Federalism Lets Fifty States Experiment, and the Constitution Planned It

Imagine a drug company that skipped the clinical trials and went straight to putting its new pill in the national water supply. Everyone would call that reckless. Now imagine a government that takes an untested idea about schools, welfare or health insurance and imposes it on three hundred million people at once. We call that leadership.

The Framers had a better idea, and it is hiding in plain sight in the structure of the Constitution. Divide power between a national government with limited, listed powers and states with broad, general ones, and you get fifty places where a policy can be tried, measured and copied or abandoned before anyone forces it on the whole country. That is not a happy accident of 1787. It is the design. Let me lay out the argument and then answer the charge that it only produces a race to the bottom.

Define the terms

Federalism is the division of governing power between a national government and state governments, each with its own sphere, each answerable to the same people, neither created by the other. The states existed before the Constitution and ratified it; the national government exists because they did.

Policy experimentation is the trial of a law or program in one jurisdiction, with results that others can observe and decide whether to copy. It is not a lab with a control group. It is something more democratic: millions of citizens living under different rules and comparing notes.

A race to the bottom is the claim that states, competing for residents and businesses, will cut taxes, protections and services until everyone is worse off. We'll come back to that.

The argument in numbered premises

1. Nobody knows in advance which policies will work. Social policy is hard. Smart, well-meaning people disagree, and the results of a new program often surprise its designers.
2. A single national policy turns every mistake into a national mistake. If Washington guesses wrong, all fifty states pay, and there is no comparison case to show what the alternative would have done.
3. Multiple jurisdictions allow trial and comparison. When states choose differently, the results can be observed side by side, and successes can spread while failures stay contained.
4. People can vote with their feet. When states differ, families and firms can move toward the policies they prefer, which gives states feedback no election alone provides.
5. The Constitution's structure preserves this capacity by limiting national powers to an enumerated list and leaving the rest to the states.
6. Therefore federalism is a mechanism for discovering good policy, and centralizing a question should require a strong showing that a single national answer is needed.

Each of those premises can be checked against the text and the record.

The constitutional design

Start with Madison's promise in Federalist 45:

The powers delegated by the proposed Constitution to the federal government are few and defined. Those which are to remain in the State governments are numerous and indefinite.

He went on to explain that the national powers would be exercised principally on external objects, such as war, peace, negotiation and foreign commerce, while the states' powers would extend to all the objects which, in the ordinary course of affairs, concern the lives, liberties and properties of the people. The full reading is in Few and Defined, the Promise About Federal Power in Madison's Federalist 45.

The Tenth Amendment, ratified in 1791, wrote the principle into the text:

The powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the States respectively, or to the people.

And Federalist 51 explained why the division protects liberty, not just efficiency:

In the compound republic of America, the power surrendered by the people is first divided between two distinct governments, and then the portion allotted to each subdivided among distinct and separate departments. Hence a double security arises to the rights of the people.

Double security. Separation of powers is one lock; federalism is the second. A national majority that wants to impose a policy on every American has to get through both. And because the states keep "numerous and indefinite" powers, they are free to try different answers to the same questions.

Brandeis gave the idea its famous name

The phrase most people know comes from a dissent. In New State Ice Co. v. Liebmann in 1932, the Supreme Court struck down an Oklahoma law that required anyone selling ice to obtain a state license showing a public need, effectively protecting existing ice companies from new competitors. Justice Louis Brandeis dissented and wrote the line that has been quoted ever since:

It is one of the happy incidents of the federal system that a single courageous State may, if its citizens choose, serve as a laboratory; and try novel social and economic experiments without risk to the rest of the country.

There is a delicious irony here that conservatives should enjoy. The experiment Brandeis wanted to protect was a licensing scheme that shielded incumbent businesses from competition, the kind of policy free-market thinkers usually oppose. The majority that struck it down was the conservative wing of the old Court. Yet Brandeis's principle is the right one, and it cuts in every direction. A state may try a bad idea, and its neighbors may watch it fail. That is the point. The laboratory works only if the experiments are allowed to run, including the ones you expect to fail.

The record: experiments that spread and experiments that stopped

Theory is fine. Here is what the laboratory has actually produced.

Women's suffrage. Wyoming Territory gave women the vote in 1869, half a century before the Nineteenth Amendment was ratified in 1920. Other Western states followed one by one. By the time the national amendment came, the experiment had been running for decades, and its results were plain to see.

Welfare reform. In the late 1980s and early 1990s, Washington granted states waivers under Section 1115 of the Social Security Act to test new approaches to cash assistance, including work requirements and time limits. Wisconsin, under Governor Tommy Thompson, was among the most aggressive, and its program became a model. The national reform signed on August 22, 1996, the Personal Responsibility and Work Opportunity Reconciliation Act, replaced the open-ended Aid to Families with Dependent Children program with Temporary Assistance for Needy Families, a block grant that gave states wide latitude. The national law codified what the states had already tested.

School choice. Milwaukee launched the first modern publicly funded voucher program in 1990. Minnesota passed the first charter school law in 1991. Both ideas spread state by state, with each legislature adjusting the design. The debates continue, and they should; the rural objection, for example, is serious and worth answering on its own terms, as School Choice Fails Rural Kids? Here Is the Rural Objection Answered does.

Notice the pattern. None of these began as a national mandate. Each started in a single state or city, produced results others could see, and spread or stalled on the merits. Imagine instead that Washington had imposed a single national welfare system design in 1988, or a single national school-choice rule in 1990. There would have been nothing to compare it to, and no way to know what the alternative would have done.

The interstate market in good government

Premise four, voting with your feet, deserves its own moment. When states differ in taxes, regulation, schools and public safety, families and businesses can move. That movement is information. It tells governors and legislators which policies people value enough to relocate for, and it puts pressure on states that drive residents away.

This is why centralizers dislike federalism. A state that tries a high-tax, high-regulation model has to keep its residents from leaving; a national government that imposes the same model everywhere does not. The only exit left is emigration from the country, which is a much higher bar. Federalism keeps the exit door close and cheap, and a government that knows its citizens can leave behaves better than one that knows they cannot.

The Supreme Court has protected this structure with what it calls the anti-commandeering doctrine. In New York v. United States in 1992 and Printz v. United States in 1997, the Court held that Congress may not order state legislatures to legislate or state officers to carry out federal programs. In Murphy v. NCAA in 2018 it applied the same rule to strike down a federal law forbidding states to authorize sports betting. The principle is simple: Washington may pursue its own enumerated powers with its own officers, but it may not conscript the states as its field offices.

The race-to-the-bottom objection, stated fairly

Here is the strongest case against everything above.

"Interstate competition does not produce good policy; it produces a race to the bottom. States, competing for mobile capital, will slash taxes, gut environmental rules, weaken labor protections and cut benefits for the poor, because any state that holds the line will lose businesses and wealthy residents to its neighbors. The poor cannot move as easily as the rich, so they bear the cost. The history of federalism includes states' rights arguments used to defend segregation. A single national standard protects the vulnerable from fifty separate auctions of their rights."

That is the strongest version, and parts of it deserve respect. Here is the answer.

Answering the race to the bottom

1. Competition runs on more than one dimension. States do not compete only on low taxes; they compete on schools, safety, infrastructure, courts and quality of life. A state that guts its schools to cut taxes may lose the families it wanted to attract. The "bottom" assumes residents care only about price, which no one believes about their own choice of where to live.

1. The corporate-law debate shows the race can run upward. In a famous 1974 law review article, William Cary argued that Delaware had won the competition for corporate charters by racing to the bottom on shareholder protection. Ralph Winter answered a few years later that states competing for charters must attract investors too, so the race could just as easily run toward better rules. That debate has never ended, but it shows the conclusion is not automatic.

1. The Constitution sets the floor for rights. Federalism never meant states could violate the Constitution. The Fourteenth Amendment binds every state to due process and equal protection, and Section 5 gives Congress power to enforce it. Brown v. Board of Education in 1954 and the Civil Rights Act of 1964 enforced those guarantees against states that defied them. Segregation was not federalism working; it was states violating constitutional rights, and the national government's enumerated power to enforce the Fourteenth Amendment was the proper answer.

1. National uniformity can fail just as badly, with no escape. When a national policy is mistaken, there is no neighboring state to flee to and no comparison to reveal the error. A race to the bottom is a risk; a single national mistake is a certainty with no off-ramp.

1. Mobility is a protection for the poor, too. People of modest means move across state lines in large numbers for work and housing. A state that drives out jobs and makes housing unaffordable hurts the poor first, and the ability to move to a state with better prospects is one of the few levers ordinary families have against a failing government.

When a national answer is right

None of this means every question belongs to the states. The Constitution assigns national defense, foreign affairs, interstate and foreign commerce, the currency, naturalization and the protection of constitutional rights to the national government, and it should handle them. The question is the default. The Framers set the default at the state level and made Washington show its enumerated power before acting. That default is worth defending, because once a question is nationalized, it almost never goes back, and the laboratory for that question closes for good.

For how this plays out in the courts that decide most of the law Americans actually live under, see State Supreme Courts Decide More of Your Life Than the Nine in Washington, and for the text that anchors the division, The Tenth Amendment Left Out One Word, Expressly, and the Omission Mattered.

The case in brief

1. Federalist 45 promised that federal powers would be few and defined and state powers numerous and indefinite, and the Tenth Amendment wrote that division into the text.
2. Federalist 51 called the division a double security for liberty, and Brandeis in 1932 called the states laboratories that can experiment without risk to the rest of the country.
3. Women's suffrage, welfare reform and school choice all started in a single state or city and spread on their results before any national rule.
4. The race-to-the-bottom objection ignores that states compete on quality as well as price and that the Fourteenth Amendment sets a constitutional floor no state may go below.
5. A national mistake has no comparison case and no exit, so the default should stay with the states unless the Constitution assigns a question to Washington.

What you can do: pick one policy your state handles differently from a neighboring state, find the bill that created it on your legislature's website, and ask your state legislator what results it has produced.