Tiebout: A pure theory of local expenditures

Tiebout. 1956. A pure theory of local expenditures. Journal of Political Economy 64 (October): 416-424.

Y: True expression of citizen preferences, thus optimal public goods provision (thus optimal taxation)

X: Citizens select a local government by moving (voting with your feet)

A response to earlier works by [separately] Samuelson and Musgrave. Musgrave and Samuelson studies whether public goods provision could ever be optimal. They determined that it could, but only if citizens expressed their true preferences when voting. Citizens don't do that, however; they understate their preferences for public goods because they don't want taxes to rise.

Tiebout points out the these analyses implicitly assumed that only central governments provide public goods. He turns to local governments. Basically, this is his model:

Local governments produce a package of public goods. Depending on the particular package offered, there is some optimal community size that can provide that package of goods at lowest cost. If citizens are perfectly mobile (they can move on a whim), some people in oversized cities will leave for undersized cities, recognizing that they can get their desired package of public goods at lower cost elsewhere.

CRITIQUE:

In my view, this argument relies on a set of assumptions that are unrealistically strict (see the 7 assumptions listed on page 419). This leads me to several criticisms:

• First, the assumption that citizens are perfectly mobile is absurd. Tiebout concedes that moving entails costs, but he compares them to the cost of driving to the mall; just as this driving cost doesn't prevent our preferences from creating efficient markets in private goods, moving costs shouldn't prevent our preferences from creating efficient markets in public goods. The analogy doesn't make sense. Moving costs are very high (see my comment on page 422). Tiebout concedes that, as moving [transaction] costs rise, public goods provision becomes less optimal. I think that moving costs are so high that we shouldn't expect Tiebout's argument to work very well at all.
• Second, he assumes that a local government's public goods provision and expenditures are fixed and known. This is not true, for a reason. Since moving costs are high, people lobby their local government for change rather than simply moving to a new community with "fixed" public goods provision. In other words, RECOGNIZING THAT MOVING COSTS ARE HIGH VITIATES THE MODEL. Rather than move to a new community, people will instead lobby for change. And if they lobby for change, we return to the quandary discovered by Musgrave and Samuelson: they will understate their true preferences in an effort to minimize taxes.

STILL:

People do select communities to the degree that they can. So you would expect public goods provision to be more optimal than Samuelson and Musgrave predict, even if it isn't perfectly optimal.