Public Choice

James Buchanan died the other day and in the eyes of many he was the father of Public Choice economics. Comments on a number of blogs reveal two things about bloggers on the left: (1) they don’t understand what contribution Buchanan made and (2) they don’t understand why Public Choice is important. (See a skeptical Yglesias, but surprising praise from DeLong.)

Let me take up the second issue first. The single most important idea in all of Public Choice is the idea of political equilibrium. It is analogous to an equilibrium price or an equilibrium quantity in economics. In the political sphere, the idea is that there is a single platform that can defeat all other platforms in a majority vote. If there is such a platform, then a politician must endorse it or risk losing an election to an opponent. That being the case, there will be a natural tendency for both candidates in two-party elections to gravitate to the same set of policies. Failure to do so will invite defeat.

The idea of an equilibrium platform to which all candidates are inevitably pushed is absolutely devastating to the liberal conception of government. Remember that the liberal case for government intervention in the economy is the idea of market failure. Markets fail, it is alleged, because there is imperfect information, imperfect competition and imperfect other things. Government, in principle, can correct these imperfections. But government cannot correct anything if the decision makers have no choice about the decisions they make. The equilibrium platform, not the platform that corrects markets, is the platform that will ultimately prevail.

[Notice that in groping toward the winning platform, the political system is also characterized by imperfect information, imperfect competition and other imperfections — on a larger scale than what we ordinarily find in the private sector.]

What if there isn’t a winning platform? Then regardless of what a ruler does, a challenger will be able to find an alternative set of policies that will defeat her in the next election. This is the paradox of voting. It implies unstable government; and that is even more devastating to the liberal view of government than the idea of an equilibrium platform.

How do we know that the equilibrium platform and the platform that corrects market failures are not one and the same? Buchanan never adequately answered this question. But I believe I have — both in my dissertation and in my work with Phil Porter.

Consider any political decision and consider the political pressure the decision maker faces. Whether it is a price or a level of output or a decision about spending, a change in the variable will always create benefits for some people and impose costs on others. Here is the condition for equilibrium: the additional political support offered by those who benefit from a higher level of the variable must be equal to the additional loss of support from those who are made worse off.

Marginal political support for a higher level of the variable has two components: the political price people are willing to pay for a dollar of gain times the marginal benefit a higher level would create. On the other side, the marginal political cost of increasing the variable is the political price people are willing to pay to avoid a dollar of loss times the marginal cost an increase in the variable would create for them.

Now here is the bottom line. Optimal government is government that gives us a set of policies for which marginal social benefit equals marginal social cost. But that can happen only if the political prices various groups of people are willing to pay are exactly equal on both sides of every issue.

What is a “political price”? By that I mean the sum total of all the efforts a group is willing to make (per dollar of expected benefit) to support the election of the candidate it prefers. These include votes, campaign contributions, get out the vote efforts, etc. Now it might seem that just about everybody would be willing to make a dollar’s worth of effort to gain a dollar from the political system. However, the election of a candidate or the passage of a law is a public good for all those who prefer it and a public bad for all who are opposed. Every political change, therefore, has a free rider problem.

Since people who benefit from a change will realize benefits whether or not they help bring it about, their incentive is to contribute nothing and become free riders on everyone else’s efforts. In politics, therefore, people inevitably understate their preferences and in most cases they understate them a great deal. That by itself does not create a problem. The problem arises because the degree to which preferences are understated is not the same for every group.

The condition for optimality requires that political prices be the same on both sides of every issue. When that condition does not hold, we get non-optimal policies.

And since that condition will almost never hold in any political system, optimal government is in principle impossible.

If the political price milk producers are willing to pay is greater than the political price offered by the consumers of milk, we will get milk price supports. If the price sugar growers are willing to pay is higher than the one offered by sugar consumers, we will get sugar quotas. We get bad government — or if you like, we get government failure — not because of bad leaders. We get bad government because of inequality in the pressures put on elected officials.

Now we come to the most important contribution of James Buchanan and his colleague Gordon Tullock (who was shamefully passed over by the Nobel Prize committee). Historians and political scientists traditionally have focused on people in leadership positions — on their personalities and even their political parties. It wouldn’t be unusual, for example, to hear one give Franklin Roosevelt credit for our Social Security system. Many economists also tend to view the world in this way. Paul Krugman is an example.

However, by the last decade of the 20th century, 95 % of all the countries in the world had a social security system that looked structurally identical to ours. If you stop and think about that for a moment, it’s hard to conclude that social security owes much to the personality or leadership of any one politician in any particular country. It would appear that forces are at work that are independent of individual leaders. Similarly, the fact that 30 countries have fully or partially privatized their social security systems in the past 20 years suggests that countervailing forces are now at work — again, independent of individual personalities.

I am not sure that either Buchanan or Tullock ever used the phrase “politicians don’t matter.” However, I have no hesitation is using it. In the idealized public choice model, candidates don’t matter. Who is elected doesn’t matter. All that matters are the tradeoffs that the economy permits and the willingness of interest groups to reward politicians for manipulating those tradeoffs.

What are the public policy implications of all this? There is government failure as well as market failure. In deciding how much power to give to politicians, we must compare the likelihood and magnitude of the two. In the last analysis, prudence is going to suggest that we constitutionally limit the types of decisions politicians are allowed to make.

Post script: Tyler Cowen answers Matt Yglesias’ question in more detail.

Comments (14)

Trackback URL | Comments RSS Feed

  1. Greg says:

    Excellent. Thanks for this post.

  2. Ken says:

    I think this is more John Goodman than it is James Buchanan.

  3. Studebaker says:

    One of the great advances in economics over the past few decades has been economic research as an applied math. Because Buchanan was not a mathematician, younger economists were slow to grasp the meaning of his work. Buchanan developed economic theories of public sector spending that seem obvious today. However, when Buchanan began his work, he had far fewer years of observations from which to make inferences. The idea that politicians want to control public funds and dispense them in ways that get the most votes is logical. But, in the days before huge entitlement programs, transfer payments and deficit spending, Buchanan had to predict what would occur under a political environment rife with perverse incentives. I’m surprised people today express doubts about Buchanan’s work given that it’s so easy to verify.

  4. Kyle says:

    I’m not sure constitutional limits mean anything to congress at this point.

  5. Kent Lyon says:

    So, is there a tipping point at which voters who get government benefits outnumber those who pay into government, and does such a tipping point herald the financial collapse of a society? Greece, for example. In this last election, Mitt Romney won a majority of votes from voters in every income category except those making less than $50,000 a year, e.g, those who get most government benefits. Does a steeply progressive income tax and an entitlement state then result in the majority ultimately voting itself benefits from the public treasury until financial collapse occurs?

  6. Grace-Marie Turner says:

    Thank you, John, for this insightful and concise description of Public Choice economics. And I couldn’t agree more with your conclusion: “In the last analysis, prudence is going to suggest that we constitutionally limit the types of decisions politicians are allowed to make.” The private economy will continue to be suffocated as long as politicians seize and regulate a greater and greater share of our GDP.

  7. Robert A. Hall says:

    When I was a young state senator, I discovered that many issues were driven by what I called “Depth of Commitment.” That is, a small committed group could win policies that might not be the “optimal platform” for a majority of voters. For example, the vast majority of voters would be fine with anyone leaving items and information in their mailboxes—but they don’t care enough about it to vote on the issue. The postal workers unions, ever eager to hamper competition from the private sector, cares a great deal, and so it’s is illegal for a local business to hand drop advertising coupons, for example, in your mail box. (They toss them on your lawn instead.) This does not apply to large, public issues like guns, abortion, etc., but to the majority of small interest group issues. I will link to this from my Old Jarhead blog. (

    Robert A. Hall
    USMC 1964-68
    USMCR, 1977-83
    Massachusetts Senate, 1973-83
    Author: The Coming Collapse of the American Republic
    All royalties go to help wounded veterans
    For a free PDF of my 80-page book, write tartanmarine(at)

  8. Ramesh Chandra says:

    John, U amaze me.
    Great Article and Well put together.
    Unfortunately there is so much illiteracy and so little commitment.
    We need to find a way of educating these masses who lost their way. Not in our hollowed and hollow schools, but in public forums like URs.

  9. Neil Caffrey says:

    Great analysis of public choice.

  10. Paul H. says:

    Good analysis. Best I’ve seen on this subject.

  11. John Sweeney says:

    Thanks for publishing. Articles like this are rare and valuable.

  12. bart says:

    The paradox of voting is only an issue when there are three or more candidates, each of whom beats one of the others in a head-to-head contest (think rock-paper-scissors).

    If only we were so lucky as to have that problem. Our first-past-the-post system of voting places any center candidate in a three-way race at such a strong disadvantage that the system can never converge on a single equilibrium. It can only ping-pong between two lesser equilibria to the left and right of center.

    The same goes for systems using runoff or so-called instant runoff elections, all of which are just first-past-the-post in multiple iterations. Since the center candidate in a three-way race tends to be the first eliminated, runoffs don’t help.

    The only workable systems that get past this are the head-to-head or Condorcet systems, which are complicated (and subject to the paradox of voting mentioned above), and a much simpler method known as approval voting.

    The latter can be described as simply allowing the voter to vote ‘Yes’ or ‘No’ on each candidate independently, as though they were separate ballot measures. The candidate with the most ‘Yes’ votes (approvals) is the winner.

  13. Jeffery says:

    I am genuinely grateful to the owner of this site who has shared this
    impressive post at at this place.

    Also visit my homepage Jeffery

  14. Danilo says:

    Nice blog here! Also your web site so much up very fast!
    What host are you the use of? Can I get your affiliate
    link in your host? I desire my site loaded up as quickly as yours lol

    Also visit my web blog: how to prevent premature ejaculation video