Search

Our model postulates that the price of a home is the capitalized value of services from its physical characteristics and local public goods less property taxes. Assuming properties are assessed at market value and that households vote so as to maximize the value of their properties, we show that if ‘tax prices’ are independent of the level of public services, the voting outcome is a Lindahl equilibrium. Even when tax prices are not constant, market value assessment reduces disagreement over the appropriate size of the local government budget.

How a developer chooses to subdivide his landholding depends on the distribution of potential buyers and their maximum offers for different sized plots. These offers depend, inter alia, upon factors outside the developer's control such as community tax rates and public services. The developer's optimal subdivision scheme for a continuum of buyers is derived, together with the implied prices of land on different sized plots. We also illustrate the effect of alternative tax regimes on the profit-maximizing density of development, when the landholding is subdivided into equal-sized plots, and exhibit the conflict of interest between developers and existing community residents.

When public expenditures for local services are determined by a referendum process, it is impossible to tell a priori whether increasing the proportion of nonusers increases or decreases the voting equilibrium [1,2]. For a given tax price and income nonusers have a lower demand than users for local services, but the tax price of local services to users is lowered by the presence of nonusers. Stigler [3] therefore asks an empirically interesting question when he seeks to ascertain the influence of Catholic parochial school parents on expenditures per public school pupil. This reply shows that Stigler's use of arithmetic averages for state-wide data probably biases his regression coefficients, and that an alternative specification provides a more powerful test of Stigler's model. Following Stigler let us assume that the demand function for expenditures per public school child is Cobb-Douglas and that all families within a given political jurisdiction pay the same taxes. Letting si denote the proportion of voting families who use the public schools in district i, Ti taxes per family desired by an "average" user, yi the income of an "average"user, and P the cost per "quality" unit of education (assumed constant across observations), Stigler shows that

The relationship between Pareto optimal (0s) and revenue maximizing (Or) tolls is examined for queuing models that permit balking. When customers have the same value for waiting time, Q, =Or provided the entrepreneur can impose a simple two-part tariff. With heterogeneous values for waiting time, Or can be greater than, equal to, or less than H,. Expanding the number of servers and charging multi-part tariffs are shown to be alternative methods for segmenting the market, and the welfare implications of these two strategies are explored.

In a recent article (1970), Robin Barlow compared the Pareto-optimal level of output for a pure public good with the output that would be chosen by majority voting. Applying his model to expenditures on public education, Barlow concluded that actual outlays per pupil were about 26 percent below the Pareto-optimal level (1970, p., 1034). This note questions the procedure by which price and income elasticities were estimated, corrects an error made in computing tax burdens by income class, and formulates the problem in terms of a "pure communal good," which is a more realistic representation of public education expenditures than Samuelson's concept of a "pure public good."