Brandeis Called the States Laboratories and Conservatives Took Him at His Word

Here is the most quoted sentence in American federalism, and it comes from a case about ice. Not ice as in a frozen lake or a polar cap, but ice as in the stuff you put in a cooler. In 1932 the Supreme Court looked at an Oklahoma law that required a government license to sell it, and a justice who is a hero to the progressive left wrote a dissent that conservatives have been quoting back at Washington ever since.

I find that delicious. The man who gave us the phrase "laboratory of the states" was defending a regulatory scheme that, by modern lights, looks like a textbook licensing cartel. And yet his line is right, and it is more useful to the people who disagreed with him than to the people who agreed. Let me explain why, and then let me tell you what the line cannot do, because a conservative who uses it as a magic wand will get himself into trouble.

The case behind the quote

The case is New State Ice Company v. Liebmann, decided in 1932. Oklahoma had declared the manufacture, sale and distribution of ice a business "affected with a public interest" and required anyone who wanted to enter it to get a license by showing the state that the public needed another ice dealer. The existing companies, you will not be shocked to learn, were happy with the arrangement. A man named Liebmann started an ice business in Oklahoma City without a license, and the New State Ice Company went to court to stop him.

The Supreme Court, with Justice George Sutherland writing, struck the law down. The majority held that making ice was an ordinary business, not a public utility like gas or water, and that the state could not use the license to close the door to newcomers. That was a victory for economic liberty, and I'd note that our friends who cheer the dissent should remember which side of the case lost.

Justice Louis Brandeis dissented. His argument was not that ice was special. It was that the Court should be reluctant to stop a state from trying something, even something unwise. He wrote that 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. He added that stopping such experiments is a grave responsibility for a court, because the nation needs room to reshape its economic practices as conditions change.

Look at what happened there. The author of the line was defending an experiment that, if you ask me, was a bad one. The majority had the better of the particular case, and the dissent has the better of the general principle. You can hold both thoughts in your head. In fact, a good conservative should, because the principle is the more valuable property.

Why the principle is conservative at its core

Here is the argument in numbered premises, since it is easier to attack a clear argument than a foggy one.

1. Nobody knows in advance which policies work. Not economists, not governors, not editorial boards. Every confident prediction about the effect of a law is a hypothesis.
2. A hypothesis is cheapest to test on a small scale. If you try a thing in one state and it fails, one state pays. If you try it nationally and it fails, everyone pays, and it is much harder to undo.
3. Different people want different things. A farm state and a financial state, a state of ranchers and a state of shipbuilders, have different needs and different preferences. A single national rule satisfies none of them well.
4. People can leave. A citizen who dislikes the rules of one state can move to another, which is a form of accountability a national policy does not offer.
5. Therefore, the more decisions made close to home, and the more states are allowed to differ, the more the country learns and the less one mistake costs.

That is not a progressive argument or a conservative one. It is an argument about how to discover things. But it lands hardest with conservatives because we are the ones who say that power should be as close to the people as possible, and that the people closest to a problem know most about it.

The Constitution backs it up

This is not a fancy theory. It is in the text. The Tenth Amendment says that 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. The federal government has the powers on a list (see Congress Can Only Do What the Constitution Lists and Madison Said So) and the states hold the rest. The Supreme Court once described the amendment as stating a truism, in United States v. Darby in 1941, meaning it says what the structure already implies. But the Court has also used the structure to protect state autonomy. The "anti-commandeering" cases, including New York v. United States in 1992, Printz v. United States in 1997 and Murphy v. NCAA in 2018, hold that Congress may not order state legislatures or officers to carry out federal programs. Congress can regulate people. It cannot conscript a state.

That is the legal floor under the laboratory. If Washington could dictate every policy, the states would be field offices, and a field office does not run experiments. For the broader story of how this fits together in a single town, see Three Towns, Three Rule Books and the Case for Letting States Differ.

How a good idea travels

The beauty of the laboratory is the part after the experiment. A state tries something. Other states watch. If it works, they copy it. If it fails, they steer clear. Over time the best ideas spread without anybody in Washington having to order it.

Look at the record, and look at it with some humility, because the laboratories have produced results across the spectrum.

• Women's suffrage. Wyoming gave women the vote in 1869, while still a territory, decades before the Nineteenth Amendment in 1920. Other Western states followed before the national amendment came.
• Workers' compensation. In 1911 Wisconsin enacted a workers' compensation law that survived court challenge, and the idea spread to the other states.
• Welfare reform. Before the national welfare law of 1996, states like Wisconsin and Michigan tried work requirements and time limits under federal waivers, and the results shaped the national bill. That story is told in How the States Rewrote Welfare Before Washington Finally Caught Up.
• School choice. Milwaukee created the first modern voucher program in 1990, and the idea spread from there.
• Tax limits. California's Proposition 13 in 1978 set off a tax revolt that reached statehouses across the country.
• Health insurance mandates. Massachusetts enacted a statewide health insurance law in 2006, under a Republican governor, and it became a template for the national law of 2010. Conservatives who dislike that national law should note the lesson: the laboratory produces things you like and things you do not, and the trouble starts when a state experiment is promoted to a national mandate before it has been tested.

That last example is the sobering one. A laboratory is useful because it contains the experiment. When a mistake is exported nationally before the results are in, the containment fails.

The freedom to fail

Now the line most people skip: the argument works only if the states are allowed to fail. That is the hard part of federalism, and it is where I'd press my fellow conservatives hardest.

If you believe in the laboratory, you have to accept that some states will try ideas you think are terrible. A state may raise taxes to the heavens, regulate a business into the ground or adopt a policy that you believe is a catastrophe. Your remedy is not to ask Congress to nationalize your preference. Your remedy is persuasion, and, if all else fails, the moving van.

I like to call this the consistency test. When your side is out of power in Washington and a state is trying something you favor, you cheer federalism. When your side is in power in Washington and a state is trying something you dislike, do you still cheer? The test of a principle is whether you hold it when it costs you. Plenty of politicians on both sides pass the test only when it is convenient.

The honest answer is that conservatives, like everyone else, have sometimes been tempted to override the states when we disagree. Federalism is not a trump card for every dispute. But it is the right starting presumption, and the burden should be on the person who wants to override it.

There is a second reason the freedom to fail matters. Competition among states disciplines them. A state that mismanages its economy loses residents and employers to its neighbors. That discipline works only if states face real differences in taxes and rules, which is the argument made in Interstate Competition Is Why States Cannot Afford to Be Lazy.

The honest limits

Here is where a serious writer owes you the other side, because "let the states decide" is not the answer to everything, and a conservative who thinks it is will eventually have to defend something indefensible.

The Fourteenth Amendment. Ratified in 1868, it forbids any state to deny to any person within its jurisdiction the equal protection of the laws, or to deprive anyone of life, liberty or property without due process. That binds the states. It is why the laboratory has walls. A state is not free to experiment with segregated schools, which the Supreme Court held unconstitutional in Brown v. Board of Education in 1954, or to ban marriage between people of different races, which the Court struck down in Loving v. Virginia in 1967. The states that tried those experiments had no right to run them. The original federal system, with slavery as a "state" matter, is the most terrible argument against treating "states' rights" as a freestanding principle. The rights at stake belong to individuals, and no state may take them.

Federal supremacy and the interstate commerce power. The Supremacy Clause makes federal law the supreme law of the land when Congress acts within its powers. States cannot undercut it, and cannot discriminate against interstate commerce.

Spillovers. Some problems cross borders: a river that runs through five states, pollution carried by wind, a criminal who flees to the next county. The laboratory metaphor assumes the experiment is contained, and when it is not, a broader authority may be needed. Conservatives should be careful here, because the claim "it's a spillover" has been stretched by the federal government to cover nearly everything.

So the right rule is not "the states always decide." It is "the states decide unless the Constitution says otherwise, and the Constitution says otherwise less often than Washington would like." The Fourteenth Amendment protects individual rights nationally. Everything else, a wide range of economic and social policy, is the province of the people in their states.

What it means to take Brandeis at his word

To take Brandeis at his word is to accept both halves of the sentence. The first half is the part we like: a single state may try novel things without risk to the rest of the country. The second half is the one we sometimes forget: that includes novel things we think are foolish.

Conservatives have done well as a movement in the laboratories. School choice, welfare work requirements, tax limits and right-to-work laws all started in a state, in some cases decades before they became a national conversation. The tools of the movement have also been tested on the ground, in places where an error cost one budget and not a continent.

So the practical conclusion is to build from the bottom. Win a school board, a county seat, a state house. Pass a reform. Let the neighbors watch. If it works, they will copy you without being asked. If it fails, you will have learned something without ruining a country. And when someone in Washington proposes to settle every question by decree, quote the dissent in the ice case. It was written by a man the other side admires, and it says what we have been saying all along.

The bottom line

• The line about a state serving as a laboratory comes from Justice Brandeis's dissent in New State Ice Co. v. Liebmann in 1932, in a case about a license to sell ice, which his side lost and which the majority arguably got right.
• Federalism works only if states may try different things and fail. The Tenth Amendment and the Court's anti-commandeering cases protect that space.
• Apply the consistency test: defend the right of a state to do what you dislike, not only what you like, and save your fight for persuasion.
• The Fourteenth Amendment and federal supremacy set real limits. A state may experiment with policy, not with the equal rights of its citizens.
• Start where you live. Reform that works in your county or state becomes a model that others copy, and that is how the best ideas have always spread.