Search

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 rapid growth of suburban communities in recent years has led to increased congestion and greater physical burdens upon the public facilities of those communities. The purpose of this paper is to examine these problems in the specific case of public parks and recreation areas and to demonstrate that the full complexity of the issues has not yet been grasped even by those legislatures and courts that have considered the problem of congestion externalities. We begin with a simple model of congestion externalities and public recreational facilities, and then review the judicial approach to statutes attempting to mitigate these effects. Finally, we present a model for internalizing the principal effects of suburban growth upon public recreational areas, and offer some guidelines for an efficient legislative solution to these difficult problems.

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."

In their recent article, Elizabeth Bailey and John Malone deduce the optimal capital-labor ratios for firms with four different management objectives (maximum profit, return on investment, sales, and output), given that these firms are subject to four types of regulatory constraint (profit no greater than a fair return on investment, on cost, on output, and profit no greater than some absolute upper bound). This note points out some flaws in their analysis, and derives the correct results by more elementary methods.