The Biden-Harris Administration did a great deal to improve economic opportunity for working Americans, first through the rapid recovery from the COVID-19 pandemic and then through historic investments in manufacturing, energy, and infrastructure. So why didn’t its policies—including hundreds of billions in expanded social safety net benefits—translate into higher favorability ratings for Joe Biden and stronger working-class support for Kamala Harris at the ballot box last November? This puzzle has sparked fierce debate. Last year, Ezra Klein noted that policies he expected to create positive feedback loops between delivery and political dividends throughout the Administration didn’t “quite work.” Some urged Democrats to abandon “deliverism”—the strategy of winning support through tangible policy benefits—while others suggested the benefits simply weren’t substantial enough in a period of high inflation. And in the election’s painful aftermath, Sam Rosenfeld and Daniel Schlozman argued in this journal that the Democrats’ “attempt to make electoral hay out of well-designed policy alone must be counted as a failure.”

As both an academic studying how policies reshape politics and a former Biden-Harris policy official, I’ve found these debates to be frustratingly abstract. As a researcher, I’ve found them to be missing concrete evidence to test their sweeping propositions. And thinking about my experiences serving in the Administration, I believe that their provocations are not specific enough to inform future policy design or implementation.

Since leaving government to return to academia last year, I’ve been trying to bring a sharper empirical lens to these questions. Fortunately, I had an ideal example to work with: the pandemic-era expansion of unemployment insurance (UI).

This intervention was among the largest relief measures enacted during the COVID-19 pandemic and the largest single expansion of UI in U.S. history. At its peak, nearly $1 trillion in benefits reached fully one-third of American workers; for a time, the typical recipient was eligible for payments exceeding their previous wages by more than 130 percent. This dramatic expansion, first initiated under President Trump’s CARES Act and later extended at a lower level by President Biden’s American Rescue Plan, transformed a traditionally stingy state-controlled system marked by decades of cuts into a generous federal safety net approximating a basic income for many workers. It had a significant material impact on American families’ well-being, keeping nearly six million Americans out of poverty and helping tens of millions more make ends meet. Beyond the temporary expansion of benefits, the Biden-Harris Administration invested $1 billion in modernizing the system itself, funding state-level improvements in technology, access, and delivery.

Using surveys of more than 4,000 Americans and dozens of interviews with workers, I have found that the political impacts of these changes were mixed. The temporarily expanded benefits did succeed in changing how workers thought about the UI system, but they did not translate into sustained political actions, such as voting, contacting elected officials, or supporting electoral campaigns. Nor did expanded benefits lead UI recipients to become more supportive of the Biden-Harris Administration or Democratic candidates. And while federal investments have made meaningful improvements in the UI system, they have not changed the underlying politics that have plagued the system for decades. Panning out from the UI system, my findings carry important lessons for how liberal activists and policymakers should think about using policy design to reshape politics in the coming years. My conclusion is that we must move to a new vision of progressive policymaking, one that might be called policymaking as power-building.

This vision of policymaking takes fundamentally reshaping politics as one of its objectives, in addition to addressing specific social or economic problems. This can include attempting to shift public opinion, with goals that range from boosting support for an individual political candidate or political party to changing views on specific public policies or broader issues. Policymaking as power-building can also involve shaping the landscape of organized interest groups: This includes efforts to strengthen particular groups or movements or to weaken or divide others. And this approach can involve changing the capacities, structure, or orientation of government agencies in ways that can facilitate further policy development and help broaden supportive political coalitions. The underlying logic is that public policy can be a tool to change the political terrain, creating new political opportunities that enable further policy victories.

Policymaking as power-building is related to deliverism, but goes beyond it. Deliverism as described by its proponents has tended to focus narrowly on using policy rollouts to increase voter support for certain politicians and, ultimately, to produce desired electoral outcomes. The approach I describe here includes such efforts but also considers a range of other pathways to building power. Indeed, as we will see, the changes to voter sentiment and behavior anticipated by deliverism’s proponents are often the hardest to engineer in our current polarized climate—and may also be the most fleeting as compared to more durable efforts to reshape organized interests and social movements. The approach I describe in this essay also builds heavily on what political scientists have dubbed “policy feedback loops,” but it departs from this academic literature to consider how political outputs can be deliberately built into policy design and implementation—not just left to chance.

Returning to what we can learn from the Biden-Harris Administration’s legacy, I argue that policymakers and pundits should focus on three factors that determine whether policy can translate into changes in political power: movements, mechanics, and messages.

Link Policy to Movements

Using the lens of UI policy, I will first consider the narrower outcomes envisioned by deliverism’s proponents. Did expanded UI benefits during the Biden-Harris Administration change the attitudes and opinions of workers who received those benefits, including how they thought about the UI system and the welfare state more broadly?

Studying a national survey of more than 1,500 Americans with recent experiences of unemployment during the pandemic, I contrasted individuals who applied for benefits in 2020 and 2021 and received them to those who applied and did not. I found substantial differences in these groups’ views of UI benefits and the welfare state. UI recipients were 17 percent more likely to say that they supported permanently expanding UI benefits; 29 percent more likely to see their benefits as earned rights, rather than as welfare; and 27 percent more likely to believe UI benefits were important for reducing poverty. Interviews with recipients indicate that they were often very grateful for the assistance, noting in some cases that they were “more than I expected them to [be]” and that they “help[ed] me in my time of need.” Recipients described how they used the benefits to make rent payments, pay for prescriptions and medical expenses, and put food on the table.

And yet, despite the stronger support for UI benefits and the welfare state that receipt of these benefits engendered, I find no evidence that workers then acted on these new beliefs, as a narrower deliverism perspective might have anticipated. UI beneficiaries were no more likely to say that they had contacted politicians to voice support for UI legislation, to have voted in elections because of UI issues, or to have talked to and engaged other beneficiaries around common issues. Nor did UI benefits change workers’ perceptions of the Biden presidency. Recipients were no more favorable toward the President even after having benefited from the UI expansions that happened during Biden’s term.

On one level, this may not come as a surprise. Past research has found that the experience of unemployment can be isolating, leading individuals to withdraw into family and disengage from society, and so even the receipt of UI might not be enough to prevent that withdrawal. But I did not find unemployment was universally isolating. In fact, one group constituted an important exception: unemployed workers, including UI recipients, who were engaged by organizations that represent workers or focus on issues affecting the unemployed.

UI recipients who were engaged by churches, community-based organizations, worker centers, and unions were substantially more likely to report translating their lived experiences with UI benefits into political action to defend and expand the program. Workers who had been mobilized by these groups were more than three times as likely to report having contacted a politician about expanded UI benefits and nearly twice as likely to report having talked with other beneficiaries about their experiences. The problem was that only a tiny minority—7 percent of unemployed workers—reported being engaged in this way by community organizations, so UI recipients’ overall participation in politics remained low. (Just 8 percent of UI recipients reported contacting an elected official to voice support for expanding UI benefits.)

This is a critical missing piece of the policy-to-politics pipeline, and one that policymakers neglect when designing public programs. Far too often—as in the deliverism approach—the assumption is that voters will perceive changes in policy on their own and adjust their political beliefs and attitudes accordingly. But this approach requires that individuals correctly interpret a policy’s benefits; understand how the benefits affect people’s lives, including their own; assess the politicians and party responsible for implementing the policy; and then conceive and undertake political actions in support of those politicians, including deciding to turn out in an election.

Yet all of these steps represent an incredibly tall order under any conditions for busy Americans who may pay only minimal attention to politics. And citizens are even less likely to take these steps in our era of partisan polarization, in which many of them apply partisan-tinged lenses to their interpretation of the world. All of these steps are also especially challenging with respect to federal policies, including UI, that are administered jointly by the federal government and states, making accountability more challenging. When the UI system fails workers, should they blame Congress, the President, their governor, their state legislators, or their UI agency? And how should they express their blame? Under these circumstances, the role of civic organizations becomes even more important to help individuals understand and interpret the impact of public policies on their lives and connect those experiences with specific political actions. This conclusion is strongly resonant with Deepak Bhargava, Shahrzad Shams, and Harry Hanbury’s obituary for deliverism, which affirmed that policy on its own without social connections will fail to inspire action. [See “The Death of ‘Deliverism,’” Democracy, June 2023.]

Where was support in federal relief legislation for civic organizations that could help workers access benefits and make sense of their experiences? During the pandemic, networks like Unemployed Action—supported by the Center for Popular Democracy—were helping to coordinate organizing of, and advocacy by, unemployed workers around unemployment insurance. But Congress and the Administration did not partner with these networks—for example, by funding these groups to get the word out about benefits and help workers access relief.

A pilot program we launched at the Department of Labor in 2022 offers one glimpse at an alternative path. The department gave seven states funding to partner with community-based organizations to help expand the reach of UI benefits to traditionally underserved communities. These organizations helped workers navigate the process of applying for benefits and provided broader support, such as counseling, help finding training opportunities, and connections to other safety net programs in their moment of crisis. My co-authors and I recently found that these UI “navigators” helped underserved and disadvantaged workers better access UI and that they also may have fostered stronger interest in workplace economic and political action. Navigator-assisted workers were more likely to express interest in banding together with their co-workers to address workplace issues, including through collective action, as compared to similar unemployed workers who had not received navigator assistance. In surveys, navigator-assisted workers in Maine reported about 17 percent higher interest in such actions than other comparable unemployed workers did. (One example: 33 percent of navigator-assisted workers said that they would likely consider going on strike if they faced a workplace-related issue or problem.)

In the end, however, this was only one pilot program. The federal government did not fund such efforts at scale—neither for UI benefits nor for the broader raft of pandemic-era relief programs supported during the Biden-Harris Administration, such as the expanded child tax credit and economic impact payments. There are some important earlier precedents that policymakers could have looked to: The Affordable Care Act, for example, has funded community navigators to help individuals enroll in insurance coverage and assist people from low-income communities and communities of color with overcoming language barriers, stigma, and distrust in government. The absence of more substantial funding for civic groups in the Biden-Harris recovery effort was a significant missed opportunity for power-building policy.

Closely related to the lack of federal support for organizing was the temporary nature of the expanded UI benefits. As other public policy scholars have noted about the broader COVID-19 relief program enacted by the Biden-Harris Administration, temporary benefits simply did not provide enough time for organizers to reach a critical mass of Americans. With more runway, groups might have been able to build relationships of trust with recipients of COVID-19 relief and then help them make sense of their experiences and connect those experiences to politics. Studying the failure of the temporarily expanded child tax credit to inspire more political support among recipients, political scientist Jamila Michener reaches similar conclusions. Michener notes that organizers did seek to embed efforts to boost voter registration and turnout within their work to increase take-up of the expanded credit among eligible communities. But, she argues, “[T]he short time frame for building political support and the cacophony of policies enacted during the pandemic did not create optimal conditions for focused and effective organizing or focused messaging to support positive feedback.”

Just as temporary benefits do not give enough time for organizing, delayed benefits can also hamper such efforts. The fact that so many of the Biden-Harris Administration’s infrastructure and manufacturing investments were to be implemented over such a long time period may have made it more challenging for organizers to point to material changes in workers’ lives. So too for some of the Administration’s biggest social policy breakthroughs. As Biden economic adviser Bharat Ramamurti noted, it was political “malpractice” that the “wildly popular” hard cap on Medicare out-of-pocket spending for elderly Americans went into effect in 2025—and therefore its benefits would not be felt until after the 2024 election.

The broader lesson for policymakers is that efforts to shift the political views and actions of individuals through policy delivery, particularly in a polarized time, need to create openings for movement organizing alongside those policies. That might include, for instance, attaching funding for civic organizations to help with delivery and take-up of benefits, as in the case of the UI navigators. It could also mean incorporating such organizations into the design and rollout of policies—for example, building in opportunities for groups to discuss implementation with policymakers so that they are then prepared to engage with potential recipients around that policy. And critically, policymakers need to foster a favorable terrain for organizing, above all by allowing sufficient time for organizing work to unfold while policies deliver tangible benefits to Americans.

Change Mechanics to Shift Power

Reshaping individual attitudes and actions is not the only way policy can build power. A second, underappreciated channel is to change the underlying mechanics of program administration. Done well, implementation of policies can build new muscles in government that enable further expansions of the policy, empower its beneficiaries, and create new political opportunities.

In one historical example, the creation of the Social Security Administration, staffed by skilled bureaucrats committed to defending the program using nonpartisan expertise, helped to build the case for the gradual expansion of benefits over time. The Roosevelt Administration won that argument by showing the public and Congress the social and economic value of the program, as well as the ability of the bureaucracy to manage such expansions.

A more contemporary example involves the design of the Consumer Financial Protection Bureau. The bureau’s congressional crafters and early leaders wanted to centralize what had previously been a diffuse responsibility for consumer protection, creating a focal point within the federal government for public organizing around consumer issues. The bureau’s designers also intended for it to have a robust public engagement structure that could facilitate proactive outreach to consumers. Political scientist Mallory SoRelle has found that these design choices were shifting the balance of power surrounding financial policymaking. SoRelle wrote in 2020 that the agency’s design had “clearly improved the fortunes of consumer advocacy groups” and may have been “increasing people’s engagement with the [consumer protection] regulatory system as well.” The Trump-Vance Administration’s efforts to “delete” the bureau may be the strongest validation of its past success engaging the public to regulate the financial industry.

With the expansion of UI benefits, the Biden-Harris Administration had an opportunity to fundamentally reshape the mechanics of the UI system too. Ultimately, the results were mixed, with notable successes but also some missed opportunities, especially looking at the long-run politics of the federal-state UI system.

For decades before the COVID-19 pandemic, experts had been ringing alarm bells about the state of UI administration. Many states were using outdated IT systems no longer supported by their manufacturers and did not hire enough staff to accommodate surges in demand during recessions. On top of the technical challenges, many states had also erected—sometimes inadvertently, sometimes deliberately—substantial barriers to applying for benefits. These barriers included lengthy and dense applications intelligible only to lawyers, complicated and clunky websites, and poor customer support for those with questions.

The COVID-19 pandemic and federal response thus amounted to a perfect storm for poorly prepared states. To help manage the short-run crisis states faced in administering benefits, as well as to address states’ longer-run issues, the Biden Department of Labor launched initiatives using UI modernization funds from Congress. New grants supported work in the states to overhaul IT systems, procure stronger identity verification platforms to prevent fraud, and introduce measures that could make access to the program more equitable.

These grants, especially the equitable access grants, were an innovation for the department and for state UI agencies. Instead of automatically depositing money in state UI accounts for states to use as they saw fit (the typical approach for UI grants), the Department of Labor solicited proposals from states about how they could address gaps in access to and delivery of UI benefits. Each state was eligible to receive a designated allocation of grant funding—but needed to have a proposal for the use of the funding approved by the department. Career UI staff at the department then worked intensively with each state to revise proposals to ensure that states’ applications would address a specific gap in the delivery of UI benefits.

In all, 41 states had approved proposals. They directed the funds to a variety of activities, including streamlining application processes; revamping application websites to make them easier to navigate (including making them mobile-friendly); collecting and analyzing data that could provide a better understanding of gaps in access to UI benefits; and more regularly consulting beneficiaries about their experiences to identify barriers. Many of these reforms will have enduring effects on UI program delivery in their respective states, improving the experience workers have when accessing benefits.

In addition to improving access to UI benefits, these changes may have political consequences. Reforms intended to build stronger consultation and engagement of beneficiaries into the day-to-day operations of UI agencies can make it easier for civil society groups to mobilize beneficiaries in an ongoing way around improving benefits. And new data collection efforts that can better characterize the gaps in access to UI benefits can help build a public case for reform while providing tools for advocates and policymakers to deploy.

But while the UI modernization initiative—and especially the equity grants—may create some future political opportunities for activists and policymakers, the lack of more fundamental reform to the creaky federal-state UI system may well limit its reach. Since its inception, the UI system—like other federal programs run in partnership with states—has created significant inequalities across the country, as states, especially those run by conservative governments in the South, have sought to limit access to the program. Without a stronger floor on state UI policy choices—or a nationalized system—many states have raced to the bottom in cutting taxes and benefits. These cuts have left millions of workers, especially workers of color, without access to the economic security they need to weather spells of unemployment. The cuts also undermine one of the goals of the UI program: to provide stimulus to the economy during downturns. The Biden-Harris Administration’s UI work, stymied by opposition in Congress, did little to alter the fundamental imbalances of political power that sustain these race-to-the-bottom dynamics.

For decades, two of the main obstacles to more generous and more consistent national standards for UI benefits have been businesses and state governments.

Businesses have a substantial interest in avoiding more generous benefits or coverage, since nearly all states rely exclusively on employer taxes to fund their systems. By virtue of their financing of benefits, business groups argue that they deserve pride of place in driving UI policy. Such arguments are often persuasive to state policymakers, and business associations are frequently the most powerful political voice lobbying for cuts to benefits and lower UI taxes.

The interest that individual states have in preserving the status quo is somewhat more complicated and reflects the peculiar evolution of the system. Over time, states have developed large UI agency workforces that have become committed to retaining a state-by-state system with less federal oversight, given that a stronger federal role would necessitate a smaller role for state UI agency staff. State agencies and their workers are well represented in pursuing these interests. In the past, state agency associations have been decisive in beating back efforts to nationalize benefit administration, set stronger national standards, and strengthen federal oversight of the program.

The Biden-Harris Administration’s UI policy—both the temporary expansions and the modernization initiative—did not change the privileged positions of either businesses or state agencies, meaning that political opposition to expanded benefits and stronger national standards is likely to persist in the coming years. Indeed, after the pandemic, and despite the $1 billion investment in modernizing UI systems and improving access to benefits, more and more states are slashing benefits and taxes. One striking example: In 2019, seven states had set their maximum duration of UI benefits below the Department of Labor-recommended 26 weeks. By 2024, that figure had doubled to 14 states, including two—Arkansas and Florida—that set their maximum duration at a miserly 12 weeks.

For policymakers and activists, the lesson is to take advantage of openings for reform, such as economic crises, to rethink program administration in ways that expand, rather than limit, future political opportunities. That could include, as the equity grant initiative demonstrated, structural changes to a program to improve delivery and engage beneficiaries in an ongoing manner. But, where relevant, this playbook should also include efforts to divide or weaken political opponents.

Efforts by the Federal Trade Commission (FTC) to ban noncompete agreements under the Biden-Harris Administration offer one instructive example. In rolling out the ban, the FTC and its outside allies sought to mobilize traditionally conservative, pro-business constituencies in support of the FTC’s proposal, such as small business owners, healthcare professionals, and religious leaders. These unlikely allies helped divide the traditional opponents of labor market regulation in the conservative movement and Congress.

Taking a page from the FTC, the Administration might have taken steps to reduce the political clout of businesses in the UI program as part of its UI reform agenda. Working with Congress, the Administration could have encouraged or required states to collect UI contributions from workers as well as businesses. By including worker contributions in the system—which is how Social Security and Medicare are funded—the Administration could deprive businesses of the argument that they deserve a privileged position in setting UI policy. Another step might have been to encourage or require states to formally include unemployed workers or their representatives in developing UI policy.

Take Messaging Seriously

The last relevant lesson from the pandemic UI experience involves the messages and narratives that emerge around public policies in the political sphere. By tapping into broadly shared values, symbols, and stories, political messages can either raise or diminish the social standing of groups, thereby empowering or disempowering them. To be sure, no one actor is usually able to drive public narratives. But failing to coordinate with allies on advancing a narrative supportive of a policy and its beneficiaries can leave that policy vulnerable to its opposition.

Consider, on the one hand, the example of Social Security old-age benefits. In the decades following the creation of Social Security, politicians and advocacy groups representing senior citizens successfully advanced a shared conception of Social Security beneficiaries as deserving constituents who had earned their benefits through a lifetime of work. That narrative made it harder for opponents to pursue cuts to the program and also made it easier to mobilize senior citizens and the general public in defense of benefits.

On the other hand, consider the mixed public perceptions of the Biden-Harris Administration’s student debt relief efforts. There is a sharp division among Americans over debt cancellation, a split that reflects divided public views on the root causes of the student debt crisis and whether students deserve assistance from the federal government. Political science research has affirmed the potency of conservative critiques of debt relief, showing that conservatives’ efforts to draw attention to assistance flowing to beneficiaries perceived as undeserving—such as higher-income recipients or graduates of elite institutions—were successful in driving down support for debt cancellation. In turn, this may have made it harder for the Administration to reap political dividends from publicly championing student debt relief efforts.

UI policy was one area where, working inside the Administration, I felt a marked shift in narratives that seemed to constrain political opportunities for reform. Back in academic life, I confirmed that these impressions were accurate. I downloaded every news story about UI benefits written during the pandemic and analyzed the content of those stories, as well as who journalists had interviewed. I found that throughout 2020, news stories about the UI system tended to focus on workers’ experiences with benefits. One representative example: In July 2020, the newspaper chain Gannett ran a story in New York about a worker, Myra Colon, who was “insulted by insinuations…that a supplemental $600 weekly unemployment insurance payment [was] discouraging her from returning to work.” Instead, Colon argued that the expanded benefit allowed her to buy food and make housing payments that she might have missed otherwise.

But during the critical period in mid-2021 when the Administration and Congress were debating further expansions and possible reform of the UI system, coverage about UI tended to focus disproportionately on claims that the system was contributing to labor shortages. Moreover, I found that articles were more likely to feature interviews of business owners than workers by nearly a two-to-one ratio. One story distributed by Clarity Media Group in the Washington, D.C media market in May 2021 was typical: It featured a pizza and sub franchise owner who complained that unemployment benefits were making it harder to attract workers and creating “a habit of laziness.” During this same period, news coverage about UI fraud also exploded, often focusing on colorful individual cases—like the rapper who wrote a song bragging about defrauding unemployment aid in California—while neglecting the more systematic organized crime efforts states were fighting.

I find that this changing news coverage may have made a considerable difference in turning public opinion against the UI system and its beneficiaries. In a recent experiment, I found that showing survey respondents news stories about individual UI fraudsters turned the respondents against recipients in general. Those stories also led them to report higher perceived levels of fraud and waste in the system and lower support for expanding benefits. Critically, I find that the fraud-centered news stories appeared to tap into racialized stereotypes about lazy and undeserving recipients of government assistance that many members of the public already found resonant—that is, fraud-centered news stories had the biggest impact on survey respondents who held stronger racial stereotypes about Black Americans.

Facing this hostile media landscape, the Administration ultimately decided to cede the narrative battle. Instead of countering narratives around labor shortages and outsized fraud, President Biden gave widely covered remarks describing steps the Department of Labor was taking to encourage workers to take available jobs. Even though research at the time had found that extended UI benefits were not major contributors to labor shortages, the Administration gave into the frame centered on fraud and shirking.

That same pressure from hostile narratives made the Administration skittish about pursuing subsequent extensions of UI benefits, even when concerns about a new COVID surge were brewing in the summer of 2021. So too did the pressure foreclose opportunities for the Administration to work with Congress on larger and much-needed reforms to the UI system—reforms that President Biden had indicated were priorities in his earlier budget proposal to Congress.

What I took away from this episode was the power of narratives to change political possibilities. It mattered that the cable news shows my White House colleagues tracked closely were beating the drum about UI benefits discouraging work and overwhelmingly flowing to fraudsters. And it also mattered that the growing conservative media ecosystem was amplifying stories that tapped into racial grievances and stereotypes and mobilizing opposition to further relief. Together, these shifts may help explain why the public mood had noticeably turned against further extensions.

What could have been done differently—and what can future policymakers learn? We cannot know for sure if a more concerted effort by the Administration and its allies to proactively message about the benefits of expanded UI would have made a difference (and it is unlikely that such efforts would have penetrated the conservative media ecosystem). But analysis I have conducted suggests it might have had an impact. In an experiment I ran earlier this year, I found that showing Americans media stories about UI benefits centered on worker experiences made them more supportive of UI benefits and of UI recipients. Those worker-centric stories also made Americans more supportive of comprehensive reform to the UI system that would expand benefits and eligibility.

The lesson is that it is not enough to simply pursue good policy and hope that the benefits of that policy will be covered favorably by the media and lead to positive narratives about target populations. Politicians and their allies need to craft and deploy positive messages about their policies—and be prepared to counter negative narratives that emerge from opponents and those skeptical of policy advances. And, over the longer run, progressives need to grapple with a conservative media ecosystem that mobilizes racial grievance in ways that undermine the expansion of economic opportunity and support.

Deliverism Defeated?

Are the deliverism skeptics right? Does the Biden-Harris Administration’s failure to reap political dividends from its economic policy agenda spell the death of this theory of politics? I believe this diagnosis is too hasty and incomplete.

What the Administration’s experience has shown us is that for a policy agenda to have downstream political benefits, policymakers need a different toolkit for policy design and implementation. These tools include cultivating the right conditions for sustained, durable organizing by civil society groups with reach into the communities likely to be affected by policies; ensuring that policies create sufficient capacities in government to enable further political opportunities; using policy and its rollout to weaken or divide opponents; and ensuring that policymakers and allies can advance narratives that improve, rather than damage, perceptions of policies and their target populations.

None of this is likely to be easy for policymakers, especially at the federal level. I saw firsthand how the structure and norms for decision-making within government agencies often made it challenging to deliberately consider these tools. It was uncommon, for instance, to think about civil society groups during policy formulation—and still less common to discuss how different policy options might enable more organizing by such groups. Conversations about the role of civil society were more likely to happen after a policy had been designed, when engagement and outreach staff were tasked with selling the policy to stakeholders. In a similar vein, conversations about narrative tended to happen among communications and political strategy staff, not necessarily those tasked with policy design. That all has to change.

Moreover, effective policymaking as power-building will also require planning across initiatives (and agencies) so that an administration can integrate disparate individual policies into a coherent whole that can be more effectively communicated to the public by organizers and government when appropriate. This approach will require coordinated legislative and executive action, since as this essay has argued, temporary and limited federal actions often complicate efforts at power-building. We will also need to grapple with a reactionary federal judiciary—above all the conservative majority on the Supreme Court—that is increasingly constraining the ability of federal agencies to implement legislation.

All of these tools will thus require changes in the internal organization of policymaking in the executive branch, both within agencies and in the White House. Democrats should not wait, however, to try out these new strategies for governance solely at the federal level. They should begin by testing them at the state and local levels in the coming years.

The absence of a Democratic debate about policymaking as power-building is ironic. The Trump-Vance Administration has already made clear its intentions to deploy its own version of such an approach. Whether it is using Schedule F to “deconstruct” the federal agencies right-wing activists oppose, weaponizing the Internal Revenue Service against nonprofits that speak out against the Administration, or steering grants and contracts away from “woke” universities and charities, the MAGA movement will be using policy levers to build durable governing power at the expense of its liberal opponents.

MAGA efforts to undermine democratic accountability and norms, weaken the capacity of the federal government to respond to public needs, and inflict harm on vulnerable populations should not be seen as the mirror image of progressive policymaking as power-building in support of a more egalitarian democracy and economy. Nevertheless, Democrats face a choice: Learn to build power through policy or cede political ground to those who will use such tools to undermine democracy.