Of Parties, Moving Costs, Hedgehogs, and Foxes: Reply to Heather Gerken

I would like to start by thanking Heather Gerken for her thoughtful response to my lead essay and book. Gerken raises two important potential criticisms of my argument that people make better decisions through foot voting than ballot box voting. First, she contends that knowledge of the two major parties’ positions can enable otherwise ignorant voters to make good decisions at the ballot box. Second, building on the important work of my colleague David Schleicher, she worries that foot voting may often be too difficult because of moving costs.

These are legitimate points, and I address both at some length in my book.[1] On balance, however, neither seriously undermines the informational advantages of foot voting over ballot box voting. 


Party labels are often a useful information shortcut. By knowing that Candidate X is a Democrat or Republican, voters can also know a good deal about his or her positions on various issues, even if they know nothing else about that person. Unfortunately, this is not enough for voters to make an informed decision at the ballot box. Even if a voter has excellent information about the policies a given candidate will pursue in office, it is also important to have some idea of the likely effects of those policies. If a given party runs on a platform of  helping American industry by increasing protectionism, or  making the streets safer by ramping up the War on Drugs, is that actually likely to help the economy or reduce crime? Voters ignorant of basic aspects of politics and economics (as most often are) are poorly equipped to answer such questions. For example, polls show that the majority of the public believes that protectionism helps the economy, despite overwhelming evidence to the contrary endorsed by economists across the political spectrum.

Even if the party information shortcut does help voters make good choices between the platforms offered by the Republicans and Democrats, those platforms themselves would likely be different and better if the public were less severely ignorant about policy. Parties choose their platforms at least in large part to maximize their chances of winning elections. Platforms that effectively appeal to a generally ignorant electorate are likely to be different from – and of lower quality – than those that might appeal to a much better-informed one. Scholars such as economist Bryan Caplan and political scientist Scott Althaus find that there are huge differences between the views of well informed and poorly informed voters, even after controlling for a wide range of other variables, such as race, gender, income, and ideology.[2]

In addition, voters often do a poor job of figuring out the positions of the parties themselves, often overlooking even major shifts in their stances. For example, over the last forty years, the public has barely altered at all in its perceptions of the ideologies of the two parties, despite such major developments as the rise of Reagan and the New Right, Bill Clinton’s centrist “New Democrat” platform,  the ascendancy of Barack Obama in the Democratic Party, the growing influence of the Tea Party movement since 2009, and others.

In one important respect parties actually make the problem of political ignorance and irrationality worse rather than better. Many voters develop strong party loyalties that skew their perceptions of reality. For example, Republicans tend to overestimate the rate of inflation and unemployment when there is a Democrat in the White House, and they underestimate it when the president is a Republican. Democrats have the opposite bias.[3] Rather than judge parties by their records, partisans judge records by their feelings about the party in power.  Obviously, swing voters usually don’t have such strong partisan biases. But swing voters are also the least knowledgeable about politics and public policy in general.[4]

Gerken interestingly contrasts my “fox”-like view that informed voting requires knowledge of a range of issues with the “hedgehog” view that all voters need to know is the difference between the two parties. It’s worth noting that Philip Tetlock’s important research on the predictive accuracy of policy experts shows that “fox” experts who take many variables into account make far more accurate judgments than “hedgehogs” who focus only on one or two big ideas.[5]  Voters obviously don’t need to know as much as policy experts. But narrowly focused hedgehog decisionmaking is unlikely to work well even for them. It is especially problematic in a world where government addresses such an enormous range of complex issues. Hedgehog voters and hedgehog policy experts might do better if the functions of government were fewer and simpler.

Foot Voting and Moving Costs

Gerken is absolutely right to point out that foot voting often involves significant moving costs. This is indeed an important constraint. But it is far from prohibitive. In the modern United States, some 43% of Americans have moved between states at least once in their lives, and  63% have at  least made moves within a state (usually switching local governments in the process).[6] Migrations towards states and localities with relatively more effective governments  and away from dysfunctional ones are quite common. The poorly governed city of Detroit’s massive loss of population  over the last several decades is a particularly dramatic example.

In addition, much can be done to reduce those costs further. [7]  By decentralizing power to local as opposed to state governments, we can offer foot voters a wider range of options within a relatively short geographic distance. In this way, people can vote with their feet without having to give up job opportunities, social ties, or the benefits of “agglomeration” stressed in David Schleicher’s work.[8] Some unusually large jurisdictions, such as Los Angeles or New York City, could potentially be broken up into several smaller cities in order to facilitate foot voting by residents and prevent any one local government from gaining monopoly power.  The same point also applies to unusually large states, such as Texas or California, though state boundaries are politically far more difficult to rearrange than local government boundaries. Heather Gerken herself has done excellent work on the benefits of pushing federalism “all the way down” to the local level.[9] My own work reinforces her case.

Moving costs can be reduced still further by leaving more issues to be decided through the private sector rather than the public. In the market and civil society, people can often switch service providers without having to move at all. Over 50 million Americans already live in private planned communities, which provide many services traditionally associated with local governments. Moving between private planned communities is not costless, but it is far easier than moving between cities or states.

We cannot decentralize or privatize all of the functions of government. Some problems are so large that they can only be effectively addressed at the national level – or even the global one. We also cannot completely eliminate moving costs. But many of the world’s best-governed nations – such as Denmark, Switzerland, Luxembourg, and Singapore, are much smaller than many US states, and, in the latter two cases, even many American cities.  This suggests that we can do a great deal of decentralization without losing much in the way of beneficial economies of scale.

My book does not offer a complete theory of federalism, or of the role of government more generally. But the informational advantages of foot voting are a good justification for shrinking and decentralizing the state more than we might otherwise.


[1] See Ilya Somin, Democracy and Political Ignorance: Why Smaller Government is Smarter, (Stanford: Stanford University Press, 2013),  93-97, 144-45.

[2] See Bryan Caplan, The Myth of the Rational Voter: Why Democracies Choose Bad Policies,  (Princeton: Princeton University Press, 2007); Scott Althaus, Collective Preferences in Democratic Politics, (Cambridge: Cambridge University Press, 2003).

[3] For a discussion of these types of partisan biases, see Somin, Democracy and Political Ignorance, 95-97.

[4]Ibid, 111-12.

[5] Philip Tetlock, Expert Political Judgment: How Good is it? How Can We Know?  (Princeton: Princeton University Press, 2005).

[6] Pew Research Center, Who Moves? Who Stays Put? Where’s Home? (Washington, DC: Pew Research Center, 2008), 8, 13.

[7] For a fuller discussion, see Somin, Democracy and Political Ignorance,  137-39, 144-45.

[8] David Schleicher, “The City as a Law and Economic Subject,” University of Illinois Law Review  (2010): 1507-64.

[9] Heather Gerken, “Foreword: Federalism All the Way Down,” Harvard Law Review 124 (2010): 4-74.

Also from this issue

Lead Essay

  • Ilya Somin argues that political ignorance is pervasive and extends to an alarming number of elementary topics. Voters simply don’t have the information that they need to make informed decisions. Increasing education seemingly hasn’t alleviated the problem, and voters tend to choose based not on facts or values, but on the basis of team-like partisan allegiances. This behavior is to some extent rational, because the payoffs are essentially zero for any one voter becoming more informed. It may thus be unreasonable to expect voters to become highly educated on matters of public policy. What we need, he argues, is instead a structural change: Make government a smaller and less consequential part of our everyday lives.

Response Essays

  • Heather Gerken finds much of interest in Ilya Somin’s work on political ignorance, a topic that has only gradually come into its own among law professors despite considerable social scientific research demonstrating political ignorance among real-world voters. She characterizes Somin’s view of political ignorance as “fox-like,” in the sense used by Isaiah Berlin: A fox knows a little about a lot of different things. Like Berlin’s fox, Somin’s voters will be apt to move from one environment to another. But she wonders whether this is a proper description of American voters. She contrasts Somin’s thesis with the work of David Schleicher, who suggests that most voters rely on party affiliation as an information shortcut. Gerken argues that the debate between Somin and Schleicher raises fundamental questions about the future of our democracy.

  • Sean Trende argues that while political elites may find voters startlingly ignorant, the system doesn’t really seem to be broken. It has often functioned quite well despite voters’ ignorance. He suggests that this is so for two reasons: First, much of what passes for expert knowledge is really nothing of the kind, so adding it to voters’ heads does not therefore constitute an improvement. Second, public opinion on many normative matters is ultimately more or less decent, and from it, relatively good policies flow.

  • Jeffrey Friedman argues that rational ignorance theory is false. Voters vote because they mistakenly believe that their votes are likely to matter. They remain ignorant not because they think their votes are irrelevant, but because they mistakenly believe that the world is a very simple place. In such a world, it would take very little knowledge to make an informed decision about politics. Sadly, that’s not the world we live in; in reality, political decisions are complex and difficult. Friedman argues that both genuinely ignorant voters and blinkered ideologues fit this model, and that it conforms better to the available data on voter motivations than does the theory of rational ignorance.