Blog

Three Towns, Three Rule Books and the Case for Letting States Differ

Three imaginary towns show how federalism lets families choose, lets ideas get tested cheaply, and what it costs when a state chooses badly.

Every year, some American family rents a truck, loads the couch, and drives across a state line for reasons that have nothing to do with the weather. Maybe the property tax bill got out of hand. Maybe the school down the road stopped teaching the basics. Maybe a cousin said that in the next state over you do not need a government license to open a shop. Nobody sends a memo to Washington. Nobody holds a referendum. The family just votes with its feet, and some governor three states away starts getting nervous when he sees where the trucks are going.

That, my friends, is federalism with its working clothes on. It is the least glamorous and most powerful idea in the American system, and it is the answer to a question the country argues about constantly: should one rule apply to 330 million people, or should different places be allowed to do things differently? I want to make the case for letting states differ, and I want to do it with three imaginary towns, real arithmetic, and an honest accounting of what it costs.

The argument in four sentences

Here is the case, stated plainly so you can check each step.

  1. A rule written in Washington applies everywhere at once, so if it is wrong it is wrong everywhere.
  2. A rule written in a state applies there only, so if it is wrong the damage stays local and the other forty-nine states can watch and learn.
  3. Families can move between states far more easily than they can move between countries, so a state that taxes heavily and delivers poorly pays a price that a national government never pays.
  4. The Constitution was built this way on purpose: it gives the national government a list of powers and, in the Tenth Amendment, reserves the rest to the states or the people.

The first two are about learning. The third is about accountability. The fourth is about law. All of them are testable, so let us test them.

The laboratories

In 1932 the Supreme Court decided New State Ice Co. v. Liebmann, a case about whether Oklahoma could limit who was allowed to go into the ice business. The majority said no. Justice Louis Brandeis dissented, and wrote one of the most quoted sentences in American law: 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."

Brandeis was no conservative, which is worth remembering whenever somebody says that federalism is a Republican trick. He wanted states to try things. The point holds no matter who is trying. In 2006, Massachusetts passed a health insurance law under Governor Mitt Romney, years before the national Affordable Care Act. In the early 1990s, several states got federal waivers to experiment with work requirements in welfare before Congress rewrote the national rules in 1996. We tell the story of that law in two vetoes, one signature and the 1996 welfare law that put marriage in statute. Whether you loved or hated those experiments, the country got to look at them before betting everything.

A second thinker finishes the thought. In 1956 an economist named Charles Tiebout published "A Pure Theory of Local Expenditures," arguing that when people can move among communities, local governments end up competing, and residents sort themselves by the bundle of taxes and services they prefer. Albert Hirschman gave the idea its popular names in 1970: exit, voice and loyalty. You can complain (voice), or you can leave (exit). A federal system gives you both.

Meet the three towns

Now let me build a classroom. These towns are hypothetical. They sit in three imaginary states, and every number is invented to show the logic, not to describe any real place. Two simplified households live in each.

Household H1 is a married couple with a house worth $250,000, a household income of $70,000, and $30,000 a year in purchases that are subject to sales tax. Household H2 is a young renter with an income of $40,000 and $20,000 a year in taxable purchases. To keep the arithmetic clean I ignore deductions, credits, and the fact that landlords pass some property tax into rent.

Town One: Oakdale

Oakdale's state has no income tax. It leans on the other two legs of the stool. The sales tax is 7 percent, the property tax is 1.6 percent of a home's value. Schools are funded by the district but parents get a choice of public, charter and private options. A state license is required to practice medicine or law, but not to braid hair, wash windows or decorate cakes.

  • H1 pays sales tax of 7 percent of $30,000, which is $2,100, plus property tax of 1.6 percent of $250,000, which is $4,000. Total: $6,100.
  • H2 pays 7 percent of $20,000, which is $1,400, and no income tax. Total: $1,400.

Town Two: Brookfield

Brookfield's state has a flat 4 percent income tax, a 4 percent sales tax, and a property tax of 0.8 percent. Schools are run by the traditional local district with a strong elected board. Hairdressers need a license requiring 100 hours of training.

  • H1 pays income tax of 4 percent of $70,000, which is $2,800, plus sales tax of 4 percent of $30,000, which is $1,200, plus property tax of 0.8 percent of $250,000, which is $2,000. Total: $2,800 + $1,200 + $2,000 = $6,000.
  • H2 pays 4 percent of $40,000, which is $1,600, plus 4 percent of $20,000, which is $800. Total: $2,400.

Town Three: Cedar Harbor

Cedar Harbor's state runs a graduated income tax that starts higher, which for this example I round to a flat 6 percent. The sales tax is 2 percent, and the property tax is 0.5 percent. The state sets a uniform curriculum and funds schools heavily from the capital. The state licenses nearly every trade, and a cosmetology license requires 1,500 hours of training.

  • H1 pays income tax of 6 percent of $70,000, which is $4,200, plus sales tax of 2 percent of $30,000, which is $600, plus property tax of 0.5 percent of $250,000, which is $1,250. Total: $4,200 + $600 + $1,250 = $6,050.
  • H2 pays 6 percent of $40,000, which is $2,400, plus 2 percent of $20,000, which is $400. Total: $2,800.

What the numbers show

Look at H1, the homeowners: $6,100, $6,000, $6,050. Almost the same bill in all three towns. A national "tax fairness" standard would have a hard time declaring any of these obviously right or wrong. They simply put the burden in different places.

Now look at H2, the renter: $1,400, $2,400, $2,800. The same young person pays double in Cedar Harbor what she pays in Oakdale. Different states chose different bargains, and a person's life stage changes which bargain suits her. A homeowner may like Brookfield, where the property tax is low and he gets a school board he can reach by phone. A renter with no kids may prefer Oakdale. A family that wants a heavy state hand in school curriculum may like Cedar Harbor.

That is the first lesson: there is no single right tax mix. There are trade-offs, and different people weigh them differently. Economists like Thomas Sowell have made the point for years that policy is about trade-offs, not solutions, and federalism lets each place make a different trade. Our piece on Thomas Sowell Asks What Happens After the Policy Passes covers that idea.

What a family decides

Let me turn the examples into decisions, the way an actual family has to. Again, these are hypothetical households.

The tradeswoman. A woman wants to open a hair-braiding business. In Oakdale she can start tomorrow. In Brookfield she needs 100 hours of training, which at 40 hours a week is two and a half weeks. In Cedar Harbor she needs 1,500 hours, which at 40 hours a week is 37.5 weeks, nearly nine months, before she can legally charge anyone a dime. If your goal is to open a business on a shoestring, you see the difference at once. The effect of that gap on ordinary jobs is the subject of the license you need to braid hair and the jobs it quietly kills.

The parents. A couple with two children chooses a town by its schools. In Oakdale they can pick a school for their child. In Brookfield they can run for the school board and change the one in town. In Cedar Harbor they accept a statewide plan, which might be excellent. The decisions differ, and so does the kind of parent who is happiest. See The Constitution Never Mentions Schools and That Is the Whole Point for why the Constitution left this to the states in the first place.

The retiree. A retired couple with a paid-off house but a modest income looks at Oakdale's heavy property tax and Cedar Harbor's heavy income tax. They run the numbers and pick Brookfield. They are not irrational or selfish. They are doing what Tiebout predicted.

Now step back. In each case the choice is made by the people who live with the consequence, at the smallest scale where it can be made. That is the old idea of keeping decisions close to home, which Subsidiarity Is the Old Idea Behind Keeping Power as Close to Home as Possible unpacks.

The honest cost: some states will choose badly

Here is where a lot of conservatives cheat, and I will not. Federalism does not guarantee good policy. It guarantees that bad policy will be local. Some state will pass a law that is foolish, wasteful or cruel, and its residents will pay for it, and some of them will not be able to leave. A child cannot choose his state. A widow with a house she cannot sell cannot exit as easily as a young engineer with a laptop. Tiebout's model assumes mobility, and mobility is not equal.

There is a darker chapter, and anyone who defends states' authority has to face it. "States' rights" was the banner of slavery before the Civil War and of segregation afterward. The answer to that history is not to abandon federalism. It is to remember what the Reconstruction amendments did: the 13th (1865) ended slavery, the 14th (1868) guaranteed citizenship, due process and equal protection against the states, and the 15th (1870) forbade denying the vote on grounds of race. These set a national floor. Federalism is the freedom to differ above that floor, not the freedom to dig below it. The line between a state's choice of tax rate and a state's denial of someone's basic rights is exactly the line the Fourteenth Amendment draws.

A second cost is spillover. Sometimes what one state does affects its neighbor, as with a river that crosses borders, or a trade barrier. The Constitution gives Congress the power to regulate commerce among the states precisely for such problems. Federalism is not the claim that Washington never has a role. It is the claim that Washington's role is limited and listed. See A Farmer's Wheat Crop and the Long Fight Over the Commerce Clause for how far that clause has stretched.

And a third: the federal government can buy its way around the whole structure by attaching strings to money. That is the subject of the next piece in this series, The Drinking Age Story Shows How Washington Buys Control of the States. The more a state depends on Washington's grants, the less it differs from the next one.

The strongest objection, answered

The best argument against all this goes like this: rights should not depend on your zip code. A person in one state should not enjoy fewer protections than a person in another. National problems need national solutions, and "let the states decide" has too often meant "let the powerful decide."

I accept half of it. Basic rights, the ones in the Constitution as amended, belong to every American, and the federal courts exist to enforce them. But the objection proves too much if it is stretched to every disagreement. If every policy question that could be nationalized must be nationalized, then the country gets a single rule on everything from school calendars to zoning, written by people who have never seen your town and enforced by people who answer to nobody you elected. The 50-lab system is not a defect that we tolerate. It is how a continent-sized republic stays free.

There is also a quiet advantage to the arrangement that rarely gets mentioned: it lowers the stakes of national elections. When one party loses the presidency, but still governs a state, its voters keep a stake in the system. A country in which every decision is made in one capital is a country where every election is a war. As a matter of social peace, federalism is the cheapest insurance policy we have, a point that runs through The Tenth Amendment Rule That Protects Red States Also Protects Sanctuary Cities. And if you want to see the same logic applied to state lawmaking, our explainer how a bill becomes a state law, and where a citizen can step in shows where a resident can plug in.

The bottom line

  • Federalism is not a slogan. It is a design that keeps mistakes local, lets states copy what works, and gives families the power to leave a place that treats them badly.
  • Three towns can tax the same family almost identically and still be worlds apart in who pays, how schools work, and who can start a business. That is a feature, because people differ.
  • Be honest about the cost: some states will choose badly, and the answer for basic rights is the national floor in the Constitution, not a national rule for everything.
  • When someone says "there ought to be a national law," ask three questions: what power in the Constitution authorizes it, what do we lose if the states differ, and what happens if the law is wrong?
  • Use your own state. Read your legislature's website, call your representative, and show up. The laboratory only works if the neighbors watch.

Latest on this topic