University of Montana School of Law

Ilya Somin*


I would like to start by thanking the University of Montana School of Law and the Montana Law Review for organizing this exciting symposium and for honoring me with the invitation to deliver this keynote speech. I also welcome this opportunity to visit the great state of Montana for the first time ever. I notice that you now actually have much nicer weather than we usually do in Washington, D.C., which is a pleasant surprise.

Tonight, I will focus on some important issues related to the subject of the Symposium: federalism and limitations on federal government power. In particular, I will suggest that in discussing these issues, we don’t pay enough attention to the relationship between federalism and voting with your feet. Paying proper attention to this connection strengthens the case for de-centralizing political power to lower levels of government. It should also affect our attitude toward judicial enforcement of federalism-based limits on the power of the central government.

In first part of the presentation, I will briefly discuss what voting with your feet is, and how it works. I will then describe how increasing opportunities for foot voting can enhance our political freedom. Empowering foot voters to decide on more issues will also enhance how well informed their decisions are relative to deciding those same issues purely through more conventional voting.

Later in the presentation, I will note some possible limitations of foot voting. They are genuine problems, but not nearly as severe as many people claim.

Finally, towards the end, I consider the implications of all this for constitutional federalism and for the appropriateness of judicial review of federalism issues for the purpose of enforcing limits on the power of the central government.


First things first. What is this voting with your feet or, as I call it in much of my work, “foot voting”?

At least as I use it in my work, “foot voting” denotes situations where we can choose what sort of government or policy we want to live under by deciding where we want to live, and, for example, we can do so by choosing among different states. Many people, for instance, historically migrated to Montana for reasons like this. You can also exercise foot voting when you choose what local government to live under. In addition, one can vote with their feet in the private sector. We do so every day when we decide what sorts of products to buy or what kind of civil society organizations we want to be associated with, such as a church or a synagogue or a private school, professional association, and so forth. I have elaborated on this in greater detail in my book, Democracy and Political Ignorance,[1] which came out last year and also in the more recent work about the relationship between federalism and political freedom.[2]

Although it is called “foot voting,” in some cases it does not require any actual movement. In the private sector, we can often vote with our feet without actually physically going anywhere. You can pick up the phone or go to a website, and get a different service provider or join a new civil society organization.

What differentiates foot voting from conventional ballot-box voting is the fact that it is an individual choice as opposed to a joint decision that you make with thousands or millions of other people where your vote is just one of a great many others. This distinction is crucial to the advantages of foot voting that I focus on later in tonight’s presentation.


Foot voting in a federal system can play a vital role in enhancing our political freedom. Most Americans have read the Declaration of Independence where it says governments “[derive] their just powers from the consent of the governed.”[3] But too often we do not reflect very much on what that means. We tend to assume that the consent of the governed and political freedom more generally can be achieved through ballot-box voting. If we get to vote for who is in government, that means we have political freedom. It also means that we have consented to the government’s policies.

Unfortunately, however, that is an inadequate account of both political freedom and the consent of the governed. In a sense, you are indeed exercising political freedom when you go to the ballot box and choose between A and B, Democrats and Republicans, or whatever third parties might happen to be on the ballot. But it is a very peculiar kind of freedom, because the chance that your vote will make a difference to the outcome of the election is infinitesimally small. In an American presidential election, the chance that your vote will have a decisive effect is about 1 in 60 million.[4] In state or local elections, it is higher than that but still extremely low.

With most other important freedoms, we would not consider that the freedom is in any way meaningful if your chance of actually making a difference is 1 in 60 million or even 1 in 60,000. Take religious freedom: if you only had a 1 in 60 million chance of being able to determine what religion you want to practice, we would not say that you have meaningful religious freedom. Further, if you only have a 1 in 60 million chance of being able to determine what sorts of opinions you can express in your speech, we would not believe that you have genuine freedom of expression. The same goes for other important forms of freedom.

With respect to political freedom, this suggests that the degree of political freedom you have, when your only way of exercising it has just a 1 in 60 million chance of making a difference—or even a 1 in 60,000 or a 1 in 10,000 chance—is extremely limited and grossly inadequate. That does not mean that the vote is completely worthless. We would probably be even less free without it. But it is far from sufficient.

Voting with your feet greatly diminishes this problem. When you decide whether you want to live in Montana or some other state, or whether you want to live in this city or one of the other cities within the same state, that is a decisive individual choice. Even if 6,000 other people or 60 million other people do not agree with you, you still get to decide which of these sets of government policies you want to live under.

Granted, many of you, like me, are married. Obviously my wife might be able to veto my choice. And probably will. Her judgment is better than mine. But, even so, 1 vote out of 2, or 1 vote out of 5 or 10 or 15, if you have a large family, that still is much more meaningful than 1 vote out of 10,000 or 1 vote out of 10 million.

Your chances of exercising genuinely meaningful freedom are much greater when you vote with your feet than when you vote at the ballot box. Foot voting is more like the way we exercise other freedoms that we consider important—freedom of speech, freedom of religion, and others—and less like the very ephemeral kind of freedom where all we can do is make a decision that has almost no chance of actually making any difference.

We reach the same conclusion when we focus on the question of consent. Is it really meaningfully consensual when we choose our government at the ballot box? It is difficult to argue that you have genuinely consented to the outcome in any meaningful way when all you have done is make a decision that has only a tiny little chance of making any difference.

Some people argue that the very fact that you voted means that you have given your consent to the resulting government policies.[5] I don’t think this actually follows in any significant way. For one thing, the government still claims authority over you even if you haven’t voted. If you choose not to vote, you can’t then say, “Well, any laws made by the winners, they won’t be binding on me.” So if it’s not even a meaningful choice in that sense, it severely undercuts the validity of the argument.

In addition, even if you did vote, it may be that you voted not because you consented to the overall structure of the system but simply because the system is in place. At least in the short run, there’s nothing you can do to change it, so you say, “Well, at least I want to vote for the lesser of the two evils that happens to be on offer.” But that doesn’t mean you’ve actually consented to either one of the evils in place; it just means that the choice is before you, and you have decided to make a decision between them, but it’s not really meaningful consent to any significant degree.

On the other hand, when you choose to vote with your feet, you do have a meaningful opportunity to make a choice that makes a real difference. You also have a potentially wider range of options before you: hundreds of different cities, and 50 different states. If it is a choice in the private sector, there are even more options.

Even with respect to federal government policies, one can argue that at least you consented to live in the country as a whole, because you have not left. That is a kind of “tacit consent.” But obviously the burden of having to leave the country entirely is vastly greater than the burden of leaving a state or a locality or simply choosing some new option in the private sector.[6] While your consent may not be perfect when you vote with your feet in the federal system, or even when you choose within a private sector, it is at least a much more meaningfully consensual choice than decisions you make at the ballot box.

Perhaps there is no way to achieve completely consensual government. But we at least get closer to that ideal when we make more of our decision by voting with our feet.

If we consider political freedom to be an important value and not just something that we intone on the Fourth of July but do not really care about, and to the extent we want to take the ideals of the Declaration of Independence seriously, foot voting enables us to realize those ideals much more fully than ballot-box voting alone. If we care about these values, we may want to make more of our decisions by voting with our feet and fewer at the ballot box. And one way we can do so is by de-centralizing political power to lower levels of government so more decisions can be made by voting with your feet and fewer at the ballot box.


In addition to giving people more meaningful choice, foot voting also leads them to make better informed decisions than they would make by voting at their ballot box. When we exercise political choice, it is important we have genuine freedom, and exercise it in at least a reasonably informed way.

A. The Problem of Rational Political Ignorance

One of the major shortcomings of ballot-box voting is the problem of political ignorance. People may be voting, but most of the time they have very little knowledge of what it is they are voting about. For example, right before the crucial 2014 mid-term election that determined control of both houses of Congress, a survey conducted by the Annenberg Public Policy Center found that only 38% of Americans knew that the Republican party controlled the House of Representatives, and a very similar number knew that the Democrats controlled the Senate.[7] It is difficult to evaluate the performance of the two major parties, the Senate, and the House if you do not know which party controls these chambers of the legislative branch.

This kind of widespread political ignorance is common. During our last mid-term election in 2010, most voters said that the most important issue they were considering was the state of the economy. Yet two-thirds of the public did not even know that the economy had grown rather than shrunk during the year right before the election.[8]

Another big issue that was much considered at that time was the government’s policies in response to the recession and the financial crisis, of which the biggest of those policies was the TARP bailout enacted in the fall of 2008. But only 34% of Americans knew former President Bush was the president who had pushed TARP through Congress. More people than that thought it was President Obama, so they weren’t even in a position to properly allocate responsibility for this major government policy.[9]

This ignorance is not limited to particular public policy issues. It also extends to the basic structure of government and the way it operates. For example, that same recent Annenberg poll found that only 36% of Americans can even name the three branches of the federal government: the executive, the legislative, and, my personal favorite, the judicial.[10]

In addition, many surveys showed that the public has very little understanding of which government officials are responsible for which issues. If you do not know that, it is actually hard to allocate responsibility for policy successes and failures.[11]

Some may think that this must be a recent problem. Maybe it is the fault of the millennial generation. Unlike previous generations, perhaps the millennials are spending too much time surfing the Internet, watching “twerking” videos, and following celebrities like Miley Cyrus. As a result, they are not studying up on important political issues like we did in our day. But the available survey data suggests that the problem of political ignorance is roughly similar to what it has been in the past. Today’s levels of political knowledge are about the same as they were fifty to sixty years ago, in the early days of modern public opinion polling. It is, however, striking that political knowledge levels have stagnated even as educational attainment has greatly increased and even as the Internet and other modern technologies make more data more easily available than probably ever before in human history.

It is even the case that IQ scores have gone up over the last 50 years.[12] It is often suggested that the younger generation must be dumber than their elders.[13] But it isn’t true. IQ scores have actually risen. The millennial generation may not be the best informed generation we have ever had, but they may well be the most intelligent generation. But that intelligence, for the most part, has not been used to acquire more political knowledge. It has instead been used for other purposes.

It is in fact rational for both millennials and the rest of us to behave that way most of the time for the very reason I noted earlier. The chance that your vote will make a difference in an election is infinitesimally small. Therefore, it is irrational to spend more than a tiny amount of time paying attention to political information if your only reason for doing so is to be a better voter. Economists call this “rational ignorance.”[14] When there is little or no benefit to acquiring more information, it makes sense to devote your time to other sorts of activities that are more meaningful and more likely to have an actual effect.

Obviously, there are people who learn about things even though they have very little chance of affecting the outcome. Consider sports fans. Sports fans love to learn about their favorite team and also to cheer for it and against its rivals. They therefore learn a great deal about their team and its opponents, even though they recognize the reality that their actions have very little chance of affecting the outcome of games.

There are people who acquire political knowledge for much the same reasons as sports fans acquire sports knowledge. In my book, I call them “political fans”: people who follow politics not because it makes them better voters but because they find it interesting and entertaining.[15] They enjoy cheering on their favorite political team—a party, an ideology, or sometimes an interest group or the like. Such political fans generally know much more about politics than the average voter, and the more interested you are in politics, the higher your level of political knowledge. Indeed, interest in politics is the most powerful predictor of how much you know about political issues, more so even than education or income or race or gender or any other variable that we can put in a regression equation.[16]

There is nothing wrong with being a sports fan or a political fan. I am some of both myself. When I am not speaking at academic conferences, I spend a lot of time watching my favorite Boston sports teams and cheering on the Red Sox, Patriots, Celtics, and Bruins. So I have no problem with sports fans. I don’t necessarily have a problem with political fans, either. But when you are acquiring information for the purpose of enhancing your fan experience, often that objective is very much at odds with the objective of seeking the truth.

Think about the way that sports fans react to new information that reflects badly on their favorite team. For instance, let’s say that a referee makes a call that goes against the team. The fan’s reaction is usually something like, “That was a terrible call! The ref must be blind, or perhaps he is on the take.” On the other hand, if the referee makes a call in favor of the fan’s team: “Well, of course that was a correct call. How could it possibly have gone the other way?”

This is exactly how political fans react to new political information. Numerous studies show that they tend to discount or completely ignore new information that counts against their preexisting political views. On the other hand, if the information counts in their favor, they tend to overvalue its significance.[17]

It’s even the case that the more you are a political fan, the stronger your interest in politics, the more you tend to talk about politics only with other fans of the same side of the political spectrum. If you are a liberal Democrat, you tend to talk about politics only with other liberal Democrats. If you are a conservative Republican, with other Republicans, and so on.[18]

Similarly, people who are strong political fans tend to follow political news only in media that has the same viewpoint.[19] If you are a conservative Republican, you may very well watch Fox News. If you are a liberal Democrat, perhaps you follow NPR or the New York Times.

All of this is highly irrational behavior if the goal of learning about politics is to get at the truth. As John Stuart Mill famously pointed out, if you are a real truth seeker, you should actively seek out viewpoints that are different from your own: “He who knows only his own side of the case knows little of that.”[20] Those information sources are more likely to give you facts and arguments that you haven’t heard of previously. If, however, your goal is not to get at the truth but something else, such as enhancing your political fan experience or having the psychological gratification of having your preexisting views reinforced, then all of this actually is very rational.

Economist Bryan Caplan calls this “rational irrationality.”[21] When the goal of acquiring information is something other than getting at the truth, it is rational to be highly biased in your choice of information sources, and the way that you evaluate the information that you learn. You can thereby achieve the goal for which you sought out the information. But at the same time, you actually diminish your chance of achieving the goal of getting at the truth and becoming a better voter.

The natural reaction of many people to hearing about the problem of political ignorance is that we must be able to solve it by somehow raising people’s political knowledge, perhaps through education. In my book, I discuss at some length why I think this is actually unlikely to happen.[22]

I will not go into this issue in detail here. I will mention, however, that it is striking that educational attainment has risen enormously over the last 50 to 60 years, yet political knowledge levels have stagnated.[23] This suggests, at the very least, that increasing knowledge through education is a much tougher proposition than we might think.

Precisely because it is not rational for people to acquire much political information, they have little incentive to use the education system for that purpose. It is also the case that government itself has little incentive to use education to inform people about politics as opposed to indoctrinating them in the views held either by the majority of the public or by powerful interest groups.[24]

B. Limitations of Information Shortcuts

Some scholars argue that we do not need to worry much about political ignorance because people can use information shortcuts to get around it.[25] They can use small bits of information to substitute for larger bodies of knowledge that they don’t know. There are many different “shortcut” arguments of this kind. Here, I will discuss only one, which has the strongest support, and exemplifies the weaknesses of the others: retrospective voting.[26]

The idea of retrospective voting is this: maybe you do not need to know much about politics or government to be a good voter. You just need to ask yourself: are things going better under the rule of the incumbents, or are they getting worse? As Ronald Reagan famously put it during the 1980 election, “Are you better off than you were four years ago?” If you are better off, then you can reelect the incumbents and reward them so that they will continue their good policies. If, on the other hand, you’re worse off, you can “vote the bastards out.” Then you can vote in a different set of bastards to replace them. The new group of bastards will have incentives to adopt better policies than the old ones because they will know what happened to their predecessors could also happen to them.

This mechanism can be useful in some cases. But it also exemplifies the severe limitations of information shortcuts. One of them is that in order to use the shortcuts effectively, you actually need some preexisting knowledge. You need to know, for example, what issues the incumbents can actually have an effect on and which ones they cannot. If you don’t know that, then you will end up rewarding and punishing the incumbents for things that they didn’t actually cause, and this is actually exactly what happens in most elections most of the time.

The biggest determinate of electoral outcomes usually is the recent condition of the economy.[27] Did it improve or get worse in the six months or year or so right before the election? This occurs despite the fact the political incumbents actually have very little control over short-term economic trends. If they did control them, they would virtually never be defeated for reelection. As a result, voters routinely reward and punish incumbents for economic conditions they had little or no effect on.[28]

And this problem is not limited to just economic voting. It also extends to other issues. For example, in farm states, there is evidence that the incumbent governor and other officials are much less likely to be reelected if there is a drought, even though these officials generally have no real control over droughts.[29]

On the other hand, it is good news for local officials if sports teams have won a championship recently.[30] If one of the local teams has won, that is great news if you are the mayor of the city where this happened, even though usually the mayor has very little effect on the team’s chances of winning. He does not usually play on the team. He doesn’t call plays. He doesn’t coach. Yet he benefits anyway.

In addition, there is a second big problem with information shortcuts. These theories implicitly assume that voters choose shortcuts based on whether the shortcut will be effective at getting at the truth about politics. In reality, they often use them for these purposes of rational irrationality—for enhancing their fan experience or for reinforcing their preexisting views.

For instance, the idea of retrospective voting implicitly assumes voters are using the condition of the world to evaluate the incumbents. If things are getting better, they give the incumbent more credit. But often what happens is actually the opposite: voters’ preexisting conceptions about the incumbents affect their view of the world. When there is a Democrat in the White House, Republican-aligned voters tend to overestimate the rate of inflation and unemployment. Democratic voters have the opposite bias.[31] They think that things are better than they actually are when a Democrat is in the White House but worse than they actually are when it is a Republican.

Ultimately what we have here is a two-level problem of political ignorance. Most of the public is simply rationally ignorant. They do not pay much attention to the issues, and therefore have very little understanding of what they are voting on. There is a minority, the political fans, who know a lot more. But those people tend to be highly biased in the way that they evaluate the information that they learn, so they get much less truth-seeking benefit from it than they otherwise might.

C. Informational Advantages of Foot Voting

What can we do about this problem? Making more of our decisions by voting with our feet and fewer by voting at the ballot box can help. And it can help for the very same reason that it can help us enhance our political freedom. When you vote with your feet, you have much better incentives to become informed about your decisions than when you vote at the ballot box.

If you are like most people, you probably spent more time acquiring relevant information and studying it the last time you bought a car or a TV set than the last time you decided how to vote for President or for any other political office. Is that because your TV is more important than who holds the presidency or perhaps because it deals with more complex issues than those addressed by the White House? Unless you have a very unusual TV set, I doubt it.

The real reason, for most people, is that when they made the decision about the TV set, they knew that it would actually make a difference. Whereas, on the other hand, when you decide to vote for Obama or Romney or the Democrats as opposed to the Republicans, the chance that it will make a real difference is infinitesimally small. Quite understandably, you take that decision much less seriously than the one about the TV set, even though it relates to a far more important outcome in terms of its impact on the world.

This is true not just for your incentive to acquire relevant information but for your incentive to evaluate that information in an unbiased way when you do learn it. Consider the following social norm that I, at least, have run into on occasion—perhaps you have as well: you are not supposed to argue about politics in mixed company.

If you come up to someone at a social event and start explaining to them why their political views are wrong, they probably will not react in a very positive way. Even if you have devastating arguments and evidence that they have never heard of before that completely destroy their position and definitively explain why they were wrong, amazingly enough they probably will not thank you for correcting their mistakes. Indeed, in many cases they will be quite hostile and annoyed. If you want to be a popular and well-liked person who gets invited to all the cool parties, you don’t want to be the kind of person that is constantly criticizing and correcting other people’s political views. Believe me. I know from painful personal experience that is not the kind of reputation you want to have.

On the other hand, if you approach those very same people and give them new information that is relevant to a foot-voting decision, such as what car to buy or what community to live in, in general their reaction will be quite different. They won’t always be grateful. But they usually will take what you are saying seriously and probably will not react with the kind of hostility that political arguments are often met with.

Why the difference? For most people, having their political views criticized is all pain for very little gain. They suffer the pain and discomfort of having their cherished beliefs criticized. That pain may actually be even greater the stronger your arguments are. If you lose an argument, you feel greater pain, because you might be embarrassed for seeming stupid or illogical compared to your critic. On the other hand, the gain from having your political views corrected is usually very small. At best, you might make a better-informed decision with respect to a choice that is unlikely to make any difference anyway.

On the other hand, when you get new information that’s relevant to a foot-voting decision, that’s news you can actually use. You might buy a better TV, you might make a better choice about where you are going to live, you might send your kids to a better school, and so forth. That will lead you to pay closer attention and try harder to control your biases.

You won’t always succeed, of course. I certainly don’t claim that foot-voters are all like Mr. Spock in Star Trek: completely rational and unbiased in their evaluation of any new information they get. But, in general, we are more logical and less biased when we evaluate foot-voting information than when we evaluate political information. There is a lot of evidence that foot voters do a better job of acquiring and using information, even under very difficult conditions, than ballot-box voters do.[32]

A great example from American history is the case of African-Americans in the Jim Crow-era South, a population that was severely oppressed.[33] Most of them had very low levels of education, and southern state governments actually deliberately tried to keep them ignorant of how conditions for African-Americans were relatively less bad in the North and West. Nonetheless, millions of African-Americans did learn the relevant information and did migrate to other parts of the country, thereby significantly improving their situations. Sadly, it fell far short of putting an end to racist oppression. But it did at least manage to make their situation significantly better than it would have been otherwise. It is a dramatic example of people successfully finding relevant information for foot voting even under highly adverse conditions.

Another good example is the movement of immigrants in the 19th and early 20th century from Europe and Asia to the United States. Many of these people were also poorly educated. Many were illiterate. In those days, they did not have the Internet. You couldn’t simply Google “economic conditions in the United States” to find out that opportunities there were better than in your home country. Nonetheless, tens of millions of people did acquire the relevant information and acted accordingly.

Today, of course, the conditions for effective foot voting are far better than they were a century ago. Moving costs are lower because transportation is faster and cheaper than in the past. In addition, people have more education so they can more easily assimilate new information. Their IQ scores are also higher than ever before. They have more raw intelligence. And the Internet and other modern technology make information much easier to find than previously.

Unfortunately, ballot-box voters have taken very little advantage of all these developments. Although information may be easier to get, it is still not rational for most people to spend a lot of time and effort learning it, in order to vote at the ballot box.

On the other hand, when it comes to foot voting, there are much stronger incentives to take advantage of these opportunities. In many ways, therefore, foot voting can be even better today than it was 50 or 100 years ago when the ancestors of most of the people who currently live in the United States took advantage of it and successfully moved to this country.


There are obviously important constraints and limitations of foot voting as well. Nothing is perfect, and certainly foot voting is not. In this section, I will discuss a few of the constraints and limitations most often cited in the literature.

The most obvious constraint is the problem of moving costs.[34] If you are going to vote with your feet by moving to a different state, that can be costly, not just in terms of money you spend to actually move. You might have to give up your job. You might lose ties to family or friends. I know of people who moved, and then their girlfriends or boyfriends ended up breaking up with them as a result. That is a genuine cost, and a real problem that might limit your ability to vote with your feet or may even make it impossible.

But the problem is not as severe as people often think it is. It has not prevented some 43% of Americans from making an interstate move at least once in their lifetime; almost two-thirds of Americans have, at the very least, made a move within a state.[35]

To the extent that moving costs are a genuine problem, it is actually one that justifies greater decentralization than we might want otherwise. Moving from one state to another is cheaper than leaving the nation entirely. And moving from one locality to another within the same state is usually cheaper than moving to a different state.

If the issue in question is decided in the private sector and not through government, often you can vote with your feet without any significant moving costs at all. You can simply live in the same place you lived in before but switch service providers. For example, you can enroll your children in a different private school, without moving to a different community, as is usually the case if you wish to shift to a different government-owned school. Often, the best way to deal with the problem of moving costs is to reduce those costs by decentralizing further.

A second standard argument against foot voting is the problem of the race to the bottom.[36] Perhaps people can vote with their feet. But businesses can do it too. And businesses are what state and local governments really want to attract. They are the ones that pay taxes. Thus, in a world of foot voting, state and local governments might cater to the interests of businesses in ways that are detrimental to the rest of us.

This argument is often advanced in the area of environmental regulation.[37] Businesses want to be able to pollute as much as they want. In order to attract them in, state and local governments will reduce or eliminate environmental regulations. That may be good for the businesses, but it will be terrible for everyone else. Ordinary people will have choices, perhaps. But most of them will be places where the air is unbreathable and the water is undrinkable and things are generally awful. The foot-voting opportunities might then be illusory for everybody but business interests and perhaps the very wealthy.

This race to the bottom argument does apply in certain instances. But, for the most part, I think it is greatly overblown. It ignores several important things. One is that state and local governments do not just want to attract businesses. They want to attract individual taxpayers as well. And taxpayers care about things like environmental quality. They are less likely to want to move to an area where the air is difficult to breathe or the water is undrinkable or the environment is generally terrible. Other things equal, they prefer areas where the environment is in good shape, like here in Missoula, Montana. That is actually attractive to people looking to move, people who provide tax revenue. And state and local governments, when they compete with each other, are sensitive to that consideration.

Even many businesses actually want good environmental conditions because they want to be in places where the sorts of people they wish to hire want to live. If the air is unbreathable or the community is otherwise environmentally degraded, workers may not want to live there. At the very least, businesses will have to pay wage premiums to make up for the fact that the community is undesirable. Other things equal, they would prefer not to have to do that.

Obviously, there are tradeoffs here. Some types of pollution are also environmentally productive. We would not want to reduce pollution to zero. But, in general, under competitive conditions, the state and local governments have to weigh these things in a reasonable way. They are unlikely to simply say “let’s allow as much pollution as possible and not care about the consequences to anything else.” From a competitive point of view, that is not going to be a good strategy for most communities. In some ways it is desirable that there be a diversity of environmental regulation levels, because people vary in the extent to which they are willing to trade income for environmental quality or for other government services. It is reasonable that there be options that people can choose from on that basis. They can thereby exercise their political freedom.

A third often-advanced objection to foot voting is one that has long been a great stain on the reputation of American federalism.[38] That is, the association between federalism and the oppression of unpopular minorities, particularly racial and ethnic minorities. Historically, state and local governments have tended to persecute African-Americans and other minorities. That has understandably led to the emergence of a conventional wisdom that holds that decentralization harms minorities, whereas concentration of power in the hands of the federal government is beneficial to their interests.[39] If we decentralize power and allow people to vote with their feet, maybe that would be good for white males. But it could be terrible for unpopular minority groups because it would just give the state and local governments free rein to abuse them.

There is some validity to this concern. There is no doubt that state and local governments do have a long history of abusing various minority groups, particularly racial minorities, but certainly not exclusively. But the conventional wisdom about the relationship between federalism and minorities also has severe flaws. One thing it tends to ignore is that the federal government also has a long record of abusing minorities.[40] Before the Civil War, the federal government did a lot more to promote slavery than it ever did to constrain it. Between Reconstruction and the Civil Rights Movement, the federal government also tended to do more to promote racial segregation than to restrict it. The one part of the continental United States that was completely under federal control throughout that period was the District of Columbia. And the District of Columbia was just as segregated as the southern states. On the same day that the Supreme Court decidedBrown v. Board of Education,[41] they also decided Bolling v. Sharpe,[42] which dealt with racial segregation in D.C. schools. Washington, D.C. was very similar to the South in that respect.

The point is not simply that the federal government also often had a bad record. At various times in American history, unpopular minorities, including racial minorities, would have been even worse off with unitary control over policy than they were under federalism. Had there been a unitary policy on slavery at the time of the Founding it probably would have meant slavery everywhere, because only one state, Massachusetts, had abolished slavery up to that point.[43]

Similarly, had there been a unitary policy on segregation between the late 19th century and, say, around 1940, it very likely would have looked more like the policy in the South than the policy in the North for various reasons, but you can see some example of what it might have looked like based on what happened in D.C. and also in some federally controlled territories as well.

In addition, during these periods, minorities to some degree were able to benefit from the federal system by having the opportunity to vote with their feet. As I mentioned earlier, African-Americans were able to vote with their feet and leave the South. That, to some extent, helped alleviate their oppression. It certainly did not fully solve their problems. But it made their situation less bad than it would have been otherwise, and less bad than it likely would have been under a unitary national policy on racial issues.

Today, in many ways, minorities have greater leverage and influence over state and local governments than has often been true in past history, as emphasized in important recent work by Heather Gerken of Yale Law School.[44] For instance, one of the main reasons for the impressive success of the movement for gay and lesbian rights over the last twenty to thirty years is that policy on most of the issues of concern to them was de-centralized. They could first make gains in those state and localities that were relatively more favorable to them, and then those gains spread elsewhere. If we had had a unitary national policy on same-sex marriage, we would not today have numerous states that have adopted it. And we probably would not be talking about the possibility of the United States Supreme Court making same-sex marriage a right that applies nationally.

I am not contending that the opposite of the conventional wisdom is true: that de-centralization is always and everywhere good for unpopular minorities. That would be too strong a claim, and cannot be justified by the available evidence. But I am suggesting that the traditional view that de-centralization is bad for minorities is incorrect. We need a more nuanced approach to the issue. At the very least, it is not a valid general argument against enhancing foot voting through increasing political decentralization.


In the last part of this lecture, I would like to briefly summarize the implications of all of this for constitutional federalism and for judicial review. One big implication is that if we want to enhance people’s ability to vote with their feet, so as to increase their political freedom and diminish the problem of political ignorance, we should de-centralize political power to lower levels of government. That way, more issues can be decided through foot voting as opposed to ballot-box voting. That, in turn, strengthens the case for judicial review of federalism issues—for judicial enforcement of limitations on the power of the federal government.[45]

This also has implications for the longstanding debate over the scope of judicial review more generally. Perhaps the most common objection to strong judicial review is the claim that it is anti-democratic because unelected judges are setting aside the will of the people as expressed by their elected representatives. Judicial review is suspect because it undercuts political freedom, as expressed through ballot-box voting. This is the so-called “countermajoritarian difficulty.”[46] But once we recognize that in many ways voting with your feet is a better way of exercising political freedom than voting at the ballot box, then it may be that judicial decisions that impose limitations on the power of the federal government actually enhance our political freedom more than they detract from it.

Political ignorance is also relevant here. It raises serious questions about whether the laws that judges strike down genuinely represent the will of the people. In a world of widespread public ignorance, it turns out that many of the laws the judges might invalidate are ones that the public either does not know about at all or has very little understanding of their effects. Thus, decisions striking down such laws are less countermajoritarian than we usually assume.[47]

On the other hand, judicial decisions that limit the power of the central government enable people to make more decisions by voting with their feet. Those issues can thereby be addressed in a framework where individuals can make more meaningful and better-informed choices.


Foot voting in a system of de-centralized federalism can enhance our political freedom by enabling us to make decisive choices about what laws we wish to live under. It can also help ensure that we make those decisions in a better-informed way.

None of this proves that we should have the maximum possible amount of judicial review. In deciding how much judicial review is optimal, there are many other issues that need to be considered other than political ignorance, the countermajoritarian difficulty, and the advantages of foot voting. It also does not mean that we should have the maximum possible degree of political de-centralization. Many other issues must be considered as we weigh the appropriate size and centralization of government.

Even if you read my book and agree with all of its arguments, you still might not favor as much de-centralization and limitation of government power as I do. But if you take political freedom and political ignorance seriously, you should support more de-centralization and tighter limits on federal government power than you yourself would be in favor of in a world where political ignorance was not a serious problem or one where ballot-box voting was somehow a more meaningful exercise of political freedom than it actually is.

More generally than that, I think foot voting and political freedom are issues that we haven’t given enough consideration to in debates over federalism constitutional theory. In my work over the last several years, I have tried to plug some of the holes in the literature. But there is a lot more work to be done on these vital issues.

* Professor of Law, George Mason University School of Law. This article is adapted from my keynote address at the Montana Law Review’s Browning Symposium on the Future of Federalism at the University of Montana on October 2, 2014.

[1] Ilya Somin, Democracy and Political Ignorance: Why Smaller Government is Smarter (Stanford U. Press 2013).

[2] Ilya Somin, Foot Voting, Federalism, and Political Freedom, in Nomos LV: Federalism and Subsidiarity 83 (James Fleming & Jacob Levy eds., N.Y.U. Press 2014).

[3] Declaration of Independence [¶ 2] (1776).

[4] Andrew Gelman et al., What Is the Probability that Your Vote Will Make a Difference? 50 Econ. Inquiry 321, 323–324 (2012).

[5] For good reviews and critiques of these sorts of arguments, see A. John Simmons, Moral Principles and Political Obligations 91–93 (Princeton U. Press 1979) and Michael Huemer, The Problem of Political Authority 22–35 (Palgrave Macmillan 2013).

[6] For a strong critique of tacit consent theories based on residency, see Simmons, supra n. 5, at 95–100.

[7] Ilya Somin, Annenberg Public Policy Center Survey Provides New Evidence of Widespread Political Ignorance, Washington Post(Sept. 28, 2014) (available at ( The survey itself is available at The Annenberg Public Policy Center, Americans Know Surprisingly Little about Government, Survey Finds, ( (Sept. 17, 2014).

[8] Somin, supra n. 1, at 21.

[9] Id. at 22.

[10] Annenberg Public Policy Center, supra n. 7.

[11] Somin, supra n. 1, at 100–102.

[12] James R. Flynn, Are We Getting Smarter? Rising IQ in the Twenty-First Century 6 (Cambridge U. Press 2012) (noting that average American IQ rose 15 points during the last half of the twentieth century).

[13] See e.g. Mark Bauerlein, The Dumbest Generation (Penguin Group 2008).

[14] The theory was introduced by Anthony Downs, An Economic Theory of Democracy ch. 13 (HarperCollins 1957). For a detailed discussion, see Somin, supra n. 1, at ch. 3.

[15] Somin, supra n. 1, at 78–79.

[16] Id. at 82–84.

[17] See Id. at 79–81 (summarizing the evidence).

[18] Id. at 81.

[19] Id. at 81.

[20] John Stuart Mill, On Liberty 36 (David Spitz ed., Norton 1975).

[21] Bryan Caplan, The Myth of the Rational Voter: Why Democracies Choose Bad Policies ch. 5 (Princeton U. Press 2007); Bryan Caplan, Rational Ignorance vs. Rational Irrationality, 54 Kyklos 3, 3 (2001).

[22] Somin, supra n. 1, at ch. 7.

[23] Id. at 171–173.

[24] Id. at 173–175.

[25] For a survey and critique of shortcut theories, see Id. at ch. 4.

[26] For a well-known defense of the retrospective voting shortcut, see Morris Fiorina, Retrospective Voting in American National Elections (Yale U. Press 1981).

[27] See generally Christopher Achen & Larry Bartels, Musical Chairs: Pocketbook Voting and the Limits of Democratic Accountability (unpublished manuscript, Sept. 8, 2004) (available at (

[28] Id. ; see also Somin, supra n. 1, at 100–101.

[29] Somin, supra n. 1, at 100–101.

[30] Id. at 101.

[31] Id. at 104–105.

[32] Id. at ch. 5.

[33] I discuss this case in some detail in Somin, supra n. 1, at 128–135.

[34] For a more extensive discussion of the problem of moving costs, see Somin, supra n. 1, at 144–145.

[35] Id. at 144.

[36] I address this issue in greater detail in id. at 145–147.

[37] See e.g. Kirsten Engel, State Environmental Standard-Setting: Is There a “Race” and Is It “to the Bottom”? 48 Hastings L.J. 271, 274 (1997).

[38] This brief discussion of the minority oppression objection to political decentralization draws on my more extensive analysis in Somin, supra n. 1, at 147–150; and Somin, supra n. 2, at 98–102.

[39] For a good summary of that conventional wisdom, seeDouglas Laycock, Protecting Liberty in a Federal System: The US Experience, in Patterns of Regionalism and Federalism: Lessons for the UK 121–145 (Jörg Fedtke & Sir Basil S. Markesisinis eds., Hart Publishing 2006).

[40] For a summary of this record and discussion of its implications, see Somin, supra n. 2, at 99–101.

[41] 347 U.S. 483 (1954).

[42] 347 U.S. 497 (1954).

[43] Somin, supra n. 2, at 99.

[44] See e.g. Heather K. Gerken, A New Progressive Federalism, 24 Democracy 37 (Spring 2012) (available at (

[45] For a more detailed discussion, see Somin, supra n. 1, at 165–169.

[46] For a discussion of the countermajoritarian difficulty and its major role in constitutional theory, see Barry Friedman, The Birth of an Academic Obsession: The History of the Countermajoritarian Difficulty, Part Five, 112 Yale L.J. 153 (2002).

[47] See Somin, supra n. 1, at ch. 6; Ilya Somin, Political Ignorance and the Countermajoritarian Difficulty: A New Perspective on the “Central Obsession” of Constitutional Theory, 89 Iowa L. Rev. 1287 (2004).

Montana Law Review Online

          Montana Law Facebook  Montana Law Twitter  Montana Law YouTube

Montana Law Review Online

Go to top