Should Benefit–Cost Analyses Take Account of General Equilibrium Effects?

• Date: 16 October 2007
• DOI: https://doi.org/10.1016/S0193-5895(07)23011-8
• Pages: 247-272
• Published date: 16 October 2007
• Author: V. Kerry Smith,Jared C. Carbone

SHOULD BENEFIT–COST

ANALYSES TAKE ACCOUNT

OF GENERAL EQUILIBRIUM

EFFECTS?

$

V. Kerry Smith and Jared C. Carbone

ABSTRACT

This paper demonstrates the importance of general equilibrium (GE)

feedback effects inside and outside markets for the measurement of the

efficiency costs of taxes in a distorted economy. Our specific focus is on

the changes in environmental amenities that can result from pollution

externalities generated from production activities. Even when amenities

are under three percent of virtual income, the error in the GE

approximations of the welfare effects of new taxes with pre-existing

distortions can increase threefold. The nature of the link between the

source of the external effects influencing amenities and the changes in

amenity services can alter the conclusions one would make about the

merits of an intervention based on benefit–cost analyses.

$

Partial support for this research was provided by the US Environmental Protection Agency

Star Grant Program under grant number Rd83-092301 through RTI sub-contract number

1-420-8892.

Research in Law and Economics, Volume 23, 247–272

Copyright r2007 by Elsevier Ltd.

All rights of reproduction in any form reserved

ISSN: 0193-5895/doi:10.1016/S0193-5895(07)23011-8

247

1. INTRODUCTION

Measures of the general equilibrium (GE) net benefits for a policy that

affects one commodity can be developed with only that good’s demand-and-

supply functions.

1

While there are a number of assumptions required to

assure the logic of the process associated with the definition of this measure

is valid, many economists seem to have ignored these ‘‘details’’ and assume

the process of incorporating GE effects is straightforward.

2

Therefore, one

might expect to find a GE dimension as a routine component of most

benefit–cost analyses for major social regulations. In fact, the opposite is

true. The Environmental Protection Agency is the leading agency in both

the quality and extent of documentation in the benefit–cost analyses

conducted for its major rules. Yet few of its Regulatory Impact Analyses

(RIAs) acknowledge GE effects. Where they have been discussed, they are

part of arguments that partial equilibrium (PE) measures of direct costs and

benefits are ‘‘good first approximations.’’

3

A notable exception can be found

in the research agenda discussion in the first Prospective Analysis required

by Section 8 of the 1990 Clean Air Act Amendments. These periodic reports

are intended to provide a forward-looking assessment of the benefits and

costs of periodic revision in the standards and the associated rules governing

the ambient concentrations of the criteria of air pollutants. The first such

report clearly acknowledges the importance of general analysis for these

types of assessments noting that:

One potentially important area where research may enhance our ability to conduct

broader assessments is the development of computable general equilibrium (CGE)

models that can be implemented in a resource-effective manner.yA well designed CGE

model may also enhance our capability to estimate the effects of the tax interaction

effect, both on the cost and the benefit side. (pp. I-2)

This paper argues that the existing literature has overlooked an important

source of error in conventional practice. Social regulations, especially

environmental policy, are intended to influence services available outside

markets. To the extent these services make non-separable contributions to

preferences or production activities they can be expected to have large

effects on the properties of PE welfare measures. These effects have

implications for the recommended methods for taking into account GE

effects. Conventional strategies for applied welfare analysis usually overlook

the importance of the feedback effects outside the market that arise from

these non-separabilities. Simply stated, our argument suggests that when

non-market amenities affect the demands for market goods and the

V. KERRY SMITH AND JARED C. CARBONE248