Public Finance, 2nd Edition

investments. But complete objectivity is virtually impossible to maintain if

the distribution of costs and beneWts matters. Consequently, many cost±

beneWt studies ignore distributional considerations entirely and attempt to

measure only aggregate present values. Ignoring the distribution problem

does not solve it, however. It merely consigns to cost±beneWt analysis the

more modest task of determining the relative eYciency of government pro-

jects, rather than ranking projects according to their contributions to social

welfare. Perhaps this is all cost±beneWt analysis can hope to achieve.

Chapter 27 explores two fundamental questions relating to the distribu-

tion of costs and beneWts. First, under what conditions can the distribution

question safely be ignored in cost±beneWt analysis? Second, if these condi-

tions do not exist, how should distributional considerations enter into the

analysis? We will consider some suggestions for incorporating the distribu-

tion of the net beneWts that have appeared in the literature. Unfortunately,

none of them avoids the problem of subjectivity. One can almost always

design a set of subjective distributional weights that overwhelm whatever

objective eYciency information a cost±beneWt study might contain, no matter

what approach is taken.

PITFALLS IN COST±BENEFIT ANALYSIS

Chapter 28 concludes Part IV by discussing a number of common pitfalls in

cost±beneWt analysis. It may seem odd to spend time on false claims, but the

sad truth is that public debate on government projects all too often focuses on

irrelevancies and misconceptions. Roland McKean long ago labeled these

errors ``secondary beneWts'' (costs), implying that people Wnd new sources of

beneWts (costs) to add to the true beneWts (costs) in order to make the project

seem all the more attractive (unattractive).

8

Unfortunately, these secondary,

bogus beneWts (costs) can be many times greater than the project's true

beneWts (costs).

We will consider Wve of the more common pitfalls:

1. The chain-reaction game, in which proWts (losses) generated through-

out the economy as a result of a government project are indiscriminately

attached to the beneWts (costs) of the project.

2. The regional multiplier game, in which Keynesian-type multipliers are

used to enhance the beneWts (costs) of the project.

3. The labor game, a variation of (1) and (2) in which the employment

eVects of the project are viewed as one of its major beneWts, perhaps even the

principal beneWt. This error has become a matter of law. Federal and state

legislation often requires that cost±beneWt studies explicitly consider the

8

R. N. McKean, EYciency in Government through Systems Analysis, Wiley, New York,

1958, chaps. 8 and 9.

728 PITFALLS IN COST±BENEFIT ANALYSIS