Cost-Benefit Analysis and Public Policy

464 Pages·2009·6.63 MB·English

by  the Association for Public Policy Analysis and Management

COST-BENEFIT
ANALYSIS AND
PUBLIC POLICY
Davidl.. Weimer,Editor
V Preface
V Introduction
David L Weimar, Cost-Benefit Analysis and Public Policy
The Issue of Standing
Whittington, D.,& MacRae,D.,Jr. (1986). The M ol standing in (mt-benefit analysis.
PAM. 5(4). 665-82.
MacRae. D.,Jr, & Whittington, D. (1983) Assesing preference ln cost-benefit analysis: Reflections on
nrral water supply evaluation in Haiti. JPAM,7(2), 246—63. 21
'll'umbull, W.N. (l990). Who has standing in cost-benefit analysis?jPAM.9(2), 20l—18. 37
Whittington. D.,& MacRae,D.,jr. 0990). Comment Judgments about who has standing in cost-benefit
analysis. JPAM.9(4). 536—47.
'l‘l'umbull, W.N. (l990). Rqaly to Whittington and MacRæ. ]PAM.9(4). 543—50. 63
Yerba.R. O.,]r. (l991).'(‘.ommenl: Doesbenefit-cost analys's stand alone? Rights and standing,
]PAM, 10(l), 96—105. 65
Discountingfor finie
Brown, P.G. (1988). Policy analysis, welfare economic. and the greenhouse effect.JPAM,7(3). 471—75.
Kcle. A, & Scheraga.]. D. (1990). Discounting the benefits and costsof environmental regulations
. JPAM. 9(3). 381—90­
Metcalf, G. E.. & Rœenlhal, D. (1995). The “'new‘ Viewof invstment decisions and public policy
analysis: An application to green lights and cold refrigerators. JPAM,14(4), Sl7—31.
Moore. M. A, Boardman, A E.. Vrning, A R., Weimer. D. L, & Greenberg, D. H. (2004). “Just give me .
a number!" Practical values for the social discount rate. JPAM,23(4). 789—812.
Cart-Barefir Analysisand Public Policy Edited by David L Weimer
o 2008 the Association for Public Policy Analysis md Management. ISBN: 978-1-405-l90l6—9
Risk and the “Value of Life”
Moore, M.]…& Vrscusi,W. K. (1988). Doubling the estimated value of life: Raults from new
occupational fatality data. jPAM.7(3). 476-90. 125
Fisher. A, Chtstnut. L G., & Violette,D. M. (1989). The value of reducing risks of death:
A note on new evidence. ]PAM,8(1), 88—100. 139
Knetsch,_|.L (1995). Assumptions, behavioral finding, and policy analysis. JPAM,l4(l), 68-78. 151
Mrozek,J. R., & Taylor, L 0. (2002). What determines the value of life. Ameta-analysis.
I61
JPAM. 21(2). 253—70.
HammilLJ. K (2002). Understanding difl‘erencs in estimates of the value of mortality risk.
JPAM,2l(2). 271—73. 179
Krupnick. A (2002). The value ol reducing risk of death: Apolicy perspective.
jPAM, 21(2). 275—82.
Non-Use Value as a Benefit Category
Rosenthal. D. H., & Nelson, R. H. (1992). Why existence value should no! be used in cost-benefit
analysis. jPAM, 11(1), 116—22.
Kopp,R.]. (1992). Whyexistence value should be used in cost-benefit analysis.
JPAM,11(1),123—30.
Quiggin,j. (1993). Existence value and benefit-cost analysis: Athird view.JPAM,12(l), 195—99.
Smith,V.K.(1993).Rethinkingthe rithmeticofdamage …!.)PAM, 12(3).589-95.
CDAin Administrative and Legal Context
Boardman, A, Vining. A, & Waters. W.G., II (1993). Costs and benefits through bureaucratic lenses:
214
Example of a highway proieuJPAM. 12(3). 532—55.
Krutilla, K. (2005). Using the Kaldor-Hicltstableau formal for cost-benefit analysis and policy
evaluationJPAM. 24(4). 864—75. 235
Kopp,R.j., & Smith, V.K. (1989). Benth stimalion goes to court The case of natural resource
damage asessment. jPAM,8(4). 593—612.
hrbe. R. 0., Jr. (1998). ls cast-bendit analysis legal? Three mlts. jPAM, 17(3). 419—56. 275
"311111810“W. ..Morgenstem, R. D., & Nelson, P. (21-1). On the accuracy of regulatory cog estimates.
JPN“. 19(2). 297-322.
CDAApplications
bong, D. A, Mallar, C. D., & 'l‘homlon, C. V.D. (1981). Evaluating the benefits and costs of theJob Corps
JPItM, 1(1). 55—76. 333
Goma-ibanez, J. A. icone, R.A. & O‘Connell,S. A (1983). Restraining auto impons: Dots
anyone win?]PAM, 2(2), l96—219.
Kamerud, D. B. (1988). Benefits and costs of the 55 MPHspeed limit: NewStimals and their
implimtions JPAM,7(2), 341—52. 370
Vitaliano,D. F.(1992). Aneconomic mme… of the social ants of highwaysalting and the
efficiency of substituting a new deicing material. JPAM.11(3). 397—418.
Devaney,B., Bilheimer. L. & Schore,]. (1992). Medicaid oœts and birth outcome: The effectsof
prenatal W1Cparticipation and the use of prenatal mm.]PAM, ll(4), 573—92.
Schwindt, R., Vining. A. & Globennan, S. (21—l). Net los: Acost-benefit analysis of the
Canadian Pacific salmon fishery. IPAM. 19(1). 23—45. M9
Chen, 6.. & Warburton, R. N. (2006). Do speed œmems produce net benefits? Evidence from
British Columbia, Canada. jPAM,25(3). 661—78.
Index 457
Prefoce
Cost-benefit analysis (CBA)holds a prominent, but controversial, place among the
techniques of public policy analysis. At one extreme, some economists view CBAas
synonymous with good policy analysis. At the other extreme, a diverse group of
political philosophers attack it as a technocratic undercutting of democratic values,
a utilitarian threat to individual rights, or a crass debasing of public discourse. Yet,
CBA is neither panacea nor fatal poison. Though often impractical to implement
and rarely fully appropriate as a formal decision rule, it provides policy analysts
with insights for organizing their thinking about the goal of efficiency and specific
techniques to help guide the measurement and valuation of the impacts of policies
in terms of the resources they require and the effects they produce. Exposure to the
fundamental issues surrounding the use of CBAand examples of its practical appli­
cation have value to current and future practitioners of policy analysis as well as to
researchers in the policy sciences. This volume seeks to facilitate such exposure by
drawing together into a convenient collection the fine articles on CBAand its appli­
cation that have appeared in the Journal of PolicyAnalysisand Management (JPAM).
As a teacher and sometimes producer or user of CBA,as well as a participant in
some of the debates over its appropriate role in policy analysis, I am honored to
have the opportunity to assemble and to introduce this volume. Asa member of the
Association for Public Policy Analysis and Management, I am very pleased that our
journal has produced an interesting, balanced, and practically oriented set of arti­
cles from which to choose. I especially enjoyed returning to some of the articles that
appeared during my editorship of JPAM—abit like becoming reacquainted with old
friends. It is my hope that practitioners, teachers, and students will share my view
that there is much in these articles to help us make better use of CBAas a resource
for good policy analysis.
David L. Weimer
University of Wisconsin—Madison
Journal of Policy Analysis and Management,
© 2008 by the Association for Public Policy Analysis and Management
Published by Wiley Periodicals, Inc. Published online in Wiley lnlerScience
(www.interscience.wiley.com)
DOI: lO.l002/pam.
Introduction: DavidL. Weimer
Cost-Benefit Analysis and
Public Policy
The economic approach to policy analysis gives a central role to efficiency.The con­
ceptual starting point is Pareto efficiency.An allocation of resources to production
and goods to consumption is Pareto efficient if it is impossible to find an alternative
allocation that makes at least one person better off without making anyone else
worse off. Reallocations are Pareto improving if they make someone better off
without making anyone else worse off. Seeking Pareto improvements has obvious
appeal; in specific circumstances, one would have to be malevolent to oppose gains
to some that require no others to bear losses. Out of practical necessity, however,
economic analysis generally measures gains in efficiency in terms of potential,
rather than actual, Pareto improvements. A reallocation is potentially Pareto
improving if it generates an excess of gains over losses so that it would be possible,
through costless transfers, to make the reallocation Pareto improving. Cost-benefit
analysis (CBA) comprises the concepts and methods for measuring benefits and
costs in a money metric to determine if proposed policy alternatives are potentially
Pareto improving.
The CBA decision rule, adopt the combination of policies that maximizes the
excess of benefits over costs, suffers from at least two limitations as a guide for pub­
lic policy. First, efficiency is rarely the only relevant value in choosing among policy
alternatives. Distributional concerns, individual freedom, and national security,
among other values, often have widely recognized substantive relevance to pruden­
tial choice in various policy areas, and political feasibility often has instrumental
importance in actual arenas of choice. Conflicts between distributional values and
the CBAdecision rule are particularly fundamental in that Pareto efficiency takes the
existing distribution of wealth as given and potential Pareto improvements do not
require that everyone actually get at least their initial shares. Second, even when
efficiency is the only relevant value, it may not be practically possible to measure it
in terms of the money metric. Often only some policy effects can be monetized. The
reliability of the CBAdecision rule depends on the comprehensiveness of the mon­
etization. Either excluding important effects or monetizing them incorrectly can
lead to the choice of policies that do not promote efficiency.
Much criticism of CBAas a decision rule involves these limitations. Nevertheless,
situations do arise, most often in the context of infrastructure investments such
as bridges, dams, and highways, in which efficiency can be reasonably taken as the
relevant value and all major impacts can be confidently monetized. CBAhas much
broader application, however, as a protocol for identifying and monetizing the
efficiency effects of policies.
Efficiency is almost always one of the relevant goals in policy analysis. CBAcon­
cepts and methods enable analysts to rank alternatives in terms of their efficiency.
When the ranking is in terms of the money metric, not only is the comparison of
Cosl-Benefil Analyst}and Public Policy Edited by David L. Weimet
O 2008 the Association for Public Policy Analysis and Management. ISBN: 97B-l-405-l90l6—9
Introduction: Cost-BenefitAnalysis and Public Policy / 3
alternatives in terms of efficiency facilitated, but tradeoffs between efficiency and
other goals can be made more easily. Thus, even analysts working in policy areas in
which CBAis inappropriate as a decision rule are likely to find it useful as a proto­
col for measuring efficiency.
The articles collected in this volume explore many of the important issues that
arise in the application of CBA.They assume some familiarity with the basics of
CBA,which can be found in a number of texts: Harberger (1972), Mishan (1976),
Sugden and Williams (1978), Gramlich (1990), Zerbe and Dively (1994), Dinwiddy
and Teal (1996), and Boardman and colleagues (2006). As the JPAM articles were
written originally for an interdisciplinary audience of scholars and practitioners
interested in policy analysis, they nicely bridge the gap between conceptual issues
and practical application. They offer policy analysts a deeper understanding of
CBA and the way that it can help them improve their craft. Though generated
through the decentralized process of journal submission, they are surprisingly
comprehensive in terms of covering the big issues that arise in the application of
CBA.The following sections provide a brief overview of the topics covered in this
volume.
THE ISSUE OF STANDING
The guiding principle in CBAfor monetizing policy effects is willingness-to-pay. The
social benefit of a policy effect is the sum of the maximum amounts that people
would be willing to pay to obtain it; the social cost of a policy effect is the sum of
the maximum amounts people would be willing to pay to avoid it. But whose
willingness-to-pay should count? Should willingness-to-pay for things that would
generally be viewed as abhorrent count?
Dale Whittington and Duncan MacRae, Jr. (1986) address these questions in their
seminal article introducing the rubric of standing. Atone extreme, standing deals with
the relatively straightforward issue of the appropriate jurisdictional extent of society
(world, country, or subnational unit). More complicated issues of standing involve the
composition of society within the jurisdiction (citizens, legal residents, visitors, or ille­
gal aliens) and the treatment of suspect preferences (criminal, ungenerous, or
immoral behavior). For example, MacRae and Whittington (1988) raise the question
of how to take account of the apparent disutility suffered by rural Haitian men as a
result of water supply improvements that make the lives of their wives easier.
In a spirited and thoughtful exchange, William Trumbull (1990a, 1990b),
Whittington and MacRae (1990), and Richard Zerbe (1991) debate some of the
more difficult cases of standing involving criminal gains, future generations, and
cross-border effects. Trumbull (1990) attempts to resolve these cases by asserting
that, just as analysts must consider physical constraints in valuing alternative
projects, they should also take into account social constraints. Zerbe (1991) seeks
resolution by arguing that issues of standing should be viewed as problems in the
assignment of rights, an argument that he develops more fully in his 1998 article
included in the section on the use of CBA in administrative and legal settings.
Although neither of these approaches provides a fully satisfactory resolution, each
offers insights that are likely to be helpful to analysts in dealing with the issues of
standing that they confront.
DISCOUNTING FOR TIME
Common sense and simple capital theory tell us that a dollar of benefit received
today is worth more than a dollar received a year from now. This preference for
earlier rather than later payments holds even if we assume that the payments are
Journal of Policy Analysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management
4/ Introduction: Cost-BenefitAnalysis and Public Policy
certain and that they are made in dollars with the same purchasing power. CBA
takes account of this by expressing costs and benefits that accrue over time in terms
of their present values. In its simplest form, the present value of a dollar of cost or
benefit 1years in the future is 1/(1 + d)t, where d is the annual social discount rate.
As the implementation of projects typically involves the accrual of costs before
the realization of benefits, the magnitudes and even sometimes the signs of the
present values of net benefits are often sensitive to the assumed value of the social
discount rate.
Although economists generally accept the notion that future costs and benefits
should be discounted to their present values in CBA,some critics object to its impli­
cations for valuing effects that occur far in the future. Peter Brown (1988) reminds
us, for example, that the use of any positive social discount rate puts vanishingly
small weight on effects that occur in the far future.
Among economists, there continues to be debate over the conceptually correct
social discount rate and how it should be applied in practice. In a real economy,
consumers trade current and future consumption at a different rate than producers
can trade current consumption for future production through investment—the for­
mer, the marginal rate of pure time preference, is generally larger than the latter,
the marginal rate of return on private investment. Jeffrey Kolb and Joel Scheraga
(1990) summarize the shadow-price-of-capital approach to discounting, which
involves converting changes in investment to equivalent changes in consumption
and discounting using the marginal rate of pure time preference.
Developments in investment theory suggest that factors such as irreversibility and
uncertainty in returns bring into question the standard private investment rules
based on present value calculations. Gilbert Metcalf and Donald Rosenthal (1995)
discuss the implications of these developments in investment theory for the evalu­
ation of government programs intended to encourage private investments. Their
discussion suggests yet another level of complication that should be considered in
the debate over appropriate discounting in CBA.
Mark Moore and colleagues (2004) provide a review of the various issues and
approaches to discounting. In addition to the shadow price of capital approach, they
consider the optimal growth rate method and the use of time-declining discount
rates for intergenerational projects. Seeking to provide guidance for practitioners,
they provide recommendations for the numerical values of the social discount rate
to be used for projects with both current and intergenerational time horizons.
RISK AND THE “VALUE OF LIFE"
Many public policies, such as highway improvements, vaccination programs, and
pure food regulations, can change the mortality risks that people face. CBArequires
that such changes in risk be monetized as people's ex ante willingness-to-pay for
them. In practice, however, analysts are rarely in a position to elicit these amounts
directly. Instead, they typically predict the number of lives saved and then apply a
shadow price for a statistical life, a "value of life," estimated from observation of the
tradeoffs people make between mortality risks and other goods in various markets.
Michael Moore and W. Kip Viscusi (1988) demonstrate this approach by relating
occupational mortality risks to wage premiums, that is, how much additional com­
pensation must be paid to workers so that they will accept riskier jobs. Ann Fisher,
Lauraine Chestnut, and Daniel Violette (1989) review value-of-life estimates based on
such labor market tradeoffs, as well as studies of consumer decisions involving pur­
chases of safety devices and surveys that directly ask consumers about their willingness
to pay for risk reductions. These studies report a range of estimates from $1.6 million
to $8.5 million (1986 dollars) for the value of a statistical life. They argue that the
Joumal of Policy Analysis and Management DOI: lO.l002/pam
Published on behalf of the Association for Public Policy Analysis and Management
Introduction: Cost-BenefitAnalysis and Public Policy / 5
appropriate shadow price for lives saved is at the lower end of this range. In a more
recent review of value-of-Iifeestimates, Janusz Mrozek and Laura Taylor (2002) con­
duct a meta-analysis of available studies to argue for a value of statistical life of
between $1.5 million and $2.5 million (1998 dollars). In the same issue, James
Hammitt (2002)and Alan Krupnick (2002)offer useful comments on the meta-analysis.
Agrowing body of empirical research on the behavior of people in circumstances
involving risk has raised some doubts about whether the “value of life” inferred
from observed behavior can be reasonably interpreted as willingness to pay for
reductions in mortality risks. Drawing on a growing body of behavioral research,
Jack Knetsch (1995) considers the implications of the finding that people seem to
value losses much more highly than gains of the same magnitude for the treatment
of risk in policy analysis more generally.
NON-USE VALUE AS A BENEFIT CATEGORY
With development of new measurement techniques and broader substantive appli­
cation, the range of policy effects valued in CBAhas expanded. It is now common to
distinguish between use and non-use values. Use includes changes in consumption
of pure private goods such as water, electricity, and day care, that have traditionally
comprised the benefit categories used in CBA. Over time, the use category has
expanded to include goods, such as visits to wilderness areas, that may be
nonrivalrous in consumption and even to local public goods, such as air quality,
that affect property values. In all these cases, the consumption of the good leaves a
"behavioral trace" that can be observed as the basis for inference about its value to
those who consume it.
The primary component of non-use value is a pure public good called existence
value. It recognizes that people might be willing to pay for such things as the preser­
vation of a wilderness area even if they expect never to visit it themselves. In other
words, they place a value on its mere existence. The development of contingent
value survey methods, which attempt to elicit people's willingness-to-pay through
direct questions, has led to the inclusion of existence value as a benefit in a grow­
ing number of CBAs dealing with environmental issues—for a comprehensive
review, see Bateman and Willis (1999). Donald Rosenthal and Robert Nelson (1992),
Raymond Kopp (1993), John Quiggin (1993), and V.Keny Smith (1993) debate the
appropriateness of including existence value in CBA.
CBA IN ADMINISTRATIVE AND LEGAL CONTEXT
The use of CBAas an input to administrative and legal processes creates incentives
for its distortion. Anthony Boardman, Aidan Vining, and W.G. Waters II (1993) illus­
trate with the case of a highway project how bureaucrats tend to adopt roles as
guardians, spenders, or analysts that lead them to classify costs and benefits in very
different ways. Kerry Krutilla (2005) illustrates how distributional implications of
alternative policies can be presented within the CBAframework using what he calls
“Kaldor—Hickstableaus." Winston Harrington and colleagues (2000) review the
record of cost estimation by federal regulatory agencies, reporting that predictions
of the total cost of regulations generally overestimate costs, while the predictions of
unit costs tend to be more accurate. Raymond Kopp and V. Ken'y Smith (1989)
consider the sources of difference between estimates of environmental damage by
plaintiffs and defendants in court cases under the Comprehensive Environmental
Response, Compensation, and Liability Act. On the one hand, controversy over the
measurement of costs and benefits may dishearten CBA advocates; on the other
hand, it should also cheer them as an indication that CBAhas practical significance!
Journal of Policy Analysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management
o / Introduction: Cost-BenefitAnalysis and Public Policy
Richard Zerbe (1998) adopts a broad perspective on the measurement of benefits
and turns to the law to resolve issues of standing. He argues that doing so removes
many of the normative objections raised against CBA, including those based on
distributional issues. Further, he suggests that CBAmay even play a role in identi­
fying ways that law itself may become more efficient.
CBA APPLICATIONS
Observing efforts to apply CBA helps convey its promise and limitations. The
articles that conclude this volume demonstrate the application of CBAin a variety
of substantive policy areas: social welfare (Long, Mallar, & Thorton, 1981; Devaney,
Bilheimer, & Schore, 1992); trade (Gomez-Ibanez, Leone, & O'Connell, 1983); envi­
ronment (Vitaliano, 1992; Schwindt et al., 2000); and highway safety (Kamerud,
1988; Chen & Warburton, 2006).
REFERENCES
Bateman, l. J., & Willis, K. G. (1999). Valuing environmental preferences. New York: Oxford
University Press.
Boardman, A., Greenberg, D., Vining, A., & Weimer. D. (2006). Cost-benefit analysis: Con­
cepts and practice (3rd edition). Upper Saddle River, NJ: Prentice Hall.
Gramlich, E. M. (1990). A guide to benefit-cost analysis. Englewood Cliffs, NJ: Prentice Hall.
Dinwiddy, C., & Teal, F. (1996). Principles of cost-benefit analysis for developing countries.
New York: Cambridge University Press.
Harberger, A. C. (1972). Project evaluation. Chicago, IL: Markham.
Mishan, E. J. (1976). Cost-benefit analysis. New York, NY:Praeger.
Sugden, R., & Williams, A. H. (1978). Principles of practical cost-benefit analysis. New York:
Oxford University Press.
Zerbe, R. O., & Dively, D. (1994). Benefit—Cost Analysis in Theory and Practice. New York:
HarperCollins College Publishers.
Joumal of PolicyAnalysis and Management DOI: 10.1002/pam
Published on behalf of the Association for Public Policy Analysis and Management