expansions—and seeks to explain why they were so short-lived.

Bagenstos is joined in his efforts to solve the mystery by commentators Professors Andrew Hammond and Abigail Moncrieff.^[4] While each author ultimately offers their own account of what brought about the rapid post-COVID retrenchment, they are not inconsistent with one another. Rather, when read together, the three sets of remarks illustrate just how many social and political forces and competing interests policymakers must navigate when seeking to shape, or reshape, the American welfare state. And while this set of remarks offers meaningful insight into why benefits extended during the COVID pandemic were revoked, they also implicitly pose another, perhaps more important, question—for policymakers seeking similar, though permanent, benefits expansions in the future, what lessons does the COVID-era and its aftermath provide that might lead to such efforts enjoying more long-term success?

When designing the policies at issue, Bagenstos and his colleagues explicitly sought to move federal policy toward more universal benefits, favoring universalism (available to all or most people) over targeted benefits (available only to those meeting specific criteria) because universal benefits can promote equal citizenship, avoid stigma, reduce administrative burdens, and, as a result, secure broader political support.^[5] By any measure, the effort to push individual entitlements toward universalism was wildly successful. Through expansions of the Child Tax Credit, ^[6] unemployment insurance,^[7] stimulus payments,^[8] and Medicaid,^[9] child poverty fell by 46% to a historic low,^[10] nearly twelve million people rose above the poverty line,^[11] and the uninsured rate dropped to 7.7%—the lowest ever recorded.^[12]

Yet America rejected this move towards universalism. By the end of 2024, nearly all these gains had been reversed. And with the passage of the One Big Beautiful Bill Act (OBBBA) in 2025,^[13] the Congressional Budget Office predicts that approximately 10 million people will lose health insurance^[14]—reversing not only COVID-era gains but also most of the health coverage expansion since the Affordable Care Act of 2010.^[15] Bagenstos and his commentators offer different, though complementary explanations for why this happened.

Bagenstos offers two primary explanations for the reversal. First, the COVID benefits expansions, particularly the generous unemployment insurance, shifted bargaining power from employers to workers.^[16] That, in turn, mobilized employers to aggressively oppose making those benefits permanent.^[17] Second, opponents of the expansion successfully framed the benefits not as the new, universal baseline for health and welfare programs, but rather as a set of temporary measures to alleviate the unique effects of the pandemic.^[18] Moreover, casting benefits as assistance for the “deserving poor” rather than universal entitlements provided the conceptual foundation for the OBBA’s even deeper cuts.^[19]

Professor Hammond’s response, Reconsidering Expansion and Retrenchment in the U.S. Welfare State,^[20] offers two friendly amendments that elaborate on and complicate Bagenstos’s employer-backlash story. First, he argues that “administrative burdens and our institutional design of devolved responsibility to state and local governments disrupted policy feedback before . . . these temporary extensions could take hold.”^[21] When benefits are distributed through fifty state systems, universalism is difficult to implement, even when that is the goal. Some states’ unemployment insurance systems crashed under the volume of applications; others simply refused to implement the expansions at all, forgoing billions in federal funding.^[22] This implementation of federalism, Hammond contends, undermined whatever universalist tendencies the legislation possessed.

Hammond also reframes Bagenstos’s characterization of the pandemic expansions as universalistic social policy. He distinguishes between the “First Pandemic Deal” (the CARES Act and related 2020 legislation under President Trump) and the “Second Pandemic Deal” (the American Rescue Plan and subsequent Biden-era laws).^[23] The First Pandemic Deal, he argues, was conventional Keynesian stimulus—countercyclical government spending to prop up an economy in freefall—not an effort to build a more generous welfare state.^[24] Members of Congress were not focused on the persistence of child poverty; rather, they were bent on avoiding the mistakes of 2009, when the stimulus package proved too small to effectively avert the Great Recession. Only the Second Pandemic Deal had genuinely universalist ambitions.^[25] Disaggregating the legislation into two distinct deals, Hammond argues, should make us less surprised that the policies did not persist past the official public health emergency.^[26]

While Hammond agrees with Bagenstos that opponents’ characterization of the COVID policies as a response to a particular, temporary emergency likely “got in the way of making them permanent,” he does not consider employer backlash the driving force.^[27] Instead, he argues, it was simply “the latest chapter in the Republican Party’s four-decade fight to undo the welfare state” established by the New Deal, the Great Society, and Barack Obama’s Affordable Care Act.^[28] President Trump and his allies in Congress had tried mightily to undermine Medicaid and SNAP during his first administration, but they did not have the votes to do it legislatively and were stymied in the federal courts when they tried to do it administratively.^[29] In 2025, however, they ran the same retrenchment playbook with a more cohesive party caucus in both houses of Congress.

Critically, Hammond argues that OBBA achieves not merely programmatic retrenchment (cutting specific benefits) but systemic retrenchment. First, it “defund[s] the welfare state by constraining the flow of revenues to future administrations” by making permanent the individual and corporate tax cuts from the 2017 Tax Cuts and Jobs Act.^[30] Second, it modifies political institutions and decision-making processes in ways that will shape future policy outcomes, encouraging states to make deeper cuts to reduce their own fiscal liability and undermining the data by which we assess the welfare state’s efficacy.^[31] The OBBA thus does not simply retract specific policies; rather, it remakes government structures in ways that embed its policy choices, making future policymakers’ efforts to reverse those choices immensely challenging.^[32]

Professor Moncrieff’s commentary approaches the question of why the dramatic COVID-era expansions of public benefits failed to become entrenched from a more theoretical perspective. In Uncertainty in the Time of COVID, Moncrieff draws on the political philosophy of John Stuart and Harriet Taylor Mill to develop what she calls the “uncertainty principle.”^[33] Moncrieff argues that perceived regulatory legitimacy hinges on certainty and observability—when the costs and benefits of a program are observable, either through sensory perception or widespread consensus, society can confidently intervene.^[34] By contrast, when costs and benefits are unobservable, intervention risks appear arbitrary, undermining public support.^[35]

The COVID pandemic, Moncrieff contends, created deep uncertainty. The harms that became observable during the COVID pandemic did not arise from bad laws; they arose from a virus.^[36] While legislation could mitigate some of the pandemic’s downstream harms, complex biological systems like pandemics resist the kind of clear cause-and-effect relationships that make regulatory interventions demonstrably beneficial.^[37]

Moreover, the government could not demonstrate that the COVID benefits outweighed their costs. School closures, for example, produced observable learning losses and mental health harms, but their success in mitigating the virus’s spread—an unobservable benefit—was impossible to measure.^[38] Under Moncrieff’s framework, the mix of observable costs and uncertain benefits made COVID interventions seem arbitrary, eroding the public trust necessary for benefits to become entrenched.^[39]

Taken together, these essays reveal that the reversal of COVID-era benefits stemmed not from a single cause but from a confluence of factors operating at different levels of analysis. Bagenstos emphasizes political economy—how shifts in workplace bargaining power mobilized employer opposition and how opponents successfully contested the social meaning of the benefits. Hammond emphasizes institutional design—how federalism fragmented implementation and how the legislation’s structure reflected different purposes at different moments. Moncrieff emphasizes epistemology—how the inherent uncertainty surrounding pandemic interventions undermined public confidence in government action.

Yet these explanations are complementary rather than competing. Employer opposition succeeded in part because implementation failures made the benefits seem less universal than intended. The framing of benefits as temporary emergency measures gained traction in part because the observable costs of intervention (inflation, labor shortages) appeared more certain than the benefits. And the uncertainty principle helps explain why employers’ arguments against entrenchment proved politically persuasive rather than merely self-interested.

For policymakers seeking to expand the American welfare state in the future, these essays offer sobering lessons, illustrating the difficulty in effecting transformative changes in an era of extreme polarization, federalized implementation, and complex causal chains.

The essays, however, also suggest paths forward. Collectively, they demonstrate that building durable benefits expansions requires not just good policy design but a strategy that grapples with the challenge along multiple dimensions—political economy, institutional structure, narrative control, and epistemic legitimacy. This entails building political power among beneficiaries, combating opponents’ narrative framing, ensuring consistent implementation across jurisdictions, and demonstrating observable benefits that outweigh observable costs. Whether future efforts to expand the American welfare state can rise to this formidable challenge remains an open—and urgent—question.