In previous writings, I have argued that “voting with your feet” is often superior to voting at the ballot box because it enables individuals to make a decisive choice, and gives them stronger incentives to seek out relevant information and use it wisely. We should therefore limit and decentralize government power more than at present, so as to expand opportunities for foot voting.

In a recent post, leading federalism scholar Rick Hills argues that foot voting in a federal system can’t work effectively without major help from ballot box voting. He agrees that foot voting has many of the advantages I emphasize, but contends that it might be futile if ballot box voters cannot incentivize politicians to continue good policies:

Unless there is a feedback mechanism inducing local officials to care about attracting migrants, we argue that interjurisdictional migration will be nothing more than a trip from the frying pan to fire. Foot-voters would either over-crowd the “good” jurisdictions or not bother moving at all, because they would have no guarantee that well-governed destinations would stay well-governed. This does not mean that foot-voting is pointless: It just means that migration should be regarded as a complement to ballot-voting, not a substitute. Ilya’s argument for foot-voting, in other words, becomes much more powerful if one recognizes that foot-voting actually improves local ballot-voters’ incentives to cast an intelligent vote. One such incentive is the prospect of foot-voters’ buying their houses. Our criticism is offered, therefore, more as a friendly amendment than a refutation to Ilya’s outstanding book.

I greatly appreciate Rick’s praise of my book, and I certainly agree that we cannot dispense with ballot box voting entirely. Well-informed ballot box voting can make a federal system function much better than it would otherwise. Rick outlines several plausible mechanisms by which foot voting can reinforce ballot-box voting, and vice versa. But effective foot voting is not as dependent on ballot box voting as Rick suggests.

I. The Danger of Killing the Goose that Attracted the Golden Egg.

It is indeed true that foot voting does not provide guarantees against the reversal of the policies that attracted the foot voters in the first place. What a state or local government gives today it could potentially take away tomorrow. Government officials with poor incentives might end up killing the goose that attracted the golden egg of foot-voting migrants in the first place.

But the risk is not as great as Rick suggests. Even if the governments in question are just selecting policies at random and do not care about their effect on foot voters, there is still a great deal of inertia in government, especially when it comes to changing long-established policy. Interest groups, bureaucracies, and others who have become habituated to the status quo will not easily accept change. If a policy has been in place for a long time, that is some evidence that it is unlikely to be radically altered in the short to medium term future.

Moreover, at least in a system where subnational governmments must raise a large proportion of their own funding, they do in fact have strong incentives to adopt policies that attract foot voting taxpayers and businesses. Otherwise, they risk losing valuable tax revenue that can be spent on projects they value, or on efforts to help keep them in power. Additional tax revenue is valuable to officials even if voters remain poorly informed and do not directly reward them for it. If nothing else, it can be useful in enriching the officials themselves and their political allies. Thus, competition can incentivize local and regional governments to cater to the needs of foot voters, even in the absence of electoral pressure to do so.

II. Crowding and Capitalization.

As Rick indicates, “crowding” might at some point limit the ability of foot voters to enter attractive jurisdictions. However, such effects need not be a major problem in a system where there are hundreds or even thousands of jurisdictions competing with each other. While some may become crowded, others will not. Moreover, crowding is not a fixed fact of nature, but often the effect of policy. It can be mitigated by changing such policies as excessively restrictive zoning, which artificially limits the amount of housing available to potential migrants.

The danger of artificially inflating land prices through zoning makes me skeptical of Rick’s and Qiao Shitong’s proposal to incentivize good policy by rewarding local officials for increasing the relative value of the land within their jurisdiction. As Rick himself has emphasized in other work, high prices often reflect exclusionary zoning rather than good policies.

Like some other scholars, Rick worries that foot voting may become pointless if the benefits of good government become “capitalized” into the value of land: housing in well-governed jurisdictions becomes more expensive to the point that there is no benefit to moving there from poorly governed ones. In my book Democracy and Political Ignorance, I explain why this concern is overstated.

Capitalization does not happen nearly as quickly as critics fear. If it did, we would not have had a long history of migration from poorly governed areas to relatively better ones. If capitalization were all that it is sometimes cracked up to be, we would never have had the recent migration of northerners to the Sun Belt or the Great Migration of African-Americans out of the Jim Crow South. The relative advantages of these destination areas should have immediately been reflected in higher prices that would deter migrants in the first place. In addition, as noted above, capitalization effects can be diminished by adopting policies that make housing more available and thereby reduce the costs of migration.

Finally, in a world where instantaneous capitalization does occur, we might indeed see little or no migration driven by foot voting. But the mere possibility of it is still likely to incentivize local governments to adopt better policies As I put it in my book:

If local governments get a substantial proportion of their revenue from property taxes—as most in the United States do—they would have incentives to adopt policies that increase the value of land within their jurisdiction so that they could generate more revenue without raising tax rates. In such a scenario, the possibility of in-migration would serve to bid up prices, even if (by assumption) the changes in price occur so quickly that migrants are deterred from ever actually coming. Indeed, the faster the increases in price happen, the faster local governments will be able to reap the benefits of better policies in the form of higher property tax revenue.

I certainly do not contend that foot voting can solve all our problems, or that ballot box voting has no useful role. To the contrary, I have argued that foot voting cannot effectively address such issues as overexploitation of immobile assets by subnational governments, most notably property rights in land. For that reason, a system of decentralized foot voting needs to be supplemented with constitutional protections for property rights, for other immobile assets, and for people with unusually high moving costs. Also, I recognize that there are some problems so large that they cannot realistically be dealt with at the regional or local level, and therefore must be addressed by national or even international ballot box voting.

Still, even without the help of ballot box voting, foot voting can accomplish more good than Rick suggests. A full response to Rick’s many insights on these issues would take more than a blog post. I hope to say more about these issues in a more comprehensive book on foot voting I am currently working on, tentatively entitled “Foot Voting and Political Freedom.”