Lecture Notes3

ECO 4554 Economics of State and Local Government Lecture Notes PUBLIC CHOICE THROUGH MOBILITY: THE TIEBOUT HYPOTHESIS Key Points 1.

Ceteris paribus, the quantity of public services or the level of public expenditure is more efficient in a relatively homogeneous community where residents have similar preferences for public services than in heterogeneous communities where residents have dissimilar preferences.

2.

Fiscal zoning improves efficiency in the supply of public services by reducing the potential for some residents of a community to free ride.

3.

The quantity of public services or the level of public expenditure can be efficient even in a heterogeneous community if differences among communities in benefits from public services and taxes are capitalized into the price of housing.

4.

Despite the absence of realism of its assumptions and the inconsistency of its implications with the real world, the Tiebout hypothesis appears to explain many features of real world economic behavior and is relevant for evaluating public policies. Synopsis

We have shown that the quantity of a public good supplied in the private sector is inefficient (less than the quantity at which MSB=MSC) whether a price is charged and non-payers are excluded (the market model) or no price is charged and no one is excluded (the voluntary contributions model). We have also shown that the quantity of a public good supplied by government (the government model) is inefficient because voting does not adequately reveal individual’s marginal benefits and because taxes create inefficiencies in the supply of the taxed private goods and activities. We begin by summarizing this argument. We then introduce the Tiebout hypothesis, which states that, if there are enough communities and if taxes in each community are based only on individual marginal benefits from public services (that is, each community imposes perfect benefit taxes), people will choose among communities based on their preferences and their willingness to pay for public services. In the Tiebout model, each community consists only of people with the same preferences for public goods or services (homogeneous communities). In each community, all individuals pay the same tax, and the quantity of public goods and services is efficient for that community. Two key features of the Tiebout hypothesis are inconsistent with what we observe in the real world, however. Most communities use non-benefit taxes, such as a property tax, rather than benefit taxes, and most communities are not perfectly homogeneous. We show next that even with non-benefit taxes and heterogeneous communities, the Tiebout hypothesis may still be valid. The problem with non-benefit taxes is that the Tiebout model equilibrium is unstable. Communities do not remain homogeneous, and when they become heterogeneous the quantity of public services is no longer be efficient. But, when non-benefit taxes are accompanied by fiscal zoning, the stability and efficiency of the Tiebout model equilibrium can still be achieved.

1

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

Even when we allow for fiscal zoning, however, Tiebout model communities are still homogeneous, unlike communities in the real world. But, if differences among communities in benefits from public services and taxes are capitalized into housing prices, the quantity of public services in each community can still be efficient. With capitalization of fiscal differentials, we can still have a stable equilibrium even though some communities are heterogeneous. How relevant is the Tiebout hypothesis? In the real world, we observe non-benefit taxes. We observe limits on the use of fiscal zoning. Most communities are heterogeneous, not homogeneous. Fiscal differentials among communities are not perfectly capitalized. Therefore, the quantity of public services in each community in the real world cannot be efficient. Does this mean the Tiebout hypothesis is just an abstract irrelevant theory? Or, does the Tiebout hypothesis have enough similarities to the real world that we can still use it to evaluate the impact of public policies on efficiency in the supply of local public services? We conclude by examining the case in favor of the Tiebout hypothesis. Lecture Notes I.

Inefficiency in the supply of public goods: Review and summary A.

A public good has joint or shared or collective consumption properties regardless of whether it is provided in the private market, by clubs, by non-profit organizations, or by government.

B.

The private sector typically supplies less than the efficient quantity of a public good either because of exclusion or because of free riders.

C.

1.

If non-payers are excluded (the market model), people whose MB

2.

If non-payers are not excluded (the voluntary contributions model), individuals have incentives to free ride and the voluntary payments are insufficient to finance the efficient quantity of the good.

Government supply of a public good (the government model) is typically inefficient because of the inefficiencies of voting, the inefficiencies of taxes, and the incentives toward overconsumption and crowding. 1.

Voting (and other political activity) is an imperfect method of obtaining information on people’s preferences or marginal benefits for public goods and doesn’t necessarily result in choice of the efficient quantity.

2.

Most government revenue to pay for public goods is collected by taxes that are not directly related to marginal benefits. a.

If the distribution of taxes is not closely correlated with the distribution of benefits, the supply of the good is likely to be inefficient.

2

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

b.

3.

D.

II.

Taxes create inefficiencies in the markets for those goods and activities that are taxed. Even if government increases efficiency in the supply of the public good, there is a trade-off between greater efficiency in the supply of the public good and less efficiency in the supply of the taxed private goods or activities.

Open access provision of a public good by government (that is, no direct charge to consumers or users) may result in overconsumption of the public good or crowding of a public facility.

Conclusion: The quantity supplied of a public good is inefficient whether the good is provided in the private sector or by government. 1.

The choice is not between inefficient private provision and efficient government provision. The choice is between alternative models, none of which is likely to supply the efficient quantity of a public good

2.

The problem is to weigh the inefficiency of private provision against the inefficiency of government provision and choose on a case-by-case basis the model we think is least inefficient.

Tiebout model (“Voting with one’s feet”) A.

While the quantity of a public good provided at the national level will typically be inefficient whether the good is supplied privately or by government, in the Tiebout model, the quantity of local public services may be efficient.

B.

Statement of the Tiebout hypothesis: Individuals choose the one community from among all communities offering the package of public services and taxes that comes closest to satisfying their preferences. 1.

If there are as many communities as there are individuals with different preferences for public services and if the supply of public services is characterized by constant returns to scale (meaning constant long-run average cost), then each individual can find a community with a package of local public services that exactly satisfies her/his preferences.

2.

Each community will be homogeneous; that is, all residents of each community will have the same preferences for public services. Therefore, the community will provide a package of services at tax-prices that just exactly satisfy the preferences of its residents.

3.

Lindahl equilibrium in the Tiebout model a. In the Tiebout model, each community is homogeneous; all individuals in a community have the same preferences and pay the same tax per unit of public services. b. These tax-prices are exactly equal to marginal benefit. They are Lindahl taxes. Therefore, the equilibrium in each community in the Tiebout model is a Lindahl equilibrium. 3

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

c. In a Lindahl equilibrium, the quantity of the good is the efficient quantity. Therefore, in the Tiebout model the quantity of local public services in each community is efficient. C.

D.

III.

Assumptions of the Tiebout model 1.

Consumers are mobile and will move their residence to the communities that best satisfy their preferences.

2.

Consumers are completely knowledgeable about the differences among communities in public services and taxes.

3.

Consumers have many communities from which to choose.

4.

Employment opportunities do not restrict or limit the mobility of individuals among communities.

5.

There are no spillovers of public service benefits or taxes among communities (no interjurisdictional externalities).

6.

Each community attempts to attract a population that is exactly large enough to take full advantage of any economies of scale in the supply of public services without being so large as to encounter diseconomies of scale.

Tiebout model communities have two counterfactual characteristics, that is, two characteristics that communities in the real world do not have. 1.

Benefit taxes: The Tiebout equilibrium in each community is a Lindahl equilibrium. Lindahl taxes are perfect benefit taxes. Each individual’s taxprice is exactly equal to her/his marginal benefit from public services.

2.

Homogeneous communities: Tiebout communities are homogeneous, meaning that all residents have the same preferences for public services and the same willingness to pay for them.

3.

We will soon show that even if communities do not use pure benefit taxes and even if each community is not homogeneous, the basic result of the Tiebout model may still apply. That is, the quantity of public services in each community may still be efficient provided that consumers are mobile and are able to choose among communities based on their preferences for public services and their willingness-to-pay for public services.

The Tiebout equilibrium is unstable with non-benefit taxes A.

There are few examples of pure benefit taxes or Lindahl taxes in the real world. In the real world, local public services are financed by non-benefit taxes where an individual’s tax payment depends on some observable economic variable, not on the individual’s marginal benefit for public services. An example of a non-benefit tax is the property tax, where an individual’s tax payment is based on house value. 4

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

B.

C.

The problem with non-benefit taxes: the property tax as an example 1.

The trivial case: If the value of taxpayer’s house values are perfectly correlated with their marginal benefits from public services, then the property tax is a virtual benefit tax. In this case, the property tax equilibrium is a Lindahl equilibrium and the Tiebout hypothesis is still valid.

2.

The non-trivial case: Suppose, however, that house values are not perfectly correlated with marginal benefits from public services. Then the distribution of taxes among individuals is not proportional to marginal benefits. The property tax is not a perfect benefit tax and the property tax equilibrium is not a Lindahl equilibrium. The Tiebout hypothesis is not valid because a.

Individuals will not perfectly sort themselves among communities according to their preferences for public services and their willingness to pay taxes for those services.

b.

Communities will not be homogeneous; residents in each community will have different preferences for public services.

c.

The quantity of public services supplied in each community will not be efficient.

The consequences of using non-benefit taxes 1.

Case 1: Different individuals have different preferences or demands for public services. Community A $100,000 $4,000 4% $4,000

House value Education expenditure per pupil Property tax rate Property tax amount

Community B $50,000 $2,000 4% $2,000

a.

Suppose initially community A and community B are perfectly homogeneous so that the equilibrium in each community satisfies the Tiebout hypothesis. Residents of community A prefer more education spending and are willing to pay higher taxes for it. Residents of community B prefer less education spending and lower taxes.

b.

Problem: This isn’t a stable equilibrium. A resident of community B can move to community A and build a $50,000 house in A. S/he pays the same $2,000 in taxes as s/he paid in B but receives almost $4,000 per pupil in education spending instead of $2,000. (Why is it “almost $4,000” instead of exactly $4,000?)

c.

Result: Residents of B move to A. Tax revenues per pupil decrease in A causing a decrease in per pupil education spending. The original

5

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

residents of A are now worse off because they pay the same taxes but receive less service. Community A is no longer homogeneous. d.

2.

We no longer have a Lindahl equilibrium. Individuals in each community are no longer “on their demand curves”. The original residents of community A pay a tax-price greater than their marginal benefit from the lower level of education expenditure. The former residents of community B who relocated to A pay a tax-price less than their marginal benefit from education.

Case 2: Different individuals have the same demands for public services but different demands for housing. Community A $100,000 $4,000 4% $4,000

House value Education expenditure per pupil Property tax rate Property tax amount

3.

Community B $50,000 $4,000 8% $4,000

a.

Suppose initially community A and community B are perfectly homogeneous so that the equilibrium in each community satisfies the Tiebout hypothesis. Residents of communities A and B prefer the same education spending and are willing to pay the same taxes for it, but residents of community B prefer smaller houses.

b.

Problem: This isn’t a stable equilibrium either. Community B residents can move to community A and build a $50,000 house in A. They pay less tax in A than they paid in B, and still receive almost the same $4,000 per pupil in education spending that they received in B. (Again, why is it “almost $4,000” instead of exactly $4,000).

c.

Result: The same as in Case 1 as residents of B move to A, tax revenue per pupil in A decreases, and per pupil education spending in A decreases. Community A is no longer homogeneous, and we no longer have a Lindahl equilibrium.

This is an application of the free rider problem. a.

Individuals move out of a community in which they pay a tax-price that just equals their marginal benefit from public services.

b.

They move into a community where they are able to free ride on the original residents by paying a tax-price that is less than their marginal benefit from public services.

c.

Because the level of education spending is now below the quantity preferred by the original residents of A, they have an incentive to relocate to another community where the level of public services and taxes is closer to their preferred level.

6

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

d.

IV.

The result is “musical communities”. A Lindahl equilibrium in a community might emerge briefly, but there are always incentives for people to move to take advantage of higher spending or lower taxes so that the equilibrium is unstable.

Solving the instability problem: the effect of fiscal zoning A.

B.

Fiscal zoning 1.

If community A adopts a zoning restriction that sets a minimum house value of $100,000, newcomers will be unable to “free ride” by building a smaller, lower-valued and lower-taxed house.

2.

Fiscal zoning converts the non-benefit property tax into a benefit tax that is consistent with a Lindahl equilibrium. The Lindahl equilibrium is now stable so that communities fiscal zoning can still satisfy the Tiebout hypothesis even if they use non-benefit taxes.

Problems with fiscal zoning 1.

There are two problems with fiscal zoning. First, courts have often limited the ability of communities to impose fiscal zoning. Second, the number of communities required to satisfy the Tiebout hypothesis with fiscal zoning is very large.

2.

Legal limitations on fiscal zoning

3.

a.

Courts have often declared zoning restrictions that establish minimum house values illegal.

b.

However, other types of zoning and deed restrictions (minimum house size, minimum lot size, two-car or three-car garage, garage not facing front of house, and many more such restrictions) are legal.

c.

These legal restrictions raise the cost of housing and may have the same effect as establishing a minimum house value.

Number of communities a.

Fiscal zoning partitions the world into as many communities as there are differences in both preferences for public services and preferences for housing.

b.

If there are two types of housing and two types of public services, there must be 22=4 different homogeneous communities to satisfy exactly the preferences of all individuals. With 3 housing types and 3 public services, there must be 32=9 homogeneous communities to satisfy the Tiebout hypothesis.

c.

Thus, the number of communities required to satisfy the Tiebout hypothesis increases exponentially with the number of different 7

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

types of preferences for public services and with the number of different types of preferences for housing (or for any other taxed private good). V.

The Tiebout hypothesis with non-homogeneous communities: Capitalization of fiscal differentials A.

In the real world, communities are not homogeneous. However, if taxes are capitalized into the value of housing, then the Tiebout hypothesis may still be result in an efficient quantity of public services even though some communities are nonhomogeneous.

B.

Capitalization 1.

3.

Capitalization means that the discounted present value of all future tax payments is included in the current price of housing. The price of a house is lower by an amount equal to the discounted present value of all future taxes. a.

In this case, owners of large houses in a heterogeneous community pay higher taxes than owners of similar houses in a homogeneous community. However, with capitalization, the difference in taxes between the two communities is capitalized into the price of large houses in the heterogeneous community. This means the price of a large house in the heterogeneous community is lower than the price of the same house in a homogeneous community. The lower price offsets the higher taxes.

b.

Similarly, owners of small houses in a heterogeneous community pay lower taxes than owners of similar houses in a homogeneous community. The tax differential is capitalized into the price of small houses in the heterogeneous community so that the price of a small house in the heterogeneous community is higher than the price of the same house in a homogeneous community. The higher price offsets the lower taxes.

Example Community A $100,000

House value Education expenditure per pupil Property tax rate Annual property tax payment

a.

$4,000 4% $4,000

Community B Half $100,000 Half $50,000 $4,000

Community C $50,000

5.33% Big: $5,333 Small: $2,666.50

8% $4,000

$4,000

There are three communities. Community A is homogeneous and consists only of large houses. Community C is also homogeneous and consists only of small houses. Community B is heterogeneous

8

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

and consists of both kinds of houses. Assume for convenience that houses do not depreciate and that the discount rate is 10 percent. b.

Initially, individuals who prefer small houses have an incentive to move to Community B where they can pay lower taxes but receive the same amount of education. However, the increased demand bids up the price of small houses in B. On the other hand, individuals who prefer large houses initially have an incentive to avoid Community B where they would pay higher taxes for the same amount of education. This depresses the demand for large houses in B so that the price of large houses in B falls.

c.

An equilibrium is reached when the price of a small house in B is $58,700 and the price of a large house is $91,300. The present value of taxes in each community is then: Big Houses $40,000 $48,700 NA

Community A Community B Community C

Small Houses NA $31,300 $40,000

[Remember the formula for the present value of an infinite payment stream: Present value = (Annual Payment) ÷ (Discount rate)]

4.

d.

Because they pay higher taxes in B than in A, buyers of large houses are unwilling to pay as much for a house in B as they pay for the same house in A. A large house that sells for $100,000 in A sells for only $91,300 in B ($100,000 minus the additional $8,700 in the present value of taxes).

e.

Because they pay lower taxes in B than in C, buyers of small houses are willing to pay more for a house in B than they pay in C. A small house that sells for $50,000 in C sells for $58,700 in B ($50,000 plus the $8,700 in the present value of tax savings).

With capitalization of fiscal differentials, differences in the prices of the same house in different communities reflect differences in taxes over the expected lifetime of the house. a.

The higher price for a small house in community B is like an entry fee that just offsets the lower tax payments in community B. The lower price for a big house in community B is like a subsidy that just offsets the higher tax payments in community B. Thus, individuals are neither better off nor worse off in the homogeneous communities, A and C, than in the non-homogeneous community, B.

b.

There is no longer an opportunity to free ride by moving to another community. The incentive for residents to relocate is eliminated.

9

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

c.

5.

VI.

A stable equilibrium now requires that each community’s residents have the same preferences for public services and the same willingness-to-pay but it does not require that all residents in a community have the same housing preferences. In other words, it requires that communities be homogeneous with respect to resident’s preferences for public services but not with respect to their housing preferences.

With perfect capitalization of fiscal differentials among communities, the quantity of public services is efficient whether even in communities where residents have heterogeneous preferences. Moreover, each community achieves a stable Lindahl equilibrium even with non-benefit taxes. a.

The Tiebout hypothesis state that if individuals are able to choose among communities based on their preferences for public services, each community will supply the quantity of public services that is efficient given the preferences of its residents.

b.

In the Tiebout model, unlike the real world, all communities use perfect benefit taxes and the residents of each community have homogeneous preferences for public services.

c.

We have shown, however, that the Tiebout hypothesis may still be valid even if communities use non-benefit taxes like the property tax and even if some communities are composed of residents with heterogeneous preferences. To get these results, communities must be able to use fiscal zoning or other land-use restrictions that have the same effect as fiscal zoning and fiscal differentials among communities must be capitalized into housing prices.

The Tiebout hypothesis and the real world A.

The assumptions of the Tiebout hypothesis are unrealistic. Communities in the Tiebout model have characteristics (perfect benefit taxes and homogeneity of preferences) that communities in the real world do not have. So does the Tiebout hypothesis have any relevance? Does the real world have any characteristics that resemble the Tiebout model and its implications? Consider, for example, these questions: 1.

Do individuals take account of differences in public services and taxes among communities when deciding where to live, at least in metropolitan areas?

2.

Do most metropolitan areas include multiple communities offering different packages of public services and taxes?

3.

Even though communities in a metropolitan area are not perfectly homogeneous, is there more homogeneity within each community than in the metropolitan area as a whole? Is there more homogeneity in metropolitan communities than in non-metropolitan communities? 10

ECO 4554: Economics of State and Local Government Public Choice through Mobility: The Tiebout Hypothesis

C.

4.

Do zoning and other land-use restrictions exclude some individuals and some types of land use? Do these restrictions limit the degree to which some residents can free ride by paying below-average taxes while consuming the same level of services as individuals who pay average or above-average taxes?

6.

Are fiscal differentials among communities capitalized into land and housing prices? Do differences in housing prices among communities correspond to differences in public services and taxes? Do houses in communities with good schools, for example, cost more than similar houses in communities with poor schools?

If there are reasonable similarities between the real world and the Tiebout model, even though the Tiebout assumptions and implications are unrealistic, then the Tiebout hypothesis still offers some important lessons for public policy. 1.

2.

Any policy that makes a community more heterogeneous increases the potential for free riding so that the quantity of public services is likely to be less efficient. a.

An example is requiring construction of low-income housing in highincome communities as a condition for approving development.

b.

The result of such a policy is that the quantity of public services in these communities may be less than the efficient quantity as higher income residents leave the community or substitute privately supplied goods and services for public services.

Metropolitan consolidation reduces the number of communities and therefore the number of different combinations of public services and taxes from which individuals can choose. a. Metropolitan consolidation occurs when several smaller communities in a metropolitan area combine into a single larger community or when one or more cities in a metropolitan area consolidate with a county (for example, Jacksonville-Duval County or Miami-Dade County in Florida). b. In an area with fewer larger communities, the communities are likely to be less homogenous than in an area with more smaller communities. Less homogeneity is likely to result in a less efficient quantity of public services.

3.

The quantity of public services is more efficient when public services are financed by user fees or by taxes that are closely related to the benefits that individuals receive from public services. Examples of local benefit taxes might include property taxes to finance fire protection and sewer charges based on water consumption.

11