State capacity is in vogue. For decades, scholars theorized and debated the concept in academic obscurity; now, talk of state capacity saturates popular policy and political discourse. And whereas the scholarship focused on states’ ability to pursue official goals autonomously from powerful social groups, today’s purported state capacity liberals in the popular press—including Ezra Klein, Derek Thompson, and Matthew Yglesias—diagnose a slightly different problem. Rather than underdeveloped state institutions, they focus on allegedly overdeveloped state institutions, specifically the procedural constraints that liberal democratic states have placed on themselves, such as environmental review for federal infrastructure projects and local planning approval for construction projects. Echoing the deregulatory mantra that they also advocate for the private sector, state capacity liberals argue that liberating the state from hidebound bureaucratic proceduralism will unleash newfound prosperity, or “abundance,” for the American people.

They rely on anecdotes to support the story of a state that obstructs the private sector and shackles itself. The paradigmatic example of government getting in its own way for these state-capacity liberals is California’s failure to build a high-speed rail system. Similarly, lengthy environmental review processes supposedly throttled the federal government’s capacity to fulfill a legislative mandate under President Biden to rapidly build out renewable energy. Meanwhile, public rules constraining the expansion of charter schools (rather than, say, gross inequalities in education funding) are to blame for poor school performance. But careful study, whether on high-speed rail or environmental review, contradicts this simple story of public regulations as barriers to abundance. In the case of California’s high-speed rail development, for instance, inadequate fiscal support and American inexperience in building such projects—in contrast to decades of experience in highway construction—have been the principal obstacles. But the anti-procedural narrative has taken hold, especially in elite circles.

Looming over the entire debate is a pervasive anxiety that China is (allegedly) beating the United States in economic and geopolitical competition because its authoritarian state is stronger, nimbler, and less encumbered by procedure than ours. In this regard, current state capacity discourse mirrors a similar anxiety liberals felt during the Great Depression, when Americans wondered if liberal democracy was too unwieldy and decadent to survive competition with a European fascism that, if nothing else, was purported to “make the trains run on time” (it didn’t, but people thought it did).

We are ardent believers in the power of government to improve people’s lives, and appreciate the frustration with unnecessary delays. In some cases, especially exclusionary zoning and other land use rules in wealthy suburbs, the current legal restrictions are in fact excessive. However, the proper answer to government dysfunction and the authoritarian threat at home and abroad is largely the same today as it was during the New Deal: Rather than railroad the public with top-down directives or megaprojects imposed by Washington and state capitols, we must renew our national commitment to democracy and public participation with ambitious acts of bold economic governance.

The Trump Administration’s ability to move with lightning speed to impose unwanted infrastructure like Alligator Alcatraz on unwilling host communities, militarize cities with masked police and troops, and increase surveillance of Americans makes clear the urgency of balancing the state’s capacity to act decisively with the need for democratic legitimacy. “State capacity” is not inherently good: The what and how of state action matter.

Moreover, in the medium and long term, the tradeoff between state capacity and democratic participation disappears—and even reverses. Public input into government decisions, including in environmental review procedures, not only restrains the worst tendencies of unaccountable state power but also enhances the capacity of the state itself, by providing valuable feedback to state technocrats about what is working and what isn’t. While turning off these feedback channels may generate a sugar high of quick results, doing so, as state capacity liberals propose, dooms effective governance further down the road. Top-down technocracy is brittle, and availing ourselves of shortcuts around the public is ultimately self-defeating. In the words of Henry Wallace, New Deal agrarian technocrat extraordinaire and FDR’s second vice president: “There can be no worthwhile result unless there is just as great a volume of thought flowing into Washington as there is flowing out of Washington.” Well-designed, effective democratic procedures are how that happens.

Understandings of State Capacity

Just about everyone agrees that the “state” having “capacity” to implement desired policies is essential to public order, economic health, and good governance. But what does that mean? During the Clinton era, left neoliberals argued for “reinventing government,” which paradoxically was supposed to increase the effectiveness of the state by shrinking it down. The shrinking proceeded apace over the next few administrations. Thus, when Donald Trump took office in January 2025, there were fewer federal employees relative to the U.S. population than there were when Bill Clinton assumed the presidency in 1993. And that was before DOGE and Elon Musk. Turning away from the capacious bureaucracy and (by American standards) robust interventionist state that liberals had built during the New Deal era, the Clinton neoliberals imagined a state that would “steer more, row less” and outsource state responsibilities to private contractors presumed to be more efficient.

In a competing vision, advocates for a “Green New Deal” have called for a revitalized Rooseveltian state that directly invests in and owns energy infrastructure and utilities, builds affordable non-market social housing, and firmly guides fossil fuel capital into ultimate extinction. The Biden Administration’s initial “Build Back Better” agenda rhetorically invoked and partially revived the New Deal vision, proposing that the state take a more powerful role in the economy through a large infusion of government spending. However, the scaled-back version of “Bidenomics” that ultimately passed Congress as the Inflation Reduction Act relied more heavily on steering—via tax and other incentives for private businesses (with some pro-labor and other strings attached)—than on direct state investment and provision.

Meanwhile, the well-financed proponents of the “abundance” agenda argue for expanding state capacity in precisely the same terms that they advocate expanding private business capacity. They call for the liberation of both corporation and state from the shackles of bureaucratic “blockages,” public “veto points,” and overly proceduralist public regulations. These choke points, they claim, allow highly motivated self-interested actors and nefarious “groups” (homeowners, environmental justice organizations, nonprofits, and labor unions) to stand in the way of governments accomplishing what they were elected to do. Within this orbit, there are even “state capacity libertarians,” who believe that a strong state is necessary to create and foster free markets (a point already conceded long ago by libertarians like Friedrich Hayek).

Can all of these competing visions claim to be versions of “state capacity”? A quick detour through the history of the concept is in order.

While he did not use the term, German social scientist Max Weber, the seminal theorist of the modern state, saw two features as essential to what we would now call “state capacity”: the state’s monopoly on coercive power and the creation of a professionalized and permanent civil service bureaucracy to administer the state’s policies. At a minimum, a Weberian state must have the sole right to legitimately apply force, and it must be staffed by trained professionals who are insulated from social pressure and follow formal rules and procedures rather than the whims of charismatic, fickle leaders. (Weber, ambivalent about the effects of the modern state on society, described a resulting “iron cage” of bureaucratic rationality that ultimately severely circumscribes individual freedom.)

Following Weber, subsequent understandings of state capacity have also emphasized the centrality of a highly functioning bureaucracy. In one of the first usages of the term “state capacity,” social scientists Theda Skocpol and Kenneth Finegold attributed the success of the New Deal-era Agricultural Adjustment Act to its ability to act through a preexisting bureaucracy, and the failure of the contemporaneous National Industrial Recovery Act to its lack of that good fortune. That bureaucracy, the U.S. Department of Agriculture (USDA), had been created by Congress in 1862, and it had established leadership, organizational purpose, and years of institutional learning under its belt. This made it unique in the federal government, given the general weakness of national governing institutions at that time. According to Weber, Skocpol, and Finegold, then, a highly trained, respected, and permanent staff for government agencies would seem to be at the core of the concept of “state capacity.”

But bureaucratic expertise is built through practice. As economists Mariana Mazzucato and Rosie Collington argue, experience cannot be purchased: Governments that outsource functions to private contractors lose the ability to learn. Organizational learning requires doing; it is not a transaction, but “an incremental and collective process.” Once lost, capabilities are difficult to regain. (Indeed, Klein and Thompson emphasize the importance of in-house state expertise in their book, Abundance).

Elon Musk’s DOGE program, in which untrained computer whizzes with nicknames like “Big Balls” set about vandalizing and dismantling that professional bureaucracy, offers one solution to frustration with government “red tape.” However, far from representing an obstacle to “getting things done,” bureaucracy is essential to effective, timely action, as the DOGE experience seems to have demonstrated to millions of Americans. While it may be tempting to blame unreasonable permitting delays on the very existence of permits, a wiser approach is to ensure there are sufficient trained staff to process them quickly. Indeed, if we are to have more state capacity, we need an even bigger, more permanent, and more professional bureaucracy. Regulators and legislators’ current reliance on external lobbyists to come up with policy ideas and even to write legislation highlights the extreme deficit in this key dimension of state capacity in the United States. The relatively low pay of American civil servants, which keeps people moving through the revolving door to more lucrative industry and lobbying posts, only makes this worse.

The What and How of State Capacity

The scholarship on state capacity emphasizes the plurality and unevenness of state capacities. For example, states can strengthen their capacities in some areas, such as repression, while self-consciously weakening their capacities in others, such as corporate regulation. States exercise their power for different ends and use assorted means, some good, some bad. Some state agencies deliver health care for millions, while others target working-class people through tax audits and imprisonment. Moreover, we care not just about the state’s capacity to act, but also about the democratic legitimacy of those actions. States make some decisions through democratic means, such as legislation and regulation based on public input and consultation, and others through undemocratic methods, such as court decisions. And while state capacity entails the ability of the state to pursue its own goals autonomously from powerful social groups such as large corporations, there are other social groups, like poor communities suffering the effects of environmental racism, that do not have enough power to influence the state.

Taxation, spending, regulation, and public provision are all aspects of state capacity. A prerequisite for a state that meets even Weber’s minimal criteria would be fiscal capacity: the ability to collect taxes and direct public resources to the state’s desired ends. On the taxation front, U.S. state capacity is clearly heading in the wrong direction, with corporations and wealthy individuals openly pursuing a multitude of tax avoidance strategies with little fear of negative consequences. At the sub-federal level, states and municipalities compete with one another for private investment via offers of deregulation and subsidies, allowing powerful corporations to choose the level of regulation and taxation they desire. On the spending side, the federal government’s reliance on private contractors and unwillingness to use its bargaining power as a large buyer means it has limited control over military procurement costs.

Nonetheless, it would not be accurate to say the United States lacks state capacity in taxation and spending in general. In fact, the U.S. state does seem to retain high fiscal capacity in some areas: While the rich and corporations routinely engage in tax avoidance, the Internal Revenue Service vigorously audits Black Mississippians, and local governments eagerly seize the assets of the poor through civil forfeiture. The federal government also excels in weapons production and warmaking, spending orders of magnitude more than any other country on increasingly high-tech weapons while engaging in military adventures (and supporting the militaries of foreign allies) around the world. These displays of state power suggest that in other realms, perhaps it is not generic “capacity” that is lacking, but the will to use it in ways contrary to the preferences of the wealthiest and most powerful.

Shaping markets and regulating firms are also aspects of state capacity. Defining and protecting property rights while setting out rules for business conduct—such as antitrust, consumer protection, and labor regulations—are key functions of the state. These rules restrict certain practices, such as firing workers for attempting to form unions, and mandate others, including that mortgage lenders provide prescribed disclosures to prospective homeowners. Today, while the federal government has granted property-like rights over a range of intangible assets and ensured the strong protection of this intellectual property around the world through trade deals, other elements of this aspect of state capacity are weak in practice. Compliance with labor laws has become all but voluntary in recent decades, and major antitrust suits can take years to resolve.

Finally, since World War II, people around the world have looked to their governments to directly provide certain broadly beneficial social goods. Virtually alone among industrialized nations, the United States does not ensure that health care is provided to every resident. While “Medicare for All,” or socialized medical insurance, currently sets the frontier of progressive health-care policy, the Department of Veterans Affairs (VA), an example of direct public provision of care delivery, quietly provides what our colleague Phil Longman has called the “best care anywhere.”

The VA is an outstanding example of democratic state capacity in action. As Longman points out, according to conventional wisdom, “[T]he VA should offer the worst care anywhere: [I]t’s a gigantic, unionized bureaucracy, micromanaged by Congress and political appointees, and beset by an uncertain budget, an aging infrastructure, and a legacy of scandal.” Yet, the VA pioneered integrated electronic medical records and adopted quality metrics based on hard science decades ago—all instituted from the bottom up, by career staff interacting with patients and often acting contrary to directives from Washington. The VA has been lauded for its exceptional safety record, deployment of evidence-based medicine, effective prevention and wellness programs, and unequaled use of electronic records and other information technology. Unlike private insurers, the VA has powerful incentives to invest in coordinated care and prevention thanks to its lifetime relationship with patients and lack of a profit motive. The only area it seems the VA does not excel in is patient billing.

On the other hand, further illustrating the existing capacity of the American state for good and ill, federal, state, and local governments are extraordinarily effective at the “public provision” of carceral violence (if not so much at the solving of crimes and doling out of criminal justice). For all the complaints about regulatory barriers to infrastructure projects, the federal government recently managed to fund the state of Florida’s construction of a torture prison in the Everglades in a matter of days, a product of Governor Ron DeSantis’ abuse of emergency powers. The United States operates the largest network of prisons and detention centers in the world. Clearly, when there is the will to do so, the U.S. state can indeed build, and build quickly. In contrast with its unwillingness to discipline the private health-care system with public power, the United States is all too willing to use violence against mostly working-class immigrants and inmates. The country has the capacity to provide health care and other essential social goods to all; what is lacking, again, is the will to take on powerful and wealthy interests.

State Capacity as Democratic Self-Governance

Skocpol and Finegold, in their seminal paper on state capacity in the New Deal, pointed out that state capacities are not static, but can change over time. The very economic interventions that demonstrated the USDA’s unique state capacity also had the effect of strengthening the conservative American Farm Bureau Federation, which then used its power to weaken later New Deal attempts at progressive redistribution and marketcraft. A similar risk looms large today. Without adequate public accountability and guardrails, policies to promote the construction of data centers, for example, may increase the power of big tech corporations to capture state agencies and even legislatures, blocking the state’s ability to act contrary to corporate interests in the future.

Indeed, the view that we must liberate the state from procedural hurdles in order to increase state capacity risks blinding us to the role of public feedback in fostering effective governance. Democratic accountability and two-way information flows are essential feedback mechanisms that enable correction of mistakes and foster effective on-the-ground implementation of state policies. When imposed without public input, top-down mandates, even if they are enacted with the best of technocratic intentions, ultimately lead to bad policies. This not only undermines the legitimacy of the state, but also, by reducing trust in state actions, risks triggering public resistance rather than the cooperation necessary for policy implementation. The result is the erosion of state capacity.

Important figures in the New Deal understood this. Key New Dealers, especially in the USDA, were inspired by the pragmatic democratic philosophy of John Dewey, who emphasized participatory democracy as a key component of democratic citizenship. They sought to combine the on-the-ground knowledge of citizens with technocratic expertise in Washington to craft development policies that not only won strong public support but were improved through public input as well. While some in the United States looked to the Soviet Union’s rapid industrialization in the face of the Depression as a model for economic governance, these New Dealers disagreed. Undersecretary of Agriculture M. L. Wilson, who spent time on a Soviet collective farm in the 1920s, scoffed at the top-down planning methods of the Soviets in 1930: “Those little communists strutting! It has to be planned from the ground up.”

USDA administrator Howard Tolley wrote in 1938 that “Planning [what we would now call “development”] in a democracy called for planning by the people, not just for the people.” Another USDA administrator, Carl Taylor, added that “[O]ur various levels of democracy, from local neighborhoods and communities on up, must be tied together in national planning and action.” If they were not, he said, “the Federal Government will tend to be an autocratic and not a democratic device.” Despite (or perhaps more accurately, because of) the urgency of economic recovery from the Great Depression, these New Dealers—many of whom came from farm communities and understood the value of both experiential and technocratic knowledge—sought to take maximum advantage of public input, including that of the poorest farmers. As detailed in Jess Gilbert’s Planning Democracy, the USDA brought together social scientists, adult educators, and hundreds of thousands of rural people. Social scientists fanned out across the country and helped direct local citizen planning groups while also offering classes in agricultural economics and even philosophy. These participatory groups then conducted research that culminated in land use maps and research reports that provided the basis for planning.

Outside of agricultural policy, democracy and local participation were important norms guiding rural electrification. In 1935, only around one in ten farms in the United States had electricity. The federal Rural Electrification Administration (REA) provided low-cost credit and technical assistance to rural communities seeking electric service. Power lines in rural areas were built not by the federal government itself but principally by consumer-owned and -operated electric cooperatives set up by farmers and other rural residents. In 1938, Michigan Governor Frank Murphy hailed rural electrification for “demonstrating that democracy is not just a political ideal but that it can be made to work in the economic phase of life as well.” Federal support for rural electrification was extraordinarily successful: Twice as many farms were electrified in the first five years of the REA as had been in the previous 50 years. By the mid-1950s, more than nine in ten farms had electricity.

By contrast, post-New Deal officials like Robert Moses sought to impose projects from above on often unwilling citizens. After early success, Moses’s urban planning efforts in New York started to unravel when the public rebelled, especially against his ill-fated plan to drive a road through the middle of Washington Square Park. Elected officials subsequently reduced his powers and he lost his administrative offices. Given this history, shouldn’t we be concerned with unconstrained public officials, especially in the Trump administration, trying to double down on car-centric infrastructure development?

Some constraints on state capacity are more democratic than others. Just as “state capacity” can mean both the power to carry out extrajudicial executions of alleged drug traffickers in the Caribbean and the ability to provide lifesaving care to cancer patients, the constraints on state capacity are not of a piece either. For example, they include both the judiciary’s power to veto state action and Congress mandating notice-and-comment procedures for certain regulatory actions. In the former, the constraint consists of unelected, generalist judges overriding the decisions of legislatures and federal agencies, while in the latter, the “constraint” of the public sharing its experience and expertise enables administrators to make better-informed decisions.

When the federal judiciary threatened to derail the New Deal, Roosevelt responded by attempting to pack the Supreme Court. He failed, but the mere threat led to a chastened Court restraining itself from its own historical pattern of overreach. In the following decades, Congress, worried about the large expansion in the power and reach of the federal government during the New Deal and World War II, imposed its own procedural constraints on federal action. It has done so with the passage of three important laws in the postwar era. Two of these interventions have enhanced state capacity in different ways, while the third has largely ended up favoring corporate interests.

All Constraints Are Not Equal

First, the Administrative Procedure Act (APA) was enacted in 1946, partly to codify New Deal practice and partly to protect business interests. The APA establishes general procedures for federal agencies, in addition to those included in each agency’s enabling statutes. When an agency wants to issue a regulation using informal rule-making procedures, it must publish a “notice of proposed rule making…in the Federal Register” that describes the “substance of the proposed rule” and “reference to the legal authority under which the rule is proposed.” Further, the agency must “give interested persons an opportunity to participate in the rule making through submission of written data, views, or arguments with or without opportunity for oral presentation.” This statutory provision establishes the familiar notice-and-comment process for rule-makings. Once it has reviewed the public comments, an agency can publish a final rule together with a “concise general statement of…basis and purpose”—an especially important requirement with an administration that is inclined to ignore the law and to do as it wants.

The APA also governs agency adjudications, which decide issues such as whether a business violated labor law or an individual was improperly denied Social Security benefits. In these matters, agencies must hold a hearing and in the notice of hearing provide the factual claims and legal authorities at issue. An employee of the agency must issue an initial decision summarizing the facts and describing the legal bases for the decision. The process often functions as a less formal version of litigation in court.

In adjudications, the APA ensures basic due process for individuals and businesses and requires reasoned decision-making by agencies. In rule-making, it generally mandates a multistep process that grants members of the public the opportunity to participate and share their perspective and expertise. To be sure, structural inequalities mean that certain groups, such as big businesses, are better represented than others, like the working class and poor. Yet, the existence of these deeper inequalities does not warrant abandoning notice-and-comment procedures any more than higher Election Day turnout among wealthy, white voters justifies abandoning elections. Moreover, administrative processes are far more inclusive and democratic than court proceedings, which are expensive, rigid, and often confusing. (Not to mention, the state does have the ability to boost working-class voices, such as by fostering people’s ability to form collective organizations like labor and tenant unions.) The notice-and-comment process upholds the public’s right to petition the government and can promote more informed, democratic policymaking relative to a system in which agencies are more insulated from public input.

Second, the National Environmental Policy Act (NEPA), which became law in 1970, requires federal agencies to evaluate the environmental impacts of major actions. These actions include issuing permits to use federal lands or navigable waterways; providing fiscal aids, such as federal loans, to private actors; and directly developing infrastructure projects. The law is procedural: It does not mandate or prohibit any substantive decision. Instead, it applies the “look before you leap” principle by requiring government agencies to fully open their eyes before undertaking or authorizing actions that will have significant environmental or social impacts.

When an agency determines an action will have a significant environmental impact, it must prepare a detailed environmental impact statement (EIS). In other instances, an agency will determine from a more abbreviated review that an action will have no substantial impact, which concludes the NEPA process. Further, agencies, on their own initiative or under directive from Congress, have granted categorical exclusions to certain activities, which means they do not require NEPA review. When federal action is likely to require the time-consuming preparation of an EIS, this threat itself can motivate agencies and private developers to redesign projects so as to reduce their environmental impacts, suggesting that NEPA plays a useful channeling function and indirectly generates important public benefits.

Contrary to the assertions of NEPA critics, the law is not commonly holding up renewable energy projects and other socially beneficial infrastructure. In a 2014 report, the Government Accountability Office estimated that 95 percent of actions that implicate NEPA are entitled to a categorical exclusion, with only 1 percent requiring an EIS. At the U.S. Forest Service, an agency that relied less on categorical exclusions than most other federal agencies, more than 80 percent of NEPA actions between 2004 and 2020 still triggered an exclusion, and only 2.1 percent required an EIS. The interminable EIS is the rare exception and, importantly, can often be avoided by redesigning a project to minimize environmental impacts.

Both the APA and NEPA are constraints on federal power that ultimately help the government make better decisions with greater democratic legitimacy. However, some restraints do impede state capacity without improving the quality of decision-making or enhancing its democratic character. Rather, they represent certain parts of government and private interests attempting to curtail and influence administrative authority. Since 1980, the Office of Information and Regulatory Affairs (OIRA), housed in the White House’s Office of Management and Budget, has played a clearinghouse role in federal rule-making. Executive agencies that issue rules generally must submit their rule-making agendas and major proposed rules for OIRA review. Created by statute at the end of the Carter Administration and directed by presidents—not by legislation—to undertake substantive review of regulations ever since, OIRA reviews these rules to ensure they are harmonious with the President’s agenda and pass cost-benefit analysis, as conceptualized and applied by OIRA staff.

This process is questionable in terms of both expertise and accountability. OIRA is a small generalist office whose staff is unlikely to possess more substantive and legal expertise in, say, environmental law than the Environmental Protection Agency (EPA) or in federal housing policy than the Department of Housing and Urban Development. And while OIRA arguably enhances presidential control over administrative agencies, that does not mean it enhances public accountability. Entirely separate from OIRA review, the President already exercises important control over agencies by making budget requests and nominating agency leadership based on shared priorities. Meanwhile, agencies also answer to Congress, and OIRA may improperly interfere with their ability to carry out the statutory mandates established by the national legislature.

OIRA review appears to serve mainly as a veto point for business interests and sympathetic staff in the executive branch. In a 2006 article, legal scholars Nicholas Bagley and Richard Revesz bluntly described OIRA as “a sheltered haven for regulated entities to advance a deregulatory agenda.” In comparison to the relative transparency of notice-and-comment rule-making, OIRA decision-making is opaque and not open to members of the public. During the Obama years, rules enacted by agencies like the EPA were held up or weakened by OIRA, apparently at the behest of powerful industries that stood to lose. An administration publicly committed to environmental protection was privately undermining its stated ambitions.

The Role of the Courts

There are also judicial constraints on state capacity. By developed nation standards, the United States has an extraordinarily powerful judiciary. It is officially an independent third branch of government and can invalidate the actions of the legislative and executive branches on constitutional, statutory, and evidentiary grounds. In recent years, the Supreme Court has restricted the power of agencies on constitutional and quasi-constitutional grounds.

The Supreme Court’s major questions doctrine represents one of the most aggressive assertions of judicial power in recent times. In a 2022 case called West Virginia v. EPA, the Court ruled that a federal agency can regulate a question of “vast economic and political significance” only if Congress gave it clear authority to do so. On these grounds, the high court invalidated the EPA’s Clean Power Plan, which would have regulated greenhouse gas emissions from power plants. Never mind that Congress had granted the EPA broad power to regulate all sorts of air pollutants in the 1970 Clean Air Act. Based on its subjective determination, the Court ruled that this power did not extend to regulating a major contributor to global climate change. Under the guise of humility and respect for the separation of powers, the Court overrode the judgment of the EPA and narrowed the broad delegation of policymaking power by Congress. Judiciary über alles.

The major questions doctrine is a flexible tool for the Supreme Court, or more specifically its conservative majority. The Court has offered little guidance on what constitutes a “major question.” Five justices can invalidate administrative actions they do not like for implicating questions of “vast economic and political significance” and uphold those they do like, in part, for involving only “minor questions.” The indeterminacy of the doctrine means agencies now operate in an environment with even more uncertainty than before. After a multiyear rule-making, the courts may hold that an agency improperly regulated a “major question” without clear authorization from Congress. In such an environment, even a bold regulator may understandably shy away from touching anything that could be construed as “major”—such as climate change, worker disempowerment, or predatory lending—and reserve their attention for matters of minor import.

This judicial impingement on state capacity is hard to defend on the grounds of either expertise or democratic governance. While OIRA suffers from deficient expertise and limited public accountability, these infirmities are even truer of federal judges. Most cannot claim any sort of expertise in areas such as environmental science or consumer protection, and by design, they are insulated from popular politics. Now empowered by the major questions doctrine, these life-tenured, nonexpert government officials can strike down policies they find objectionable and more broadly deter aggressive action by the federal government.

It should be clear by now that not all constraints on administrative governance are alike. Simply calling for the removal of “blockages” on administrative action or “red tape” in general erases the qualitative differences between types of constraints. Some procedures, like those required by NEPA, promote more democratic and better-informed government decision-making. Launching major infrastructure undertakings before studying impacts seems imprudent, at a minimum. Further, prospective administrative study is superior to after-the-fact dispute resolution through piecemeal litigation. The history of twentieth-century infrastructure development—whether dam construction that destroyed fish stocks in the Pacific Northwest and displaced tenant farmers in the Southeast or highway building in poor urban neighborhoods that functioned as a tool of segregation—also cautions against development without an advance examination of impacts and conscious planning to reduce adverse effects and minimize waste.

Democratic State Capacity in Action

The 2024 ban on noncompete clauses issued by the Federal Trade Commission (FTC) encapsulates some of the most important strengths and weaknesses of the contemporary federal government. Noncompete clauses restrict the labor market mobility of an estimated 30 million workers. For instance, a doctor bound by a noncompete could be barred from joining a rival practice or opening her own clinic in the same city for 18 months after leaving her current employer. The practical effect of these contracts is to keep workers in place, even when they have better options. In January 2023, the FTC proposed a ban on noncompete clauses for all workers going forward.

This proposal attracted 26,000 public comments, of which more than 25,000 were supportive of the FTC rule. Thousands of people shared their personal experience with noncompete clauses and how they impede career progression and delay or deny entrepreneurial dreams. Through the comment process, the FTC and the public acquired a fine-grained understanding of how these contracts work in practice and their economic, psychological, and social consequences. The comments were supplemented with letters from pro-worker and pro-employer organizations sharing their own expertise on noncompete clauses and related contracts. Informed by this public input, the FTC published an improved and refined 570-page rule in April 2024 that stood to free tens of millions of workers from noncompete clauses and related restraints, such as training repayment agreement provisions.

Yet, this rule-making also reveals the weaknesses of the American state, especially its comparatively progressive parts such as the FTC during the Biden years. First of all, the rule never took effect. In August 2024, a judge in the Northern District of Texas held that the rule was illegal and blocked it nationwide. She concluded that the FTC lacked both the statutory authority and the evidence to issue the rule. This decision is hard to square with the plain language of the FTC’s enabling statute, which gives the commission the power “to make rules and regulations for the purpose of carrying out the provisions” of the Federal Trade Commission Act, including regulating and prohibiting contracts as “unfair methods of competition.” Indeed, a federal court of appeals previously characterized this language as being “as clear as it is unlimited.” Further, the FTC had reviewed the extensive body of quantitative and qualitative evidence and found that it supported a complete prohibition. Its decision was not “arbitrary and capricious,” as the judge concluded.

Moreover, the rule took far too long to complete. The noncompete issue had been on the FTC’s radar since at least March 2019, when our organization led a coalition that petitioned for a blanket regulatory prohibition on noncompete clauses. The final rule, as noted above, came out in April 2024. This hardly represents government efficiency. To be sure, prudence explains some of this delay. The FTC appears to have undertaken the rule with full awareness of the legal and political environment, publishing hundreds of pages of analysis and relying on both neoclassical economics and political economy to support its decision. This belt-and-suspenders approach did not save the rule from hostile judges who treat laws like the APA as a license to reject rules they find personally objectionable and an aggressive employer class that jealously guards its prerogatives.

Yet, this is likely not the full story. The half-decade-long process suggests that the FTC’s own capacity to write regulations and execute major policy initiatives is inadequate. We can only speculate as to the reasons, but we suspect understaffing and lack of planning and prioritization at the top may have contributed to the delay.

State capacity is critical for building a just and sustainable society. But vague calls for increased state capacity shouldn’t blind us to the specifics of what the state does and how it does those things. The United States has extraordinary capacity to coerce and kill people at home and abroad but relatively weak authority to police corporations and tax the wealthy. Immigration and Customs Enforcement and the Consumer Financial Protection Bureau are both examples of state capacity, but they are entirely distinct in their missions and methods. The Trump Administration is enthusiastically augmenting the power of the former and destroying the latter. This stark contrast shows that state capacity by and for the multiracial majority is radically different from state capacity that disciplines and cages them.

The United States faces enormous economic challenges. Tackling them will require not just bold state action, but democratic structures that secure public buy-in and support for a progressive economic vision. As Ben Beachy recently explained in these pages with regard to climate policy, an “everything bagel” strategy that actively engages and empowers stakeholders like unions, advocates, and aligned companies rather than short-circuiting procedures of democratic accountability will have a better chance at surviving both legislative sausage-making and policy implementation. In many cases this will mean introducing more process, not less. It is vital for the U.S. state to have the capacity to impose policies on recalcitrant powerful interest groups, like large corporations. But there are also social groups who don’t hold enough power, and their active engagement in policymaking will foster not only more egalitarian outcomes, but a more effective government overall.