Voting with Our Feet

Ilya Somin

Fall 2021

America today faces three serious, interconnected problems. The first is the powerlessness of individual citizens in determining which policies they wish to live under, particularly in national elections. Second, over the last several decades, opportunities for the poor and lower-middle class have become increasingly constricted. Finally, growing partisan bias and hostility have resulted in a situation where people on the right and left not only oppose each other on policy, but view the other side as a menace to the republic.

There is no single solution to these problems, of course. But all three can be substantially mitigated by empowering Americans to "vote with their feet" to a greater extent than is now the case.

People vote with their feet in three major ways: through international migration (not considered in this essay), by choosing which jurisdiction to live in, and by making decisions about which institutions to participate in, such as schools and planned communities. These types of foot voting are often considered in isolation, but they have much in common — including the fact that they represent mechanisms for exercising political choice.

If we want to augment political freedom, increase opportunity for the poor and the disadvantaged, and reduce tensions caused by polarization, expanding opportunities for Americans to vote with their feet can be of great help.


Most people believe ballot-box voting is the ultimate expression of political freedom. It is how we exercise the power to decide what government policies we will live under. But ballot-box voting has two serious weaknesses: Individual voters have almost no chance of affecting the outcome of an election, and for that reason, they have little incentive to make well-informed decisions.

The odds that an individual vote will make a material difference in an election are miniscule: about one in 60 million in a presidential election, to take one example. A person's chances are somewhat higher if he lives in a swing state, but lower if he does not. The odds are also better in elections for state and local offices. Yet in all but the tiniest of jurisdictions, the odds against casting a decisive vote are still many thousands, or even millions, to one.

Meaningful freedom requires the ability to make a decisive choice. A person does not have real religious liberty, for instance, if he has a one-in-60-million chance of being able to determine which religion to practice. Similarly, a one-in-60-million chance of deciding which views one is allowed to express in public is not meaningful freedom of speech. Even a one-in-100,000 chance (the odds of casting a decisive vote in some smaller elections) is not enough to provide anything like genuine choice.

What is true of freedom of religion and speech also applies to political freedom: A person with an infinitesimal chance of affecting what kind of government policies he is subject to has little, if any, genuine political freedom.

Of course, ballot-box voting is not the only way citizens can influence government policy in a democratic political system; they can also do so by engaging in political speech, activism, fundraising, demonstrations, lobbying, and other such activities. In some cases, then, political participation beyond the ballot box enables citizens to increase their influence over public policy. Moreover, prominent activists, intellectuals, campaign donors, and the like can have an even greater influence on policy than the average voter through these mechanisms.

But opportunities for such "beyond voting" forms of participation are highly unequal. Only an estimated 25% of Americans engage in them at all, and those who do so effectively enough to have more than a minute influence on election and policy outcomes are a much smaller group. Even if access to these forms of participation could be equalized, we would still be left with the reality that each individual citizen would have only a miniscule chance of influencing results. After all, if each citizen's participation beyond voting had equal weight, individual participants would have no better odds of changing the outcome by such methods than they would by voting. In both cases, then, increasing the influence of some necessarily means diminishing that of others. Participation beyond voting thus does not, and cannot, overcome the difficulties inherent to the insignificance of any single vote.

The near-powerlessness of individual voters incentivizes them to make little or no effort to become informed about political issues. Surveys consistently show that many voters are ignorant of even basic aspects of the political system and public policy. Only about a quarter to a third of Americans, for example, can name the three branches of government. This is not because voters are stupid, but because they are "rationally ignorant," as economists put it. For most people, there is little payoff to becoming well-informed about government and public policy. For that reason, most Americans quite reasonably prefer to devote their time and attention to other activities.

Perhaps even more troubling is the fact that voters also have incentives to be "rationally irrational" — to do a poor job of evaluating the political information they do have. If a vote has a tiny chance of making a difference, there is little incentive for people to control their cognitive and partisan biases when assessing candidates and policies. This helps explain why studies regularly show voters are highly biased in evaluating political information, often to the extent of believing ridiculous conspiracy theories that conform to their biases. On the right, a troubling percentage of Republican voters believe that the 2020 election was "stolen" from Donald Trump through voter fraud. Left-of-center voters are prone to espousing conspiratorial beliefs of their own, including even such extreme theories as 9/11 "trutherism" — the belief that the Bush administration knew about the September 11th attacks in advance and deliberately allowed them to happen.

Things are very different when people "vote with their feet." When individuals and families decide which jurisdiction to live in, for instance, they have real control over the outcome. This in turn creates strong incentives for them to seek out information relevant to their decision.

The same rule applies to choices about private-sector purchases. Most people devote more time and effort to deciding which television set to buy than to deciding whom to vote for in any election. The reason is not that the television is more important than who governs the country, but that the decision about the former has a real impact on the outcome: The television a consumer chooses will be the one he gets to use. His vote for president, on the other hand, will have a minimal effect on who wins the election. This intuition is backed by both historical evidence and laboratory experiments indicating that people faced with foot-voting decisions seek out greater information, and evaluate it more rationally, than do those assessing political information for use at the ballot box.

While foot-voting opportunities are not fully equal, they can be made available to a wide range of people without undermining their effectiveness. And unlike ballot-box voting, foot voting is not a zero-sum game: My ability to vote with my feet does not diminish yours, and vice versa. If we want to give people meaningful political choice and incentivize them to make informed decisions, we have to give them more opportunities to vote with their feet.


Foot voting is crucial not only to political choice, but to ensuring opportunity — especially for the poor, the disadvantaged, and the oppressed. Throughout American history, it has provided a vital outlet for these populations. Perhaps the best-known example is the movement of millions of African Americans from the Jim Crow South to the North and West, where they benefited from greater opportunity and relatively low levels of racial discrimination. Another is the 19th-century migration west, which not only offered disadvantaged groups greater economic opportunity, but also greater freedom. Several newly formed western states were pioneers in granting equal legal rights to women, including allowing them to vote and to enter professions they were barred from elsewhere. In more recent times, foot voting has been a boon to gays and lesbians seeking to live in more tolerant states and localities.

The history of immigration is also a dramatic example of how foot voting is a massive boon for the underprivileged. The vast majority of Americans today are either immigrants themselves or descendants of those who fled poverty and oppression elsewhere. Despite its flaws, the United States was far preferable to the regimes they left behind — as their immigration to the country illustrates.

This history undercuts oft-heard claims that foot voting is of use primarily to the affluent. To the contrary, its greatest beneficiaries are at the opposite end of the socio-economic spectrum. The latter have far more to gain from moving to locations with a better quality of government. They are also far less likely than the wealthy to own valuable, immobile assets, such as property in land, that they cannot take with them if they choose to move.

Foot voting under federalism can work even more effectively when state and local governments have incentives to compete for residents by offering lower taxes, cheaper housing, and better public services. We can incentivize this competition by limiting federal subsidies to states and localities such that they are forced to rely as much as possible on revenue raised from their own taxpayers. In turn, this result would create stronger incentives for state and local governments to make policy decisions that attract new taxpayers and persuade current ones to stay put.

Foot voting within a federal system can still have great value even if there is little or no competition. If variation in policy is driven purely by the preferences of current residents, or even just by random chance, it can still offer a wide range of options for would-be foot voters, thereby expanding freedom of choice. But incentivizing competition can increase both the quantity and quality of the options available.

Foot voting in the private sector is a less familiar idea than foot voting in federal systems. Nonetheless, it is an important phenomenon. Private organizations of various types offer a variety of services traditionally associated with regional and local governments, the most prominent of which include private communities like condominiums and homeowners' associations. These organizations provide residents with services like garbage disposal, education, security, and environmental amenities. As of late 2019, almost 74 million Americans lived in such communities — a figure that gives lie to claims that private communities are just a tool for the wealthy to wall themselves off from the rest of society.

As a source of foot-voting opportunities, private communities have important advantages over traditional state and local governments. One such advantage is lower movement costs. A given area can fit many more private communities than political jurisdictions. As a result, it is often possible for people to move from one private community to another without giving up employment, family connections, or other benefits and opportunities. Another asset private communities offer is that the services they provide are often of better quality than those offered by the state.

While private communities are far from being the exclusive preserve of the wealthy, it remains true that they are much less available to the poor than to the middle and upper classes. Much can be done to make this form of social organization more available to Americans of all income levels. Examples include loosening zoning restrictions, which would make it easier to establish new communities, and ending the "double taxation" of private-community residents, which occurs when the latter are forced to pay for public services of the same type as those they consume (and pay for) in their private community.

Private communities are far from the only form of private-sector foot voting; there are many other ways the private sector can compete to provide services traditionally associated with state and local governments. One prominent example is in school choice. Many studies show that private schools provide better educational services than their public equivalents, even when controlling for students' socio-economic backgrounds and similar variables. Competition from private schools under voucher programs also leads public schools in the same areas to improve their services.

The advantages of foot voting in education were dramatically illustrated during the Covid-19 crisis. During the 2020-2021 academic year, numerous public schools shut down long past the point where evidence showed they did not pose a significant risk of spreading the disease — often at the behest of politically powerful teachers' unions. Most private schools, on the other hand, remained open and continued to serve students with little or no added Covid-19 spread. The obvious difference is one of incentives: Private-school administrators and teachers are only paid if they provide useful services to students and their families, while their public-school counterparts can subsist on taxpayer dollars even if they provide nothing but grossly inadequate "virtual" education — which has been especially harmful for poor and minority students.

Private-sector foot voting cannot fully replace government services. But much can be done to expand its scope. For instance, policymakers can make it easier to establish private communities and expand school-choice programs. The same goes for promoting other private alternatives to government services.

It may seem as if foot voting is unlikely to remain important in a world where people can increasingly work remotely. During the pandemic, Americans witnessed how working remotely is more feasible than ever before. Nowadays a lawyer, an accountant, or an IT specialist can work for a New York firm even as he lives in Kansas or Montana — at least theoretically.

But this should not blind us to the reality that the vast majority of people — including those who work in service industries, medicine, manufacturing, and agriculture — still need to live near where they work. Even many high-skill white-collar workers are more productive when they can interact with colleagues and clients in person. A recent McKinsey study concluded that only about 22% of American workers can work remotely three to five days per week without loss of productivity. Even those who need to work from their office only once or twice per week still need to live reasonably close to their employer in order to do their jobs effectively.

For these and other reasons, there remain enormous "place premiums" that enable workers to be vastly more productive in some locations than in others. This makes it imperative to break down barriers to mobility that block workers from moving to those locations. The inability to do so not only impoverishes the workers themselves, it also reduces the growth of the economy as a whole.

Even if mobility becomes less important for employment purposes, it can still be a valuable tool for foot voting. To the extent that work becomes a less significant determinant of locational choice, it could open up opportunities for people to vote with their feet based on other aspects of public policy, such as education, law enforcement, and environmental quality.

At the same time, it would be a mistake to conclude that movement based on job opportunities and other economic considerations is not a form of political choice in itself. To the contrary, economic conditions and job opportunities are often heavily influenced by government policies on taxation, employment, and other issues. Thus, even if people vote with their feet based purely on economic considerations, it often represents a political choice as well.


Over the last several decades, foot voting in America has become more difficult and costly — especially for the economically and otherwise disadvantaged, who have the most to gain from expanding opportunities. Fortunately, much can be done to alleviate these obstacles.

The single greatest impediment to foot voting is the rise of exclusionary zoning, which makes it difficult or impossible to build new housing in response to demand. If people cannot afford to live in areas with economic and social opportunities, they will remain locked out of them — and often trapped in failing communities where it is difficult to escape from poverty.

Here, there is a strong (though often unrecognized) common interest between members of the increasingly Republican white working class and their overwhelmingly Democratic African American and Latino counterparts. After all, poor and lower-middle-class workers of all races have much to gain from being able to move to areas where they can be more productive and earn higher incomes.

Economists and land-use scholars across the political spectrum have long known that restrictive zoning cuts off millions of people from housing and job opportunities. Sadly, zoning restrictions are particularly severe in many left-leaning jurisdictions on the east and west coasts. As economists and housing-policy experts on the left are increasingly recognizing, these constraints are at odds with progressives' emphasis on alleviating poverty and improving social mobility.

New evidence suggests that the harm caused by exclusionary zoning is even greater than previously thought. It turns out that economists Chang-Tai Hsieh and Enrico Moretti, authors of some of the most influential work documenting the harm caused by zoning restrictions, greatly underestimated those effects. My George Mason University colleague, economist Bryan Caplan, discovered some significant calculation errors in their path-breaking 2019 article "Housing Constraints and Spatial Misallocation." Hsieh and Moretti have graciously acknowledged these mistakes.

When the calculations are corrected, it turns out that zoning restrictions reduce productivity and economic growth several times more than previously thought. If zoning restrictions in New York, San Francisco, and San Jose (three of the most restrictive major cities) had been in line with the national average between 1964 and 2009, U.S. GDP by the end of that period would have been some 14-36% higher than it actually was. The gains would have disproportionately gone to the poor and the lower-middle class — the very groups most likely to be priced out of housing markets. Extending this analysis to other areas with exclusionary zoning regulations only increases the magnitude of the potential gains from zoning reform.

Although not as severe as the impact of zoning, state-by-state occupational licensing also greatly reduces mobility and thereby undercuts foot voting. A 2017 National Bureau of Economic Research working paper found that people in occupations with state-level licensing are 36% less mobile between states than those in other fields. Licensing has expanded to the point where some 30% of American workers need licenses to do their jobs; in some states, even florists, interior designers, and tour guides have to have a license. The evidence suggests that most of these laws do far more to suppress competition than to protect consumers.

During the 2020 campaign, no less a figure than future president Joe Biden highlighted the harm overzealous occupational licensing causes. As he put it in a speech to a union audience, "why should someone who braids hair have to get 600 hours of training? It makes no sense....They're making it harder and harder in a whole range of professions all to keep competition down." Though it didn't use the term, his campaign website added that occupational licensing reduces foot voting:
"If licensed workers choose to move to new states for higher-paying jobs, they often have to get certified all over again."

Biden was right. In recent years, such states as Arizona, Ohio, Florida, and Pennsylvania have recognized these problems and enacted valuable licensing-reform measures. But much remains to be done on that front.

There has also been important progress on reducing zoning in several parts of the country. The Biden administration included some useful incentives for state and local governments to cut back on zoning in its initial infrastructure proposal, though it remains unclear if those will feature in the final bill. Here, too, much more needs to be done.

Another major obstacle to expanding foot-voting opportunities is the increasing centralization of power in Washington, D.C. Federal spending accounted for a whopping 31% of U.S. GDP in fiscal year 2020. Though this figure was inflated by relief spending related to the Covid-19 pandemic, federal expenditures were already high in 2019 — at 21% of U.S. GDP.

What's more, federal regulation now covers almost every major aspect of the economy and society, from health care and education to pensions and even toilet flows. People cannot vote with their feet against one-size-fits-all federal policies unless they leave the country entirely — which is usually far costlier than moving from one state or locality to another. By leaving more issues to be handled by states, localities, and the private sector, we could expand foot-voting opportunities and thereby empower more people to make meaningful, better-informed choices.

Not every issue can be decentralized, of course; some are so large-scale that only the federal government can address them effectively. National defense and climate change are obvious examples. Yet there remain many opportunities to decentralize without losing such economies of scale. If small nations like Denmark, Switzerland, and New Zealand (all generally considered to be among the best-governed in the world) can have their own independent health care, education, and retirement-security policies, the same is true of the many U.S. states (and even some cities) with comparable or larger populations.

In sum, by decentralizing political power and breaking down barriers to freedom of movement — most notably zoning and occupational licensing — we can expand Americans' freedom of choice and give them more control over their lives.

It may seem contradictory to simultaneously push for reducing federal power for purposes of decentralization while also seeking to curb state and local government authority over zoning and licensing. But any seeming contradiction is more apparent than real. Reducing federal spending and regulation decentralizes power by leaving more decisions to lower levels of government and the private sector. Cutting back on zoning regulations and occupational licensing at the state and local levels also decentralizes power — in this case, by leaving more decisions to private individuals, firms, and civil-society institutions. Likewise, allowing property owners and employers to decide how to use their land and what qualifications they want their employees to have represents a greater degree of decentralization than leaving these matters under the control of states and localities.

State and local authority — like federal power — is not an end in itself. Rather, the purpose of all three is to promote the liberty and welfare of the people — to "promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity," as the Constitution puts it. To the extent foot-voting opportunities are a major element of that liberty and welfare, it makes sense to curtail those powers of government that stand in the way.


Much has been written about the dangerous growth of polarization and partisan hostility in recent American history. The 2020 election and its aftermath have highlighted the depth of that hostility. Many Democrats and Republicans now regard the opposing party as not merely wrong, but as a threat to the entire American way of life.

Back in 2016, columnist Megan McArdle argued that the increasingly hostile relationship between red and blue America had come to resemble a failing marriage. More recently, conservative political commentator David French warned that this marriage may be headed for a contentious divorce — in the form of secession or civil conflict.

This sorry state of affairs has multiple causes. But one of the most significant is the enormous growth of federal power, which has increased the number of issues subject to one-size-fits-all, centralized policies. In an age where the two major parties stand far apart on ideology, many find it increasingly unacceptable to let that power fall into the hands of their political rivals. Such fears are part of the reason why alarmingly high percentages of both Republicans (44%) and Democrats (41%) told pollsters that violence would be at least "a little" justified if their side lost the 2020 presidential election.

Greater devolution of power to states, localities, and the private sector would not put an end to such fears. But it could significantly reduce them. Decentralization, combined with expanded foot-voting opportunities, could enable more Americans to live under their preferred policies, regardless of who sits in the White House or which party controls Congress.

Some argue that the way to defuse polarization is to promote the values of compromise and conciliation. But compromise is easier between people who do not disagree on so many fundamental issues to begin with. Building on McArdle's marriage analogy, one of the most important predictors of a successful marriage is spouses having similar goals and values. Unfortunately, the "marriage" between red and blue America is one between groups with widely divergent values and preferences. In such situations, it is far easier to make concessions when each side knows it will not have to make them too often. The more issues the parties have to decide together, the greater the potential for conflict.

Fortunately, a political union need not be as close a relationship as a marriage usually is. Red and blue America may not be able to spend some time apart, but they don't have to do so many things together. Greater decentralization and expanded foot voting would enable this state of affairs.

Some fear that such measures might aggravate political conflict by exacerbating the "Big Sort" — the supposed tendency of people to cluster in communities of politically like-minded individuals. In his influential 2008 book of the same name, journalist Bill Bishop argued that this trend increases political polarization and our already strong tendency to ignore or dismiss opposing points of view.

This issue cannot be lightly dismissed, especially in an age of rampant partisan bias and hostility. Nonetheless, the concern is overblown. One problem with the Big Sort theory is that the data don't seem to support the notion that we are more ideologically segregated than we were several decades ago. This doesn't rule out the possibility that increased foot voting could cause more of a big sort. But it should make us skeptical of the theory overall.

A deeper problem with the Big Sort hypothesis is that people who vote with their feet tend not to make choices that align neatly on a left-right or red-blue spectrum. Instead, the data suggest that many people gravitate toward jurisdictions that are both socially tolerant and have relatively free-market policies that lead to cheaper housing and more job opportunities. Many are also willing to move to jurisdictions that do not match their partisan biases. In recent years, red states like Texas have attracted many migrants from blue states on the east and west coasts, since the red states' less-restrictive zoning rules lead to lower housing costs.

More generally, there is considerable variety in the amount of diversity (of many types) that people prefer to have in their communities. That, too, cuts against any claim that Americans have a monolithic tendency to move to areas populated largely by those who share their political views. In fact, many Americans today prefer to live in ethnically diverse communities, as they are often economically dynamic and offer a wide range of cultural opportunities. And in recent years, traditional wariness of racial and ethnic integration among many whites has declined. Since communities with greater ethnic, racial, and religious diversity are less likely to be monolithically red or blue, these trends further reduce the likelihood that foot voting will result in dangerously high levels of homogeneity.

As already noted, expanding opportunities for foot voting enables people to make more decisions in a setting where they have stronger incentives to seek out relevant knowledge and curb cognitive biases. People are less likely to act as knee-jerk partisans when they decide where to live or which school to send their children to than when they vote at the ballot box. The more important choices we can make through foot voting, the less impetus there is for partisan bias and hostility.


Expanding opportunities for foot voting is not the only factor that must be considered in determining the size, scope, and concentration of government power. But it is a crucial objective that is all too often ignored in debates about the role of government in our society.

By empowering more people to vote with their feet, we can expand political freedom, increase opportunities for the underprivileged, and help alleviate political polarization. That's not a bad start to building America back better — and making it great again. Indeed, foot voting was a major part of what made America great in the first place.

ILYA SOMIN is professor of law at George Mason University and author of Free to Move: Foot Voting, Migration, and Political Freedom, from which this article is in part adapted.


from the


A weekly newsletter with free essays from past issues of National Affairs and The Public Interest that shed light on the week's pressing issues.


to your National Affairs subscriber account.

Already a subscriber? Activate your account.


Unlimited access to intelligent essays on the nation’s affairs.

Subscribe to National Affairs.