Public Finance, 2nd Edition

any given industry. In any event, one would not want to represent the short-

run changes as recurring annually for the life of the project.

Once these two factors are admitted, it is easy to see that the secondary

chain-reaction proWt game is far overdone. We believe the safest strategy is

simply to ignore proWts and losses in other industries even though changes in

the general equilibrium proWt function are a legitimate part of beneWts. Can

any researcher really hope to trace through all the pure proWts and losses

arising from a given project, both in the short run and long run? The question,

in eVect, answers itself. The cost±beneWt analyst would be well advised to

accept the proposition that aggregate production exhibits (approximately)

constant returns to scale (CRS) and assume no lasting pure proWteVects in

other industries. Assume, also, that short-run gains are completely oVset by

short-run losses as resources move throughout the economy.

At the same time, any pure proWts or losses directly associated with the

project ought to be included. The need to Wnance a decreasing-cost service in

part out of general tax revenues, if price is set equal to marginal costs, is

certainly a relevant project cost. Indeed, the government often invests in a

particular industry precisely because it exhibits decreasing costs. Hence, it

would be foolish to assume that government production necessarily exhibits

CRS.

THE REGIONAL MULTIPLIER GAME

The regional multiplier game attaches a Keynesian-style multiplier analysis

to the basic cost±beneWt framework. Continuing with the example of the

hydroelectric project, suppose the dam site was formerly a wilderness area.

Presumably, the construction and continued operation of the dam will sup-

port all sorts of ancillary services. The people associated with the project have

to be clothed, housed, and fed. Indeed, a small town might spring up around

the dam site, generating a continual annual Xow of income in a region

formerly devoid of economic activity. By the very nature of the Keynesian

multiplier, these secondary income eVects will be a multiple of the project's

direct costs and/or beneWts. Therefore, if they are included as project beneWts,

the project would necessarily have a positive present value.

Something must be wrong here, for these regional multipliers appear to

suggest that virtually any project placed in an underdeveloped region of a

country will be worthwhile. The crux of the problem is that Keynesian

multiplier analysis is simply irrelevant to the fundamental goal of cost±

beneWt analysis, which is to determine the best use of society's scarce re-

sources. Cost±beneWt analysis begins with a presumption of full employment

so that, strictly speaking, it is concerned with the maximum expansion of

society's production-possibilities frontier. Keynesian multiplier analysis, on

the other hand, is concerned with moving an unemployed economy to its

822 THE REGIONAL MULTIPLIER GAME