Public Finance, 2nd Edition

The Full Employment Assumption

The Wnal principle is that cost±beneWt analysis implicitly assumes a fully

employed economy, unless speciWcally stated otherwise. It is intended pri-

marily as an exercise in microeconomic analysis, designed to help govern-

ments select among alternative uses of scarce resources. One immediate

implication of the full-employment assumption is that government invest-

ment projects not only compete directly among themselves for scarce re-

sources. Each project must also dominate all alternative private sector uses

of the same scarce resources to be deemed worthwhile.

The formal structures of public sector cost±beneWt and private sector

investment analysis may be identical, but the former is ultimately more

diYcult and, in many ways, more subjective than the latter. Broadly speak-

ing, the diVerences between them in application center around the point that

market prices do not always reXect marginal social values. Private investment

analysis typically ignores this fact, but the government's cost±beneWt analysis

cannot. Indeed, the very name ``cost±beneWt analysis'' suggests that govern-

ment project evaluation goes well beyond the calculation of proWts at

current and expected future market prices and interest rates that one nor-

mally associates with private sector investment analysis. ProWtability is gen-

erally not the proper criterion for selecting among alternative public sector

projects.

In trying to evaluate the present value formula, government policy-

makers face the same issues that beset private investment analysts, plus

another whole set of diYcult issues unique to the public sector. The common

issues are the choice of an appropriate rate of discount and the general

problem of uncertainty. Even so, their solutions are often markedly diVerent.

Let us consider each of them brieXy by way of introduction.

ISSUES COMMON TO COST±BENEFIT AND PRIVATE

INVESTMENT ANALYSIS

The Discount Rate

By inspection of Eq. (23.1), it is obvious that choosing the proper rate of

discount is crucial to any investment analysis. Using a rate that is too low

(high) introduces two types of errors. One is that too many (few) projects pass

the present value test. The other is that the present value formula biases

selection toward projects whose net returns occur later (earlier) rather than

earlier (later), given equal undiscounted streams of returns. Furthermore,

project selection tends to be very sensitive to even small absolute errors as a

general rule. Choosing a discount rate of 5% when the ``true'' rate is 10% or

722 ISSUES COMMON TO COST±BENEFIT AND PRIVATE INVESTMENT ANALYSIS