Your current web browser is outdated. For best viewing experience, please consider upgrading to the latest version.


Send a question or comment using the form below. This message may be routed through support staff.

Email Article

Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
Main Error Mesage Here
More detailed message would go here to provide context for the user and how to proceed
search DONATE
Close Nav

Devolution: Four Proposals to Empower States and Reduce Washington’s Political Strife

back to top

Devolution: Four Proposals to Empower States and Reduce Washington’s Political Strife

April 15, 2021
Public SectorReinventing Government
OtherCulture & Society

Executive Summary

America’s population has grown to 330 million—the third-largest in the world—and the nation spans an entire continent, 3,000 miles from coast to coast. Yet the federal government in Washington, D.C., decides which sidewalks to pave, which community centers require renovation, how teachers in local schools everywhere interact with children, and a fair price for kidney dialysis in Wichita.

Meanwhile, the U.S. Constitution was designed to require a large degree of cooperation and consensus in enacting large national reforms—yet partisan politics and a growing ideological chasm between Republicans and Democrats have made cooperation increasingly difficult.

For example, populist rage over Washington’s health-care debates caused control of Congress to flip in 1994 and again between 2010 and 2014. Yet this 30-year war resulted in just two major, transformational health reforms: the 2003 creation of the Medicare prescription drug benefit; and the 2010 Affordable Care Act (ACA). ACA saw a disastrous rollout, and a push to repeal the law dominated domestic politics for the next decade. Much of the 2020 Democratic presidential primary was riven by an intraparty war over whether to support Medicare for All, which has no chance whatsoever of congressional passage anytime soon.

And divisive battles are not just about health care; similar partisan wars are regularly fought on issues such as welfare, taxes, the minimum wage, transportation, and education.

This paper asks: Does Washington have to decide all these issues? Did Vermont voters elect Bernie Sanders to the Senate so that he could impose his socialist vision on Texas? Did Texas voters send Ted Cruz to the Senate so that he could bring conservative policies to Vermont? Some subjects and areas are under Washington’s sole control, such as national defense, international relations, regulating commerce among the states, coining money, and immigration and naturalization. In others, it has a leading, though not exclusive, role, such as pollution and the protection of individual rights and liberties. But the U.S. Constitution establishes a federalist structure; it assumes that state and local governments will retain significant power.

Nevertheless, Washington has steadily assumed greater and greater control over many areas, though its performance leaves much to be desired. The No Child Left Behind Act (2001) dramatically increased funding and regulations for public education; yet student performance continued to stagnate, and a bipartisan majority reversed its key provisions in 2015. Infrastructure improvement has remained largely underfunded, stagnant, and subject to expensive delays driven by Washington micromanagement. ACA had the aforementioned failed rollout, exchange enrollment has remained far below projections, the law has required emergency (and possibly illegal) repairs, and key parts of the law were declared unconstitutional. The welfare system remains expensive, fragmented, and lacking any comprehensive vision or goals. Washington’s college-student financial-aid programs lead to higher tuition increases rather than student affordability.

In short, the federal government inspires little confidence in its ability to solve problems. Congress remains largely paralyzed and unfocused, and the federal bureaucracy remains largely distant and unaccountable to the local communities that it is charged with managing. Perhaps it is time to seriously consider moving decision-making in some areas from Washington to the states. This paper does not claim that decentralization—more commonly referred to as “devolution”—is a panacea for every social, economic, and political ill afflicting the country; nor does it ignore the reality that the national government has a necessary, active role to play in the country’s affairs. Nevertheless, it offers four areas where state and local governments can and should play a significantly stronger role than they now have, and where devolution can lead to better outcomes and lower levels of political strife. These areas are transportation, K–12 education, welfare, and health care.


The Disunited States of America

In the summer of 2019, state governors had a collective approval/disapproval rating of 51/32. By contrast, President Donald Trump’s rating was 43/54, and Congress held a 13/84 approval/disapproval rating.[1] More broadly, a 2016 Gallup poll revealed that 55% of respondents preferred that power be concentrated in state governments, versus 37% preferring the federal government.[2] This preference is long-standing. A series of surveys compiled by the Cato Institute found that a majority of Americans prefer that the federal government lead in areas such as defense, immigration, Social Security, protection of civil liberties and civil rights, food safety, and cancer research. However, Americans strongly preferred state and local governments to lead on health insurance (62% versus 38% preferring federal leadership), welfare (68/31), unemployment (55/26), education (75/25), prekindergarten education (71/25), transportation (78/22), law enforcement (73/20), job training (75/20), housing (83/18), and paving roads (77/9). Surveys show that these preferences are driven by the widely shared beliefs that state and local governments are much more competent, fairer, and less wasteful than the federal government.[3] Simply put, voters want more power to devolve to state and local governments.

Voters’ preference for state and local governments has likely been influenced by decades of geographic partisan sorting. Over the past few decades, America has politically, culturally, and religiously sorted itself into different communities. Political commentator David French has noted that the number (and geographic clustering) of evangelicals is growing—yet so is the number and clustering of nonbelievers. Many workplaces and industries—from the technology sector to colleges and universities—are increasingly politically homogenous. Social media have increasingly walled Americans into like-minded communities that seemingly inhabit completely different worlds.[4] Republicans and Democrats get their news from different sources, read different books, and watch different television shows.

Geographic sorting has accelerated as people have chosen to move to like-minded communities, and even longtime residents have increasingly converged politically. While 27 states split their U.S. senators between different parties in 1979, only six did so in 2021.[5] Despite being just eight years apart, 43 states voted for a different presidential party in 1972 than in 1964—with these states swinging from an average of supporting the Democratic nominee by a 23-point margin to supporting the Republican by 26 points. By contrast, in the six presidential elections since 2000, 35 states (and the District Columbia) have each voted for the same party—and nine more states flipped parties just once. At the congressional district level, according to the Cook Political Report, the number of House “swing districts” (defined as between R+5 and D+5, compared with the national average) declined from 164 to 72 over the past 20 years. That means that more than 80% of districts are firmly Democrat or Republican. From 1997 to 2017, the median Democratic House seat went from D+7 to D+14, while the median Republican seat went from R+7 to R+11.[6]

At the county level, 2,474 of America’s 3,113 counties in 2016 showed a presidential candidate win by at least 20 percentage points. These “landslide counties” include 193 million people and 60% of all votes.[7] Since 1992, the number of “extreme landslide counties” (with margins exceeding 50 percentage points) has risen from 93 to 1,196, more than a third of America’s 3,113 counties. Meanwhile, just 303 counties had a 2016 presidential winning margin in the single digits.[8]

This partisan sorting is also reflected in the electorate. A national survey by the Pew Research Center showed that, from 1994 to 2014, the percentage of Democrats whose political values were to the left of the median Republican increased from 64% to 92%. Similarly, the percentage of Republicans whose political values were to the right of the median Democrat increased from 70% to 94%. In that time, the percentage of Democrats whose views were “consistently liberal” rose from 5% to 23%, while the percentage of Republicans who were “consistently conservative” rose from 13% to 20%.[9]

While America has sorted itself into like-minded communities with shared values and political views, Washington has increasingly imposed one-size-fits-all solutions. This has led to political warfare: lawmakers from California and Alabama battle over which side gets to impose its national solution on the other side. Centralization has also encouraged poor policies: there is little reason that Brooklyn and Alaska require similar approaches on poverty or education policy. Nor is there any rational reason that Hawaii and South Dakota must have similar health-care systems.

Local communities should be allowed to tailor local solutions based on their values and preferences. Instead, we have a rhetorical and partisan civil war. Southern Republicans aim their vitriol at House Speaker Nancy Pelosi and her “San Francisco values,” while California Democrats see their policy preferences thwarted by Senate Republican Leader Mitch McConnell, based on Kentucky values.

In a 2017 reaction to Trump’s election, a short-lived “CalExit” movement saw 44% of California Democrats express support for their state seceding from the United States and forming a separate country.[10] While secession is obviously unrealistic, it was legitimate for left-wing California to question its micromanagement by a federal government run by a president antithetical to the state’s politics.

State and Local Advantages

The long-term trend toward increasing Washington centralization raises the question: Are states merely the administrative divisions of the federal government? In reality, state and local government flexibility and experimentation serve three purposes.

First, state and local governments better understand local needs and can tailor local solutions to local problems. What works in Miami, Florida, may not be the best policy for Grand Forks, North Dakota. Relatedly, governments closer to the people provide more accountability and information. Local leaders work with local stakeholders, and if they impose the wrong solutions for the community, they can be more easily replaced by the affected constituents.

Second, state and local governments are generally more bipartisan and compromising than Washington. Local lawmakers often have smaller ideological differences, as the gap between relatively conservative and liberal politicians is much smaller in state and local governments than in Washington. Liberal states like Massachusetts and Maryland nevertheless often elect Republican governors because those states’ political center of gravity is already to the left, and successful Republicans are typically only slightly to the right of that median level. Similarly, conservative states like Louisiana have recently elected Democratic governors who are more conservative than national Democrats. These smaller partisan gaps make bipartisanship more likely.

Third, state and local control emphasizes the importance of experimentation. States have regularly been called “laboratories of democracy,” where new policy approaches can be test-driven in individual states. Over time, successful policies can be replicated in other states, while failed approaches can be abandoned after harming only one state instead of 50.

Why Not Federalism?

Despite unhappiness with the central government, calls to transfer more power to state and local governments often meet resistance, particularly among the most politically engaged citizens. Devolution is, first of all, unsatisfying to ideologues so convinced of the superiority of their policies that nothing less than their implementation from sea to shining sea is acceptable. Partisans of both parties believe that a wave election will sweep the other party aside and provide them with full control of Washington’s levers of power, from which they can impose their tax, health-care, and safety-net agenda on all 50 states. This approach is usually accompanied by a belief that the other party’s policy approach is so destructive that no American should live with its consequences, even if local voters unwisely choose to vote for it.

On one level, advocating devolution requires humility, modesty, gradualism, and acceptance of other parts of the country choosing a policy direction that one may find harmful. However, on another level, discussed later, devolution is dynamic and confident. Rather than continue with fossilized debates in the nation’s capital, devolution empowers state experimentation with new approaches—and the confidence that one’s own policy approach will work successfully enough to be replicated in others.

Another criticism raises the question of competence. In this view, state governments are the minor leagues: amateurish, lacking expertise, and unwilling to provide minimum safeguards to protect families from extreme policies. Of course, the first question one must ask when assessing state government competence is: compared with what?

The federal government failed to anticipate the 9/11 attacks and the 2008 housing crash, failed to run a competent health system for returning veterans, and failed to build a successful health-care enrollment website for the ACA in 2014. Federal programs waste billions of dollars annually, cannot pass an audit, and are filled with hundreds of overlapping, duplicative, and contradictory programs. The federal tax code is more complicated than ever, incentivizes tax-gaming rather than efficiency, and fails to collect $500 billion in owed taxes each year.[11] Despite long-term warnings that it was only a matter of time before a serious pandemic might occur, Washington was caught unprepared for the coronavirus, and the early response was marred by failures and incompetence, such as the CDC’s erroneous diagnostic tests.[12]

While competence levels do vary, the large majority of state governments are no less competent than the federal government. Many U.S. states have economies, populations, and geographic sizes that are comparable with those of European countries. If all states were divided into separate countries, California would be the world’s fourth-largest economy, while Texas would be ninth and New York eleventh. Moving down the list, the size of Virginia’s economy roughly matches that of Poland, Louisiana’s roughly matches Finland’s, Montana’s matches Serbia’s, and the state with the smallest economy—Vermont—approximates the Baltic states of Latvia and Estonia in their economies. If those countries are large enough to determine their own policies, so are American states.

European comparisons are especially relevant to American liberals. Political scientists and economists have pointed out that larger welfare states and social democracies are more likely to flourish in smaller, more homogenous countries with higher degrees of social trust.[13] Sweden’s policies were enacted in a country with a population similar in size to Georgia’s, and many of its policies have proved more popular and enduring when imposed on this scale. Put differently, national health care is highly unlikely to be imposed on a diverse nation of 330 million people but could be much more plausibly enacted on a state-by-state basis.

States such as Illinois, Mississippi, and Louisiana have traditionally been exposed for wasteful, illegal, or irresponsible policies and politicians. But that is no reason to give Washington suffocating power over all 50 states. Instead, Washington’s primary role for many policy functions that it provides funding for should be to guarantee basic safeguards and accountability standards—as well as protect individual civil liberties and civil rights—across all 50 states. Beyond that, some states will make mistakes or have different policy approaches from other states. Nevertheless, in a democracy, states should be accountable to their voters and their policy preferences. Those who desire their state to reflect local consensus views must grant other states that same opportunity.

Surely not all federal policies should be devolved to states, and those with significant spillover effects to other states—such as pollution, interstate communications, and interstate crime—should remain federal. Washington should also maintain control of obvious national policies such as defense, international relations, homeland security, space exploration, macroeconomic stabilization, the Federal Reserve, broad financial regulation, health and safety regulation, federal research, and the federal tax code. Federal minimum standards can avoid “races to the bottom,” whereby states drastically cut their welfare systems to motivate their caseloads to move to neighboring states. Many other policies can be devolved with the help of federal offices of interstate coordination across various policy areas. For example, with the federal highway system having been completed in the 1980s, states can maintain their portion of the interstates (as required by Washington) and use a federal office of interstate transportation coordination to collaborate on cross-state projects. Health-insurance markets already work well within states.

Critics can come up with countless reasons that certain state and local governments might make mistakes on a given policy. The more important question is whether the 50 states would perform worse than the federal government on a given policy, or with less accountability to their constituents. If each European nation is capable of running its education, health, transportation, and safety-net programs, shouldn’t we be able to trust California, Wisconsin, Florida, and Massachusetts, as well?

Devolution in Transportation

The federal highway system is the easiest and most obvious candidate for devolution. States today control two-thirds of all gasoline tax revenue and surface transportation spending without any federal involvement.

Nonetheless, states also collect and send to Washington the 18.4-cent-per-gallon federal gas tax—which, in turn, sends most of that money back to the states, with numerous strings attached. A commonsense solution would eliminate the federal middleman and allow states to retain and spend those gas tax revenues on transportation projects of their own choosing.

In 1956, Washington passed the Federal Aid Highway Act—aka the National Interstate and Defense Highways Act—which authorized the creation of a 41,000-mile network of highways that would stretch from coast to coast and be funded by a 3-cent-per-gallon federal gas tax (the network was later expanded to nearly 47,000 miles). Originally, the system was to be completed by the early 1970s, at which point, the federal program and its taxes would end, and states would take over.[14]

Instead, Congress expanded the program to include state highways and roads, maintenance of all types of roads, mass transit, nature trails, bike lanes, sidewalks, parking lots, museums, landscaping, flower plantings, safety grants, metropolitan planning, university research, historic preservation, and environmental initiatives.[15] Today, more than a third of federal taxes paid by motorists are diverted from general-purpose roads to other uses.[16] This has allowed Washington politicians to take credit for a wide array of projects well outside the creation of the interstate system—and to abandon the original plans to dramatically scale back the federal role after the (delayed) 1980s completion of the federal highway system.

Washington’s micromanagement of local transportation decisions peaked in the 1990s and 2000s. Congress began overruling U.S. Department of Transportation (DOT) decisions, as well as those of state and local governments, using legislative earmarks to direct funds for specific projects. By 2006, Congress was earmarking 8,000 projects, thus dictating which intersections would have a streetlight and which town sidewalks to pave—leaving many to ponder the remaining purpose of mayors and city councils. Congress banned earmarks in 2011, although they are staging a comeback.[17]

Washington’s transportation meddling surely contributed to a survey showing that two-thirds of Americans believe that the government does not spend transportation dollars efficiently.[18] This lack of faith in Washington has driven the opposition to raise the federal gas tax over the past 27 years—though voters during those years nonetheless frequently raised their state gas taxes.[19]

The federal highway and transit system suffers from many shortcomings. Among them are:

Large Deficits. Today, Washington imposes an 18.4-cent-per-gallon gas tax (plus a 6-cent surtax for diesel)—of which 2.86 cents is earmarked for transit (such as buses, subways, and commuter rail). Altogether, the federal highway and transit trust funds collect approximately $43 billion annually and spend $57 billion. The gap is made up by general fund bailouts of the highway program that have topped $140 billion since 2008. Over the next decade, the Congressional Budget Office (CBO) projects that the highway and transit trust funds face a $190 billion shortfall—and even that spending level would be insufficient to maintain highway conditions and performance.[20]

Mission Creep. Although the interstate system was completed in the 1980s, the federal highway program has expanded its jurisdiction to cover 1,077,777 route-miles—meaning that 95% of the federal highway program’s jurisdiction is over state highways and roads that are outside the interstate highway system.[21] Washington owns less than 1% of all public roads—yet it enjoys spending jurisdiction over 85% of all vehicle miles of travel (VMT).[22] There is no proper federal role to micromanage in-state roads and highways, and even improvements on existing interstate highways can be financed by the states themselves.

Political Meddling. After states collect and send the federal gas tax revenues to Washington, getting these tax dollars back requires crafting state transportation plans that must be approved by federal bureaucrats in Washington—at which point the typical project will receive an 80% reimbursement. It is unclear why state departments of transportation—which also manage their own large state highway and road programs with state gas tax revenues—should need the permission of a federal agency to get back their state’s federal gas tax revenues to spend on projects of their choosing.

State Inequities. After Washington collects the federal gas tax revenue from each state, it redistributes those funds based on a complicated formula that does not adequately reflect state gas tax contributions.[23] For example, over the first 54 years of the federal highway program, Texas received 80 cents back for each dollar of federal gas taxes sent to Washington, and Florida received 87 cents—despite each state’s growing populations and rising transportation needs.[24] More broadly, the Highway Trust Fund has historically redistributed money from low-income to high-income states.[25] Even the current requirement that states receive at least 95 cents on the dollar nevertheless deprives several states of hundreds of millions of dollars. Transit dollars are distributed even more inequitably, as the 2.86-cent-per-gallon transit gas tax largely redistributes money from middle America to the coasts, especially to the Northeast.[26]

Red Tape. One reason that American infrastructure is much more expensive and more slowly built than the rest of the world is because of federal regulations. The Davis-Bacon Act, which requires contractors to pay the local prevailing (i.e., union) wage on public works, raises wage costs by as much as 22%.[27] Buy America regulations raise costs, too. Boston’s infamous “Big Dig”—whose cost exceeded the initial estimate by 500%—was mostly federally funded.[28] Federal Environmental Impact Statements—which were typically 22 pages long in the 1970s—now commonly exceed 1,000 pages and require seven years to complete (compared with one to two years in Canada and 3.5 years in the European Union).[29] Congress has also threatened to block states from receiving back their own federal gas tax contributions unless they obey Washington diktats on drinking ages, drunk driving laws, and other matters. If states want to join the rest of the developed world by raising transportation funds with creative approaches such as tolls or public-private partnerships, they run into federal restrictions.

Misaligned Incentives. The formula that Washington will fund 80% and states will fund 20% of the typical federal highway project produces poor incentives. Because the federal reimbursement feels like “free money” (even as it is funded by state motorists), states may determine expensive transportation to be worth its cost as long as their citizens value the project above 20% of its national cost. This encourages states to be more cavalier in spending federal transportation dollars than they are with their own funds. Each state feels as though it is free-riding off its neighbors. The result is projects like Alaska’s notorious “Bridge to Nowhere.”

Another example of misaligned incentives is that, according to Chris Edwards, “Federal aid for urban transit covers about 40% of capital costs, on average, but just 6% of operating costs. That bias has induced local governments to buy expensive rail systems rather than more-flexible and efficient bus systems.”[30] States are obviously capable of managing transportation. They collect gas taxes that average 36 cents per gallon and spend $131 billion annually on surface transportation.[31] State and local governments currently finance a slight majority of capital spending and nearly all operations and maintenance of national highway spending within their borders.[32] They are well equipped to take over much of the federal share of running the highway program.

Washington can devolve its role in surface transportation over several years by gradually reducing the federal gas tax from 18.4 cents per gallon to approximately 3 cents; states can, of course, raise their gas tax accordingly. The remaining 3-cent federal gas tax would finance federal lands and Indian reservations, federal safety programs, and highway or transit projects that truly require national leadership. The federal DOT could establish an office of interstate coordination to help states plan projects that cross borders.

In return, states would be required to maintain the interstate highways that pass through their jurisdictions. Also, states should receive an $80 billion payment from Washington to cover the existing multiyear transportation commitments that the federal DOT has already made.[33]

States would also benefit financially from a reduction in federal red tape and expensive mandates (although the Clean Air Act, Clean Water Act, and Endangered Species Act mandates would still apply). Washington should remove restrictions and encourage states to implement creative financing mechanisms for building and maintaining roads—such as the public-private partnerships (P3), tax credit bonds, expanded tolls, and infrastructure banks that have become widespread throughout Canada, Europe, and Australia.[34]

Some proposals would automatically devolve the highway program, while others would provide states the choice to opt out of the system and retain the gas tax.

Modest challenges remain. States like Alaska that receive very large transportation funding would require a much higher state gas tax to replace this funding. States would need to provide some level of assurance that their interstate highways would be maintained.

The proposal to “turn back” federal gas taxes is not far-fetched. The original interstate highway system enacted in 1956 was intended to sunset by the early 1970s. When the interstate highway system was finally completed in the 1980s, President Reagan’s Advisory Commission on Intergovernmental Relations endorsed “turn back.” Since 1996, Congress has seen regular proposals to devolve the highway program, including a 2002 proposal by Senator James Inhofe (R-OK), who later became chairman of the Senate Environment and Public Works Committee, which oversees the federal highway program (he later abandoned the proposal).[35]

Transportation experts agree. According to the Reason Foundation’s Robert Poole, “a key rationale for devolution is that the funding approach developed to build the Interstate system is now obsolete. That approach transfers large sums from larger and faster-growing states to smaller and slower-growing states…. That is exactly backward of what a real user- fee system would do—which is to generate and spend large sums in the places with huge problems of congestion and insufficient highway capacity.”[36] Even former transportation secretary Mary Peters has argued forcefully for drastically reducing Washington’s role in surface transportation.[37]

Washington is paralyzed regarding highways and transit: the gas tax has been frozen since 1993 (and lost 44% of its purchasing power), highway reauthorization bills are regularly delayed, and federal micromanagement and red tape have stalled America’s infrastructure growth and created highway trust fund shortfalls. While the rest of the world innovates, Washington blocks new approaches.

Devolution in K–12 Education

K–12 education is a rare policy area where many Republicans and Democrats have begun acknowledging the failures of Washington’s centralization and returning some power to state and local governments. The bipartisan backlash against the No Child Left Behind law and other federal interventions of the past few decades have exposed the ineptitude of Congress’s attempts to serve as the school board for 100,000 public schools.

Empowering state and local governments to run K–12 education is vital because the educational needs of America’s schools and children are far too diverse for one-size-fits-all mandates. Governors, school boards, and principals better understand local needs and, crucially, can be held more accountable to parents and the local community. Innovation, compromise, and more realistic goals tend to move more easily from the ground up than from the top down. While local control is far from perfect—teachers’ unions often have far too much power, school boards can be parochial, and local approaches can become rigid and stale—the promise of a federal cure-all is fool’s gold.

America’s schools have faced long-term mediocrity. International testing shows that the U.S. ranks 23rd (tied) in reading and 24th in science and is tied for an abysmal 39th in math. These scores compare particularly poorly against the other 36 developed countries of the Organisation for Economic Co-operation and Development.[38]

The federal National Assessment of Educational Progress (NAEP) test also confirms this mediocrity. Among 12th-graders, reading scores have continued on a slight downward trend since the modern test began in 1992 (with only 37% of students achieving proficiency), while math scores have remained mostly flat since the modern test began in 2005 (with 25% achieving proficiency).[39]

Along the way, fourth- and eighth-graders saw modest mathematical gains in the early 2000s (with most still scoring below proficient), while reading scores in those grades have remained stagnant and only one-third score as proficient. That said, NAEP scores have shown modest gains among racial minorities over this period.[40]

The modern federal role in education widened with the 1954 Brown v. Board of Education Supreme Court ruling that banned racial segregation in schools, and accelerated with the 1958 National Defense Education Act, a Cold War measure to keep up with the Soviet Union, which began pouring hundreds of millions of dollars into public education with an explicit goal to “[e]nsure trained manpower of sufficient quality and quantity to meet the national defense needs of the United States.”[41] Seven years later, the Elementary and Secondary Education Act authorized federal spending to local public schools with a high percentage of students from low-income families. Over the years, Washington expanded mandates and funding that targeted aid to low-income school districts, disabled kids, and recent immigrants—yet was content to let state and local governments drive education policy.[42]

In 1983, the National Commission on Excellence in Education issued “A Nation at Risk,” a report warning that U.S. schools were falling behind and required drastic reform. Governors responded with more standards, accountability, and rigorous course requirements (and later signed legislation authorizing charter schools and private school vouchers). However, most traditional public schools continued to graduate students whose test results were mediocre. As the years passed, Washington took increasing ownership over broad education outcomes, leading to higher funding levels and a new consensus that K–12 education was a key federal responsibility. The result was Republican President George W. Bush teaming up with Democratic Senator Ted Kennedy to enact the No Child Left Behind Act (NCLB) in 2001.

NCLB significantly increased federal funding to public schools; it also required them to annually test all students in grades three through eight (and one time in high school) in reading and math, with test scores broken down by race, ethnicity, income, disability status, and limited English proficiency. The law decreed that schools make “adequate yearly progress” toward the goal of 100% student “proficiency” in English and math by 2014—or face a cascade of sanctions that could eventually lead to state takeovers or shutdowns. Additionally, NCLB added two large new federal mandates: 1) that every teacher be “highly qualified” (which generally meant a college degree and state certification in the subject taught); and 2) that schools employ “scientifically” based teaching standards.[43] As Michael Petrilli of the Hoover Institution and Chester Finn of the Thomas B. Fordham Foundation point out, even these seemingly justifiable mandates saddled states with red tape and also limited innovation, inducing compliance games with federal regulators.[44]

The utopian commandment that schools achieve 100% student proficiency in English and math backfired badly. In the first place, it was designed backward. Instead of setting uniform accountability standards and letting states determine how to meet those standards, NCLB micromanaged the classroom approaches but allowed states to define “proficiency.” Not surprisingly, states simply dumbed down the tests and the meaning of “proficiency,” while also gaming the accountability standards. One notorious episode: after test scores in Atlanta rose sharply, 178 educators were implicated (of whom 35 were indicted) in a massive scandal that involved changing incorrect student test answers into the correct ones.[45] On the national level, NAEP test scores remained relatively stagnant, but many states exploited loopholes to show the required adequate yearly progress toward “proficiency.”

Despite the gamesmanship, more than one-third of public schools missed their annual NCLB testing benchmarks. By 2011, 82% of schools were in danger of coming up short.[46] Faced with the reality that no school could achieve 100% student proficiency by 2014, the Obama administration began granting waivers to states exempting them from NCLB’s punitive enforcement on the condition that governors impose the preferred policies of the Department of Education (ED). As Frederick Hess of the American Enterprise Institute has written, “[ED’s] conditions—including federally preferred teacher-evaluation systems and adoption of the Common Core or a federally approved alternative—had no statutory basis.”[47] The ineffectiveness of Congress in controlling schools had given way to unelected federal bureaucrats imposing their vision.

Washington’s hammer swung even harder in 2013 when the Obama administration sued Louisiana’s school voucher program on the basis that allowing a handful of African-American elementary school students to enroll in local private schools would alter the racial composition of the schools that they left. A federal judge ruled that the program may continue.[48]

By 2015, the NAEP test showed most average scores slightly falling in grades four and eight.[49] Funding has not been the problem. Public schools now spend $15,424 per student annually, which is up 280% since 1960, and up 79% since “A Nation at Risk” was released (all figures are adjusted for inflation).[50] The City of Washington D.C. is estimated to spend as much as $27,000 per student annually, with disappointing results, including one-third of students not completing high school.[51]

NCLB burdened school districts with mandates and a standard of student achievement that led states to focus obsessively on math and reading tests, game the proficiency standards, and then accept even more federal oversight.

Next up: the Great Recession’s 2009 economic “stimulus” legislation created the $4.35 billion “Race to the Top” program. The first $350 million went to fund tests for what later developed into “Common Core” teaching methods; the rest was set aside as a competitive grant process for states that promised to invest in a series of specific educational goals laid out by ED, much of which encouraged adopting Common Core standards. State applications were exceedingly sloppy and error-ridden, filled with empty jargon and vague promises meant to win over Washington’s educrats (the federal application process was so complicated that the Gates Foundation spent millions helping states navigate it). One ED review remarked: “We knew the states were lying. The trick was figuring out who was lying the least.”[52] Ultimately, both the Government Accountability Office and the liberal Center for American Progress reported that not a single state successfully implemented their Race to the Top plan as described in the application, much less achieved the promised results. Washington’s attempt to dictate and control innovation failed.[53]

The $4 billion in “Race to the Top” grants was exceeded by another $7 billion spent between 2010 and 2014 on federal School Improvement Grants (SIGs) that targeted failing schools. President Bush was responsible for this program; President Obama expanded it to offer grants to low-performing schools that promised to adopt specific education reforms laid out by his administration.[54] A few years later, ED’s evaluation admitted that “across all grades, we found that implementing any SIG-funded model had no significant impacts on math or reading test scores, high school graduation, or college enrollment.”[55] In fact, the department could not even show that SIG caused schools to alter their existing educational plans at all.[56]

In 2015, a bipartisan majority in Congress—after hearing years of complaints from parents, teachers, school boards, and governors—replaced NCLB with the Every Student Succeeds Act (ESSA), which eliminated the requirement that states show “adequately yearly progress” toward 100% student proficiency. ESSA also eliminated federal mandates in areas like teacher evaluation and addressing low-performing schools and restricted ED’s authority to use waivers and grants to regulate classrooms. However, in the interest of accountability, the new law retained NCLB’s requirement that students be tested annually in math and reading in grades three through eight and once again in high school (a science test was also added). ESSA was premised on the observation that testing was necessary for accountability, but schools should be freer to decide how to improve their performance.

ESSA was a step in the direction of devolution, but more could be done. Michael Petrilli and Chester Finn have proposed that the federal government be limited to four roles in K–12 education: 1) measuring schools with testing; 2) funding data and research to disseminate test results and best practices; 3) distributing federal funds by formula; and 4) protecting civil rights.[57] Federal student tests should be broadly uniform across the country so that results can be compared across schools, districts, and states, at present and over time. The $190 million budget for ED’s Office of Innovation and Improvement could be significantly increased to ensure that all schools have access to the most successful ideas and blueprints from elsewhere. This includes promising reforms like the “Colorado Growth Model,” which provides superior data on student progress, and the “Florida Formula,” which bans social promotion for children who are significantly deficient in reading and includes school vouchers.[58] Several of the remaining small educational grant programs could be consolidated and distributed to states by formula.[59] Charter schools and private school vouchers should be entirely in the domain of state and local governments.

Devolution in Welfare

For the past 30 years, the most furious devolution debate has involved the federal welfare system. The 1996 reform—which replaced the open-ended Aid to Families with Dependent Children (AFDC) program with the Temporary Assistance for Needy Families (TANF) block grants to states—exposed the radical differences between conservative and liberal approaches to poverty relief that make a case for state customization and experimentation based on local values. AFDC was the first broad national welfare entitlement (it was part of the 1935 Social Security Act); replacing it was the first revocation of a major entitlement for low-income families.

While TANF is only one piece of the federal welfare system, it was a major departure from the past and does inform the federalism debate: first, its existence was made possible by the example of welfare reforms undertaken in Wisconsin, Connecticut, and New York City in the early 1990s, using federal AFDC waivers to impose work requirements and time limits, as well as require child school attendance;[60] and second, TANF displayed many of the promises (and challenges) of welfare devolution.

TANF replaced the open-ended AFDC cash entitlement for families (with minor children) earning below a certain income with a fixed-amount block grant that allowed states to create their replacement system—but one that must require work activities and include a five-year lifetime limit on benefits (both with significant exemptions). Beyond that, states were given broad latitude to design their welfare systems as long as they met four federal goals that generally involved assisting needy families, promoting work and marriage, and reducing out-of-wedlock pregnancies. The law also required states to contribute their own funds to the new system—80% of what they had spent in the old AFDC system.[61]

On the positive side, TANF has shown that block grants can succeed if Washington sets broad goals and provides states with flexibility in program design. Despite hysterical predictions of mass starvation, third-world poverty, and death,[62] TANF has exceeded even its most optimistic predictions. Following the passage of welfare reform, the number of families collecting AFDC/TANF quickly fell from 4.5 million to 2 million—and then to 1.2 million by 2018.[63] Many of these families moved into full-time jobs that paid much more than welfare and began the journey toward self-sufficiency.[64] The poverty rate quickly fell to its lowest level in nearly 30 years; the black poverty (and black child poverty) rates fell so far that even the “Great Recession” of 2007–09 did not raise poverty rates to pre-TANF levels.[65] While critics credit the economic boom of the late 1990s for this progress, previous economic expansions did not produce such large poverty and caseload reductions, and subsequent recessions showed a lower poverty ceiling than before 1996 for most populations.

Still, the TANF devolution was imperfect. States exploited loopholes in the law to exempt most families from work requirements and ignore the five-year time limit.[66] The federal grant has been frozen at $16.5 billion since 1996, despite cumulative inflation of 64%—although sharply reduced caseloads have roughly maintained per-family spending, adjusted for inflation.[67] A few outlier states have offered very low cash benefit levels, rejected as many as 98% of TANF applicants, and diverted funds into questionable or illegal uses.[68] The percentage of families in deep poverty has not fallen.

The strongest criticisms of TANF have come from liberal states that object to the federal work requirements and time limits. These states prefer to put more resources into education and training (or even return to open-ended assistance with no obligations for recipients) and not be constrained by federal goals such as promoting marriage or reducing out-of-wedlock births. The result of these complaints was the Obama administration’s offer to waive (with questionable legality) some of the work requirements for some states.[69]

To be sure, state and local customization is vital for a well-functioning welfare system. Every state and local community has unique poverty drivers and needs.

Some communities may be plagued by high housing costs, extraordinarily expensive child care, or energy costs. Moving welfare recipients into jobs may be more feasible in one local economy than another. Even within a community, each low-income family has its own story and poverty solution. Furthermore, liberal and conservative communities have their preferred poverty approaches based on local values. These considerations do not imply that the country (or low-income Americans) would be better off by returning to AFDC; they argue, instead, for better rationalization of the federal welfare system in the direction of greater devolution.

In any case, the federal welfare system is a bureaucratic mess with seemingly no one in charge. It consists of approximately 80 different federal means-tested programs spread across a half-dozen federal departments and dozens of federal agencies. For example, the Department of Agriculture runs several food programs, the Department of Housing and Urban Development (HUD) operates housing and urban aid, the Department of Health and Human Services (HHS) contains a multitude of cash programs, and the Treasury Department coordinates low-income tax credits. Many federally funded welfare programs are overseen by different congressional committees and reauthorized in separate bills that are enacted years (or decades) apart from each other. Lacking any consistent approach, these programs contain wildly divergent eligibilities, benefit formulas, and requirements from recipients—even the programs that interact with each other. Some programs block grant money to states, others rely on matching grants, and others bypass states entirely. Countless state welfare programs have their own approaches, eligibilities, and benefit levels, as well. In sum, federal and state welfare programs spent $1.1 trillion in FY 2016—enough to grant $100,000 per year to each of the 10 million families with incomes at or below 125% of the federal poverty threshold (or to provide $67,000 per family excluding health-care benefits).[70] In 2016, just the cash, food, and housing assistance came to 140% of the amount needed to bring every American family out of poverty.[71] In fact, much of it does—as the federal poverty rate does not include most federal benefits. Additionally, antipoverty programs assist millions of struggling families whose earners exceed the official poverty line of $21,960 for a family of three.

From the standpoint of devolution, the pertinent issue is what states could do with all federal antipoverty funds consolidated into one block grant that roughly matches how much each state is already receiving. Each state could combine the federal block grant with its own funds to build a coordinated system of assistance to vulnerable families. Rather than navigate a maze of dozens of federal programs, families could visit one local welfare office and enroll in one coordinated system that is tailored to their needs. These offices could also assist with education, job training, and job placement. Policy scholar Oren Cass has called such a system a “Flex Fund,” noting that such a consolidation “increases the likelihood of constructive reform by increasing accountability and eliminating unfunded mandates from above or efforts to game the system from below.”[72]

Not all federal means-tested benefits can be part of this block grant. Earned Income Tax Credit (EITC) would remain part of the federal tax code. However, aggressively expanding EITC can push enough working families out of poverty so that fewer families would need to enroll in the new state fund. Some may suggest a separate fund for health-care benefits (discussed below).

This block grant concept is not new; yet it can address criticisms of past proposals. For instance, while the TANF block grant has been frozen at $16.5 billion for nearly 25 years, this new block grant can be designed to expand automatically annually by a formula that accounts for state poverty populations and the cost of living. It can also include recession triggers for additional increases. As with TANF, states can be required to continue contributing to this welfare system at their previous levels. And if states truly fear being cheated by Washington on the size of the grant, they can be given a frequent choice between the current system or accepting this grant.

Perhaps the most challenging issue is that of federal performance standards and goals. Federal work requirements have been successful in TANF, yet more liberal states have supported more of an education and training approach, with additional work exemptions. Liberal states have also preferred to focus more on reducing poverty (by any means) than encouraging work or reducing welfare caseloads. This can all be negotiated while emphasizing that devolution means trusting states to experiment with new approaches that reflect their values.

Relatedly, states that accept a federal block grant must accept basic accountability rules to ensure that funds are not misspent and that a minimum safety net is maintained. States that set absurdly low benefit levels or reject nearly 100% of applicants must be subject to federal sanctions. The continued failure to produce a welfare system that meets minimum federal standards should result in the revocation of the state’s block grant and its re-enrollment in the current federal system that sets eligibility and benefit levels for most programs.

The vast majority of states are clearly equipped to design and run their welfare system if they choose. Federal accountability standards can address any outlier states that may misspend funds or attempt to eviscerate their safety net. Some states may choose more education, training, and open-ended benefits, while other states more aggressively push work requirements. Each state should reflect its own values and discover which approach works best.

Devolution in Health Care

Health care is probably the most complicated policy to devolve. No other domestic policy has been as controversial and contentious over the past three decades. Partisan wars over health care flipped control of Congress in 1994 and again between 2010 and 2014. Meanwhile, two major health reforms were enacted over this lengthy period: the 2003 Medicare Modernization Act (which added the drug benefit, Part D); and the 2010 Affordable Care Act (ACA). Both laws barely made it through Congress; ACA set off three major Supreme Court decisions and a failed, seven-year battle for repeal and replacement.[73] Even so, much of the Democratic voter base is demanding a spectacularly expensive “Medicare for All,” while Republicans seek to move in the polar opposite direction of consumer choice and deregulation.[74]

Perhaps there is more to be gained by declaring a truce on the federal level. Large single-payer health-care movements have already emerged in Vermont, California, and Colorado—and have been proposed in 20 state legislatures (Vermont enacted a law mandating the creation of this system in 2011, before abandoning it in 2014).[75] Rather than fight it out in Washington, these states should go ahead—on their own. At the same time, states such as Indiana, that are pioneering consumer-based reform should do likewise.[76]

After all, much of the 2010 ACA had its roots in an earlier, creative Massachusetts reform. Yet ACA also shows the perils of federalization, as the crashing enrollment websites, unpopularity in several states, and repeated emergency (and often illegal) bailouts and last-minute changes to avert chaos showed that the federal version was not ready for prime time.[77] And when a federal program is poorly designed, it harms 50 states instead of one.

The main political barrier to health-care devolution may be the refusal of some states to abandon Medicaid’s current nonsensical funding formula. Under current law, states set their own Medicaid funding commitments and then, for each state dollar spent, receive a federal match of between $1 and $9. This creates an enormous incentive for states to raise their Medicaid spending to maximize their federal dollars. States consequently expand Medicaid during periods of budget surplus (which expands their federal match) and hesitate to cut Medicaid under budget deficits (which would also reduce their federal matching dollars). Additionally, nearly all 50 states have resorted to gimmicks that create the illusion of higher Medicaid spending to receive more federal dollars.[78]

This system also disadvantages relatively poor states. While they may receive a higher federal match rate than wealthier states, they also have fewer state dollars to put into their Medicaid system and get matched. Thus, Chris Pope (a Manhattan Institute colleague) notes that, even though Louisiana and Massachusetts have tax systems that collect similar shares of state income, Louisiana’s smaller tax base means less state tax revenue, which translates into fewer dollars to contribute to its Medicaid system. The result: “Massachusetts received $15,482 in federal Medicaid funds per resident under the federal poverty level in 2018, [while] Louisiana received only $9,474—though both are ‘expansion’ states.”[79]

Nevertheless, states have resisted replacing this matching grant system with a block grant that would likely be required for true devolution—preferring to game the federal system rather than innovate on their own. Washington can encourage state innovation by no longer rewarding state matching-fund gimmicks.

The U.S. is unique in its attempt to tightly regulate health programs across a nation of 330 million people. European countries with populations and economies comparable with both large and small U.S. states each run diverse health systems. Nor is Canada’s government-run health system fully centralized in Ottawa. As with Medicaid, Canada once relied on a federal-provincial matching system that brought large cost increases. In 1977, it was converted to a per-capita block grant to provinces, in which funding grows near the rate of inflation. The Canada Health Transfer has reduced cost, gained efficiencies, and gives provinces enormous flexibility to create their unique systems as long as they meet national standards of “comprehensiveness, universality, accessibility, and portability.”[80] Canada’s health system surely has problems, but even its government recognizes the need for local flexibility.

Nor should block grants and devolution seem radical in America. The federal government’s Children’s Health Insurance Program (CHIP), enacted in 1997, incorporates much of this approach. CHIP is intended to ensure health care for children from low-income families that earn too much to qualify for Medicaid. The program is funded by a generous federal matching formula up to a capped federal contribution—which, for states that provide the maximum match, essentially turns the program into a block grant with a state maintenance-of-effort requirement (a requirement for state funds as a condition for federal funding). CHIP provides states with wide flexibility to design their programs as they wish, including capping enrollment, requiring cost-sharing, and determining minimum benefit packages. CHIP has proved to be a popular program, showing that states can creatively design their own programs and Congress can maintain some control over costs.

Devolution in health care could provide states the option to merge their federal Medicaid, CHIP, and ACA exchange premium funding into one package that (when combined with the state’s traditional health funding) can be used to build a new low-income health-care system. States also would have the option of returning to the traditional federal system to protect them from future federal mismanagement or underfunding. This option can be designed as an optional block grant program or incorporated into a new “super-waiver” in which the federal government would approve all waivers that meet basic federal performance standards (with the burden of proof on HHS to justify any disapproval).

A challenge would be designing the federal performance standards and goals. As with welfare devolution, discussed earlier, Washington must ensure that funds are not misspent and that a minimum safety net is maintained. Washington can mandate—just as Medicaid does—that states ensure a minimum level of health benefits for families at or below a certain income. Washington need not mandate the tying of each required benefit to an insurance plan—as long as states can ensure that creative approaches such as subsidized health savings accounts are adequately funding required benefits for families.

Federal funding formulas may ultimately determine states’ receptivity to such a block grant. A commonsense formula would tie each state’s funding to the number of residents enrolled and some measure of medical inflation. Participating states would also be required to supply their own funds at a level comparable with that of the previous systems. The funding formula would automatically enlarge the grant during recessions and can include additional triggers to expand the grant size to cover the entire cost of recessionary caseload expansions.[81]

Many states may not rush in to the new system. Designing new health systems is challenging, and 26 states still have not bothered to set up their own ACA health exchanges, defaulting to the federal exchanges instead.[82] Nevertheless, over time, the policy stagnation of Washington—as well as the budgetary squeeze of surging health-care spending—might motivate more states to pursue new approaches. States that wish to pursue a Medicare-for-All system should also be given some leeway by the federal government to address barriers created by other federal laws and the tax exclusion for employer-provided health care.83 Other state reforms to employer-sponsored health coverage should be permitted when feasible, as should reforms to broadly reduce health-care costs for families and businesses. Demonstration projects should be encouraged, with successful approaches disseminated.


Over the past century, the U.S. has experienced a dramatic centralization of power in Washington. Some of this was brought on by war, depression, and the necessity of protecting civil rights. Nevertheless, bureaucratic failures and incompetence at the federal level, along with deep political fissures that currently exist between the parties and across the country, have made Washington politics truly resemble war by other means. State and local governments have shown themselves capable of undertaking reforms and programs that, if hardly perfect, at least better reflect the values of their citizens. Partisan sorting has created more politically homogenous states and communities that should be given greater freedom to govern themselves, rather than be governed by one-size-fits-all federal programs.


See endnotes in PDF

Photo by davidevison/iStock