Fixing Federalism

This article was previously published as part of The Reconstruction Papers. The entire book is available in print and as a PDF.

There is a persistent myth on the right that the federal government long ago crushed the independent power of state governments. Some blame the civil rights movement, which by necessity involved the use of federal power to bring Jim Crow regimes to heel. Others go back to the New Deal, and the birth of the vast federal executive branch. Still others believe this tragedy began with the outcome of the Civil War. You may note that this provides a convenient angle to cast aspersions on all three of these historic moments without necessarily claiming to be against racial equality, social safety nets, or the abolition of slavery.

Whatever the origin story, whatever the reason for adopting those stories, what they all have in common is that they are wrong. States have substantial power in our system, far more so than subnational units in most other countries. Moreover, local governments wield tremendous power as well, despite not holding the same standing as states in constitutional law. In part, this is because the Supreme Court has sided with state and local governments in some cases where the federal government attempted to encroach upon their turf. But conservative political scientist Martha Derthick emphasized a more important reason for this persistence: State and especially local governments have far more administrative capacity than the federal government, and more often than not the federal government has sought to keep things that way.¹

We have entered what is perhaps the most profound crisis of our system since the Civil War. On the one hand, a federal government controlled by Trump and his cadre is running amok illegally impounding funding required by Congress and sending masked agents to kidnap civilians in broad daylight. On the other hand, interstate dynamics, and the dynamic between states and the federal government, have taken on many of the features associated with the fugitive-slave controversy. Bounties are set for catching state residents who cross state lines in order to take advantage of differences in the laws. Local police are ordered by state governments not to cooperate with federal law enforcement. The fragile yet functional working relationship between political entities in our system appears to be coming apart completely.

The causes of this crisis can be traced back decades. Federal law has piled up unfunded mandates, commands given to states to take on burdens that the federal government will not pay for. On the other side, what is funded by the federal government is often credited to the elected officials there instead of the state and local officials who may put it to use. Weak party organizations and strong partisanship put our system at the mercy of a minority of voters who are active in primary elections rather than incentivizing parties to focus on governing for majorities. And our two-at-most party system diminishes political competition and cripples the basic cooperation between levels of government needed to provide essential public services.

The federalist system of government can work, and work well. We may have been a pioneer, but many other nations have experimented with variations of it in the centuries since. To make a serious attempt at reconstructing our system, we must go back to the drawing board with what we have learned at home and abroad.

I propose three basic principles for this overhaul:

• State governments should be greatly strengthened at the expense of local governments.
• The federal government should more aggressively act as the guarantor that state governments are democratic in their design, in a form that encourages multiparty competition to emerge.
• There should be improved fiscal federalism that seeks to equalize the per-person tax base across states without rewarding states with low taxes or punishing states with high taxes.

These will smooth the path to delivering fully funded public services, break up our uncompetitive and gridlocked party system, and reduce the points of conflict between federal and state governments. They will bring us closer to a more desirable, Madisonian balance in which federal and state governments serve as bulwarks against each other’s worst tendencies without spiraling apart entirely, and both can work together against narrower, more local interests.

Muscular state governments

It may seem peculiar to assert this in 2026 of all years, but the persistent problems in American politics are not truly at the national level or even at the state level. They are at the very local level. Land-use law at this level retains characteristics set in the era of open racial discrimination.² More generally, the high degree of discretionary power available at the most local level in a country as vast as the United States has left incredibly wide variance in areas such as education and policing. Among the more than 13,000 school districts³ are some of the best K-12 schools in the Organisation for Economic Co-operation and Development, as well as many of the worst. Among the more than 12,000 police departments in the country⁴ are highly professionalized ones and highly patrimonial ones.

States are closer to these matters than the federal government but in many ways, from a practical point of view, face very similar challenges in obtaining compliance. The administrative capacity of localities in aggregate, relative to states, should not be underestimated.

Personnel is a decent proxy for this capacity. As of March 13, 2026, the Office of Personnel Management put the federal civilian workforce at just over two million people.⁵ In 2024, the total number of employees across all state governments was almost three times the federal number, at 5.5 million. And the total number of employees across all local governments was almost three times that number, at 14.4 million.⁶

If Congress passes a law mandating some particular behavior from state governments, there is a great deal of room for state officials to pay lip service to the policy and then simply behave as they please. The DOJ might notice and take them to court, and the court might force some change in behavior. But if thirty states are in a similar degree of noncompliance, there is no conceivable way the federal government can pursue enforcement against all of them. This is more or less what the federal government faced in the effort to defeat the Jim Crow regimes in the second half of the 20th century. Success required a decades-long effort, the full support of Congress, a specialized division of the DOJ, and an effective campaign of mass persuasion.⁷

Similarly, when a state passes a law in an attempt to end the practice of raising revenue through excessive fining, it encounters great difficulty making the law stick. Localities are very motivated to find ways to keep that revenue and have the people on the ground, while state governments only have so many people to monitor them and enforce the law.⁸

At the local level, governments are also far more vulnerable to narrow interests. A single police union can dominate a town to a greater degree than it could hope to dominate a state. Local economic interests have also formed a critical backbone of Trump’s support and have acted as a reactionary influence on our politics for far longer.⁹

Many of these problems become more palatable if we integrate all of local administration into statewide administrative bodies.

America is a federalist system, the first of its kind. Far more common are unitary states, where the central government has unquestioned authority over all regional or local bodies. France, the United Kingdom, Japan, and South Korea are examples of unitary states. When these states engage in good policy, it becomes the law of the land, the whole land. Japan’s land-use laws are the envy of many American urbanists for balancing the need to build with regulations that make cities quiet, pleasant places to visit and live.¹⁰ But the difference between Japan and the U.S. is not just that Tokyo has better zoning than New York City or San Francisco. Japan’s zoning system was created by the national government, and applies across the country.

Of course, when the national government pursues bad policy, that, too, is uniformly imposed. And in 2026, with the federal government deploying masked agents of the states to terrorize our own communities, one can certainly appreciate the merit of a system that creates multiple centers of power and authority.

The problem with America’s system is not its federalism. The problem is that it takes the principle too far, turning each of our states into little federal states themselves when instead we ought to strive for a federation of unitary states. To return to the example of predatory fining, it is one thing when the state attorney general threatens to sue noncompliant police departments. It is something else entirely when they, or the governor, or someone else in the state executive branch can simply fire the police chief they suspect of covering for the noncompliance. They can then replace them with someone of their choosing, who is more likely to be responsive to their demands.

Similarly, implementing regular rotations in the localities officials serve in is one way of undercutting interest-group capture of local arms of government. Individuals fired for being bad actors will also find it much harder to just move over a county or two when all nearby counties are run by the same umbrella employer. Setting a higher floor of work quality and professionalism is far more achievable under these conditions.

Statewide integration should not be sold as a matter of bringing localities to heel. Teachers, police officers, and administrators of all kinds stand to benefit from the taller career ladder afforded by employment in a much larger organization. There is also the flexibility of allowing internal transfers within the same organization if, say, a government worker’s spouse gets a job opportunity in another part of the state.

Without understating the political and logistical difficulties of pursuing such an ambitious integration, this is an effort that can begin today. The California attorney general has had to sue the state’s own localities in an effort to get them to comply with state housing law.¹¹ Rather than passing laws they’ll attempt to enforce on local governments, California legislators ought instead to pass a bill putting all local authorities tasked with approving or blocking new construction under the direct authority of the state executive branch. And then they should fire a long list of the worst offenders to send a clear message, replacing them with officials committed to actually solving California’s housing problems.

Statewide consolidation does not necessarily mean the end of any kind of local governance, even local elected governance. France has locally elected councils and the U.K. has elected mayors, for example, so the existence of authorities at this level is not at odds with functioning as a unitary state. But it is the central government (or in our case, state governments) that sets that authority, and has the power to end that authority at any time. Devolution is very important from a practical administrative point of view; it is clear that Albany cannot manage the daily governance of New York City, any more than Springfield can do so for Chicago. The main thing that is consolidated is administration, including the ability of the state government to directly remove officials who do not comply with state law.

Localities could still pass ordinances and local rules of various kinds, of course—as they do in unitary states today. But their authority to do so would be formally recognized as having been delegated from the state government, to cover areas that the latter either has not covered or has not covered in sufficient detail for local needs. And all such rules could be reversed by the state legislature, or by a state executive body given the statutory authority to do so.

The federal government should facilitate this consolidation. But first and foremost, it should ensure that the governments these consolidations occur under are in fact accountable to a majority of their voters.

Guaranteeing a republican form of government

With the slow death of the Voting Rights Act equilibrium, an increasing number of state governments in the 21st century came to resemble contemporary Hungary. In 2023, the Brennan Center for Justice calculated that North Carolina’s gerrymandering was so egregious that “Democrats could win a solid majority of the ballots cast for Congress, but their candidates would win less than 30 percent of seats thanks to Republicans’ carefully engineered gerrymander.”¹² Wisconsin was an earlier pioneer of the model, but it has thankfully been defeated there.¹³

Article IV, Section 4 states that “the United States shall guarantee to every State in this Union a Republican Form of Government.” It’s time for the government of the United States to act like an actual guarantor on this score.

We should not be content with a new Voting Rights Act to replace the old one. We need federal legislation that is prescriptive and specific about what is and is not the republican form of government, and a federal government that enforces this energetically.

America is not the only country in the world. History did not end in 1788. Proportional allocation methods did not yet exist then; they do now and they are used the world over. It’s time to actually use them here, and use them extensively. This will not only avoid extreme outcomes such as Wisconsin’s and North Carolina’s competitive authoritarianism; it will spell the end of our two-at-most party system, in which huge swaths of the country experience no party competition whatsoever. This, more than administratively burdensome and slow methods of enforcing relatively fair single-seat districts, will guarantee that states remain a republican form of government.

This is an important goal no matter what, but it’s especially important if we seek to consolidate local administrative power within state governments. The two reforms must go hand in hand if we want consolidation to be a real improvement for the American people. But consolidation also lessens the burden on the federal government as guarantor. If the state government is the one that truly matters for outcomes in that state, then the federal government needs to monitor only fifty governments, not 90,000.¹⁴

A bill guaranteeing the republican form of government could allow some flexibility for the “laboratories of democracy” (as Supreme Court Justice Louis Brandeis famously called them). But a limited range of options should be outlined, in detail. Broadly, this should focus on:¹⁵

• Mandating measures to lower the barrier to voting: automatic and election-day voter registration, early and mail-in voting, a fixed minimum number of polling places and workers per voting-age resident, and state holidays for every election day.
• Decreeing that at least one state legislative chamber be proportionally allocated. This will ensure it is impossible for a party to seize the entire legislature with less than a majority, and encourage multiparty competition.
• Requiring any single-seat office to have a runoff, either through two rounds of voting or through what is colloquially called “ranked choice voting” (which involves “rounds” of calculation rather than voting).
• Inviting states to consider eliminating separately elected executives in favor of parliamentary modes of selecting them.

It bears emphasizing that any serious reconstruction effort will need a cooperative Supreme Court. Whether through a one-time expansion or Senator Ron Wyden’s more politically palatable approach,¹⁶ a committed post-Trump trifecta will have to create a cooperative Court if we are to have any hope at moving forward on any important front.

There is no reason a Court should consider the above prescriptions unconstitutional. Guaranteeing the republican form of government is in the original constitutional text, and the Court itself has intervened in brazenly undemocratic districting arrangements for decades. Indeed, reading the case law that accrued over the course of the VRA’s implementation, one cannot help but feel the Court was repeatedly trying to force single-seat districting to behave as if it were proportionally allocated.

Once the republican form of government has been guaranteed at the state level, it can be trusted with more fiscal support with fewer strings attached.

A new fiscal federalism

American liberals often point out that the burdens created by means testing for SNAP or unemployment end up excluding people who should be eligible.¹⁷ What few realize is that the compliance requirements placed on federal funds sent to state governments are far, far more onerous. In education, a significant portion of federal funds is ultimately spent on salaries for individuals whose job is little more than ensuring compliance.¹⁸ Moreover, the way that funds are obtained, and spent, in practice reduces accountability and agency. When no one person in the chain of decision-makers, in political entities from the state government to the school board, has discretion over the use of funds, it is very easy to deny individual responsibility. In many ways, it is simply accurate to do so.

Niskanen Center Director of Social Policy Joshua McCabe points out that other federal systems like Canada’s grant their subnational units “equalization block grants” that are “totally unconditional,” along with other policy-area-specific grants that have conditions, but far more minimal ones than we are used to in the United States. Their outcomes are not worse than ours.¹⁹

Like our peer nations, we ought to use the federal government to set an effective floor on state budgets per resident. What that floor should be is a matter of politics, social science to some degree, and judgment. But a floor should exist. States can then set higher or lower taxes to raise revenue beyond that floor, but there should be a floor that every American is guaranteed to stay above, no matter where in the country they live.

It is important to emphasize that, to a large extent, this is what federal policy already does and what state policy does to an even greater degree with respect to localities.²⁰ It just does it extremely poorly compared with peer federalist systems. We can, and should, do better.

The tax revenues of the federal government, state governments overall, and local governments overall, respectively, run inverse to the size of their workforces. In 2021, the federal government raised $4.3 trillion in tax revenue, state governments overall raised $2.5 trillion, and local governments overall raised $1.6 trillion.²¹ 

At $4.1 trillion combined state and local, it would seem that subnational units overall have a fiscal capacity roughly on par with the federal government’s. The problem is, that revenue is wildly unequal in its distribution. If the wealthy choose to cluster in a few key places, there is only so much that can be done to get rid of tax-revenue inequality per se, especially in a geographically large country like America. But this is precisely why the aim should be to set a generous per-resident floor on state budgets.

Statewide consolidation can help with this, by creating a single chain of authority within the state and allowing school-level administrators more discretion, with the understanding that the state government can remove and replace them if their performance is lacking.²² But as things stand, federal funding adds complexity, cost, and confusion. We can do better.

By guaranteeing that the elected branches of state governments are actually responsible to a majority of voters, we ensure a degree of trustworthiness in those governments. This, along with reducing costs for all parties involved, justifies relaxing federal oversight over the use of funds.

Having greater discretion to use federal funds can also aid states in their efforts to consolidate governance. States ought to appropriate a great deal of the fiscal capacity of their localities whose administrative systems they are co-opting. One of the stubborn causes of inequality in this country is the clustering of affluent families not just in particular states but in particular localities, too, which then use their superior tax base to outperform local governments with poorer residents. Just as the federal government should step in to set a floor on state budgets per resident, states should raise that floor higher for all of their residents, not only the ones in rich counties.

An important caveat on this strategy is the different approaches that must be taken to urban, suburban, and rural localities. You are not going to fund a subway in rural counties, and you are unlikely to do so even in suburban ones. On the other hand, given the different population densities involved, spending more on cities in absolute terms does not necessarily equate to spending more per resident.

Why it matters

The same conservatives who prematurely grieved the death of state governments have also long suspected liberals of seeking to impose a unitary national government, or of already having done so. There is indeed much to be impressed with in the successful unitary liberal democracies around the world. But at this very moment, the federal government is far too dangerous to be trusted with power of that kind. Moreover, a successful reconstruction of our federalist system can help anchor the federal government itself in a safer equilibrium.

The proposal of this essay is to get the benefits of unitary governments at the state level while making the federal government a guarantor of state governments’ democratic design and fiscal capacity to discharge their duties. It uses this balance to introduce a more competitive party system and a more majoritarian system overall, as well as an arrangement more likely to take down the temperature of intergovernmental relations, to the benefit of ordinary Americans.

This does not mean that liberals ought to abandon more robust federal policy. But so long as the Senate remains in place and the legislative process calls for the entire trifecta to sign on, the federal process will always be supermajoritarian at best, requiring a far more demanding politics to achieve anything. State governments, meanwhile, are far more malleable, and can be forced into a form that is actually responsive to majorities. Furthermore, the task of consolidating administrative capacity at the state level, while by no means an easy one, is far less ambitious than doing the same at the national level.

In short, this three-part proposal can be put into law entirely by statutes—federal and state ones. No federal amendment is required. The result would be a more unified, and significantly more democratic, system than the one we have now, without putting all of our eggs in the basket of a federal government currently engaged in ethnic cleansing.

No institutional design can replace politics, and politics is what it will take to navigate our way out of our current crisis. But the approach outlined above offers one set of institutional routes to consolidating the gains of that hoped-for political victory.

Appendix: Guaranteeing the Republican Form of Government

Single Member Plurality (SMP): Any elected office where one seat is filled by whoever obtains the most votes, no matter how divided the vote is. This means that a race with a large number of candidates can result in a winner obtaining much less than a majority.

Proportional Representation (PR): A family of methods that allocate multiple seats at a time in a way that ensures the percentage of seats a party controls hews closely to the percentage of votes they receive.

Two-Round System (TRS): A traditional runoff, in which a second election between the top two candidates is held after the first, but only if no candidate receives an outright majority initially. This is to ensure that whoever wins is able to truly obtain the support of a majority of voters.

Instant Runoff Voting (IRV): Often called “ranked choice” by contemporary election reformers and in popular discussions, this attempts to avoid the turnout loss and costs of doing a true second round by calculating majority support based on ranking.

Mixed Member Proportional (MMP): This mixes SMP districts with a single PR district that ensures the overall allocation of the chamber is proportional. Because single-member districts are tied to narrower geographic areas, there is believed to be value in preserving elected officials picked that way so that constituents have an individual representative.

Proposed range of forms state governments may take

The status quo in most states is two legislative chambers and multiple separately elected executives, including the governor. These are elected through SMP or TRS. The legislation guaranteeing the republican form of government should mandate that these be modified into one of the following forms:

• One chamber allocated through a statewide PR district, the other chamber allocated through TRS or IRV districts, all separately elected executives selected through TRS or IRV. Legislative districting in the non-PR chamber must be drawn by an independent state agency and submitted to the federal DOJ for preclearance.
• A unicameral legislature that is either PR or MMP, along with separately elected executives selected through TRS or IRV. If MMP, the districts must be drawn by an independent state agency and submitted to the federal DOJ for preclearance.
• Either of the above legislative arrangements, but with executives picked by simple majority vote of the PR or MMP legislative chamber.

1. Martha Derthick, Keeping the Compound Republic: Essays in American Federalism, Washington, DC: Brookings Institute Press, 2001, 63.
2. Richard Rothstein, The Color of Law: A Forgotten History of How Our Government Segregated America (New York: Liveright Publishing, 2017).
3. National Center for Education Statistics, “Table 214.10. Number of Public School Districts and Public and Private Elementary and Secondary Schools: Selected School Years, 1869-70 Through 2022-23,” Digest of Education Statistics, U.S. Department of Education, last modified January 2024.
4. Mark Berman, “Most Police Departments in America Are Small. That’s Partly Why Changing Policing Is Difficult, Experts Say,” The Washington Post, May 8, 2021.
5. U.S. Office of Personnel Management, “Federal Workforce Data,” accessed April 14, 2026.
6. Kimberly Ennis et al., “Annual Survey of Public Employment & Payroll Summary Report: 2024,” G24-ASPEP (Washington, DC: U.S. Census Bureau, March 27, 2025).
7. Adam Gurri, “Law and Social Action,” SSRN, March 20, 2021.
8. Mike Maciag, “Addicted to Fines,” Governing, August 19, 2019.
9. Patrick Wyman, “American Gentry,” The Atlantic, September 23, 2021.
10. James Hansen, “Japan’s Zoning: Creating Mixed-Use, Walkable Cities,” Strong Towns Langley Library, accessed April 14, 2026.
11. California Department of Housing and Community Development, “Legal Action,” accessed April 14, 2026.
12. Michael Li, Peter Miller, and Gina Feliz, “Anatomy of a North Carolina Gerrymander,” Brennan Center for Justice, October 27, 2023.
13. Michael Li and Thomas Wolf, “5 Things to Know About the Wisconsin Partisan Gerrymandering Case,” Brennan Center for Justice, June 19, 2017.
14. Amy Smaldone and Mark L. J. Wright, “Local Governments in the U.S.: A Breakdown by Number and Type,” Regional Economist, March 14, 2024.
15. See a more detailed technical discussion in the Appendix.
16. Office of Sen. Ron Wyden, “Judicial Modernization and Transparency Act,” bill summary, accessed April 14, 2026.
17. Ashley Fox, Wenhui Feng, and Megan Reynolds, “How Increasing Administrative Burdens and Means Testing in the U.S. Safety-Net Punishes the Poor,” LSE United States Politics and Policy (blog), August 22, 2022.
18. Paul Manna, School’s In: Federalism and the National Education Agenda, American Governance and Public Policy (Washington, DC: Georgetown University Press, 2006), Kindle locations 1573–1576.
19. Joshua McCabe, “Rich State, Poor State: The Case for Reforming Federal Grants,” Niskanen Center, December 4, 2019.
20. Ibid.
21. Tax Policy Center, “What is the Breakdown of Revenues Among Federal, State, and Local Governments?” The Briefing Book, accessed April 14, 2026.
22. See the example of how the English reform was designed in Michael Barber, “English Perspectives on Education Governance and Delivery,” in Education Governance for the Twenty-First Century: Overcoming the Structural Barriers to School Reform, ed. Patrick J. McGuinn and Paul Manna (Washington, DC: Brookings Institution Press, 2013), 211–12.