ANALYTICAL STUDIES FOR THE U.S.
ENVIRONMENTAL PROTECTION AGENCY
OLUME vm
Noise
Abatement
Policy
Alternatives for
Transportation
The
National
Research
Council
-------
ANALYTICAL STUDIES FOR THE U.S.
ENVIRONMENTAL PROTECTION AGENCY
VOLUME VIII
Noise
Abatement
Policy
Alternatives for
Transportation
A Report to the
U.S. Environmental Protection Agency
from the
Committee on Appraisal of Societal Consequences
of Transportation Noise Abatement
Assembly of Behavioral and Social Sciences
National Research Council
NATIONAL ACADEMY OF SCIENCES
Washington, D.C. 1977
-------
NOTICE: The project that is the subject of this report was approved by the Governing Board
of the National Research Council, whose members are drawn from the Councils of the
National Academy of Sciences, the National Academy of Engineering, and the Institute of
Medicine. The members of the Committee responsible for the report were chosen for their
special competences and with regard for appropriate balance.
This report has been reviewed by a group other than the authors according to procedures
approved by a Report Review Committee consisting of members of the National Academy
of Sciences, the National Academy of Engineering, and the Institute of Medicine.
This study was supported by the Environmental Protection Agency.
Contract No. 68-01-2430
International Standard Book Number: 0-309-02648-2
Library of Congress Catalog Card Number 77-87121
Available from
Printing and Publishing Office
National Academy of Sciences
2101 Constitution Avenue
Washington, D.C. 20418
Printed in the United States of America
-------
COMMITTEE ON APPRAISAL OF SOCIETAL
CONSEQUENCES OF TRANSPORTATION NOISE
ABATEMENT
WILLIAM BAUMOL (Chairman), Departments of Economics, New York
University and Princeton University
DONALD E. BROADBENT, Department of Experimental Psychology, Uni-
versity of Oxford
ELIZABETH A. DEAKiN, Center for Transportation Studies, Massachusetts
Institute of Technology
ARTHUR DE VANY, Department of Economics, Texas A&M University
KENNETH ELDRED, Principal Consultant, Bolt Beranek & Newman, Inc.
ALAN FREEMAN, School of Law, University of Minnesota
MARCIA GELPE, School of Law, University of Minnesota
DAVID GLASS, Department of Psychology, The Graduate School and
University Center, City University of New York
DAVID M. GREEN, Department of Psychology and Social Relations,
Harvard University
CALVIN s. HAMILTON, Director of Planning, City of Los Angeles
EDWARD K. MORLOK, Department of Civil and Urban Engineering,
University of Pennsylvania
JON P. NELSON, Department of Economics, The Pennsylvania State
University
WILLIAM SAMPSON, Department of Sociology, Northwestern University
ALAN A. WALTERS, Urban Projects Department, World Bank (on leave
from Department of Political Economy, Johns Hopkins University)
STAFF:
JEROME E. SINGER, Study Director
EDWARD I. FRIEDLAND, Senior Research Associate
DONNA c. GOSNELL, Administrative Secretary
CAROL BEERS, Secretary
111
-------
Contents
FOREWORD JX
PREFACE XI
SUMMARY AND RECOMMENDATIONS 1
I POLICY AND LEGAL ISSUES 13
1 The Choice of Policy Instruments 15
2 Legal Issues 32
II TRANSPORTATION NOISE: ITS MEASUREMENT,
SOURCES, AND PROSPECTS 43
3 Noise Indices 45
4 Noise from Transportation Sources 60
5 Projections of Transportation Activity 81
III BENEFITS AND COSTS OF TRANSPORTATION
NOISE ABATEMENT 103
6 Benefits of Noise Abatement 105
7 Monetary Measures of the Benefits of Abatement:
Property-Value Analysis 126
8 Costs of Noise Abatement 161
9 Cost-Benefit Analysis: Some Illustrations 183
-------
Foreword
This report is one of a series prepared by the National Research Council
for the U.S. Environmental Protection Agency.
In June 1973 the Subcommittee on Agriculture, Environmental, and
Consumer Protection of the Appropriations Committee of the U.S.
House of Representatives held extensive hearings on the activities of
EPA, and the ensuing appropriations bill for fiscal year 1974 directed the
Agency to contract with the National Academy of Sciences for a series of
analytical advisory studies (87 Stat. 482, PL 93-135). EPA and the
Academy agreed upon a program that would respond to the
Congressional intent by exploring two major areas: the process of
acquisition and use of scientific and technical information in
environmental regulatory decision making; and the analysis of selected
current environmental problems. The Academy directed the National
Research Council to formulate an approach to the analytical studies, and
the National Research Council in turn designated the Commission on
Natural Resources as the unit responsible for supervising the program.
The other studies in the series, and a diagram of the structure of the
program are presented on the following pages. Each of the component
studies has issued a report on its findings. Volume I of the series,
Perspectives on Technical Information for Environmental Protection, is the
report of the Steering Committee for Analytical Studies and the
Commission on Natural Resources. It describes in detail the origins of
the program and summarizes and comments on the more detailed
findings and judgments in the other reports.
vii
-------
Components of the NRC Program of Analytical Studies for the
U.S. Environmental Protection Agency
Project
Project
Chairman
Sponsoring Unit of the NRC
Steering Committee for Analytical R. M. Solow
Studies (SCAS)
Environmental Decision Making J. P. Ruina
(CEDM)
Environmental Research Assess- J. M. Neuhold
ment (ERAC)
Environmental Monitoring J. W. Pratt
(SGEM)
Environmental Manpower (CSEM) E. F. Gloyna
Energy and the Environment (CEE) S. I. Auerbach
Pesticide Decision Making W. G. Eden
Multimedium Approach to H. O. Banks
Municipal Sludge Management
Societal Consequences of Trans- W. J. Baumol
portation Noise Abatement
Disposal in the Marine Environment D. S. Gorsline
Review of Management of R. W. Berliner
EPA's Research Activities
Commission on Natural Resources
Environmental Studies Board,
Committee on Public Engineer-
ing Policy
Environmental Studies Board
Committee on National Statistics,
Environmental Studies Board,
Numerical Data Advisory Board
Commission on Human Resources
Board on Energy Studies
Board on Agriculture and
Renewable Resources,
Environmental Studies Board "
Environmental Studies Board
Assembly of Behavioral and
Social Sciences £>
Ocean Affairs Board
Commission on Natural Resources
aln cooperation with the Building Research Advisory Board.
*In cooperation with the Building Research Advisory Board and the Transportation
Research Board.
viii
-------
Structure of the NRC Program of Analytical Studies for the U.S. Environmental Protection Agency
National Research Council
National Academy of Sciences
Commission on
Natural Resources
Steering Committee
for Analytical Studies
*
t I
Committee on
Environmental
Decision Making
Environmental
Research
Assessment
Committee
|
Panel on Sources &
Control Techniques
1
Panel on
Fates of
Pollutants
1
Panel on Effects
of Ambient
Environmental
Quality
1
Panel on
Environmental
Impacts of
Resources
Management
Study Group on
Environmental
Monitoring1
1
Panel on
Ambient
Monitoring
1
Source
Monitoring
1
Panel on
Effects
Monitoring
Committee for
the Study of
Environmental
Manpower2
1
Panel on
Legal Aspects
1
Panel on
Methodology &
National Data
Aspects
1
Panel on Federal
Aspects
1
Panel on State &
Local Aspects
1
Panel on
Industry &
Private Sector
Aspects
I t
Committee on
Energy & the
Environment
|
Panel on Electric
Power (Coal &
Uranium)
1
Panel on
Automotive
Interactions
[Oil & Gas)
Committee on
Pesticide
Decision
Making
1
Panel on
Acquisition &
Use of Data at
Federal & State
1
Panel on Impacts
of Regulatory
Decisions in
the United States
1
Panel on Impacts
of Regulatory
Decisions in
Other Countries
Committee on
Multimedium
Approach to
Sludge
Management
Committee on
Societal
Consequences of
Transportation
Noise Abatement3
Ocean Disposal
Study Steering
Committee
Review Committee
on the
Management of
EPA's Research
Activities
*
^Assembly of Mathematical and Physical Sciences
Assembly ol Behavioral and Social Sciences
-------
Preface
This report is intended to serve as an aid in the formulation of policy in
the control of noise, particularly noise contributed by transportation
vehicles. It is not meant to constitute a piece of original research; rather,
it is intended to assemble from available information and analyses a
relatively systematic and nontechnical overview that lays out the
character of the problem, indicates the extent of the available knowledge
and its gaps, and reviews and evaluates the instruments that can be used
in formulating effective policy.
This report has been designed to satisfy the request of the Environmen-
tal Protection Agency (EPA). It is intended to assist the EPA hi the
formulation and execution of its own programs for the control of noise. It
is hoped that it will also be helpful in the design of future legislation
about noise, both at the national and local levels.
No report on noise can be truly complete; the topic is too vast for a
single volume. Consequently, there are a number of subjects that the
Committee does not include in its discussions.
• There is no survey of EPA's involvement in noise abatement, neither
an assessment of its organization nor of its strategy in addressing the
problems of noise.
• There is no attempt to survey the network of noise abatement
agencies or to describe or analyze the division of labor and law among the
various federal agencies, e.g., the EPA, the Occupational Safety and
XI
-------
Xll
Preface
Health Administration, the Federal Aviation Administration, etc., and
states, regions, and municipalities.
• The report is confined to policy issues for the United States. While it
offers several references to studies in various Western European
countries, it is bound to the current state of affairs in the U.S. by its
premises about the legal basis for policy, the nature of the aircraft fleet,
the mix and type of the automotive vehicles and trucks currently in use,
and other similar factors.
As background, the report discusses the distribution of noise in the
United States, the trends in noise generation and the methods of
measurement of noise. It also offers a long description and assessment of
the methods available for the measurement of noise abatement and
provides some illustrative calculations of its benefits and costs. The report
begins, however, with its central topic, the policy and legal issues in noise
abatement policy.
The work of the Committee was focused on noise produced by
transportation sources. There is evidence that transportation is the major
source of noise in this country as measured in terms of the number of
people annoyed by it, and the sound emanating from the operation of
transportation vehicles has been a prime subject of regulatory concern.
Nevertheless, much of our discussion is applicable also to noise emitted
from other sources, and so a considerable part of our discussion is
concerned with noise in general, not just with transportation noise.
Reports written by committees must all begin with individual contribu-
tions. Each of the chapters in this report was drafted or revised by
particular Committee members or the study director: Chapter 2 was first
prepared by Marcia Gelpe, Chapter 3 by David Green, Chapter 4 by
Kenneth Eldred, Chapter 5 by Edward Morlok, Chapter 6 by Jerome
Singer and William Sampson, Chapter 7 by Arthur De Vany and Jon
Nelson, Chapter 8 by Kenneth Eldred and Jon Nelson, and Chapter 9 by
Arthur De Vany, Jon Nelson, and Alan Walters. I tried to incorporate the
Committee's overall views in the Summary and in Chapter 1.
The Committee wishes to thank members of its staff for their assistance
throughout the long process of research, meetings, writing, and rewriting:
Jerome Singer, the study director; Edward Friedland, senior research
associate; Carol Beers, secretary; and Glenn Davis, research assistant
during a summer internship. Donna Gosnell, the Committee's adminisr
trative secretary, deserves special commendation for her able shepherd-
ing of the report through to its completion. In addition, the Committee
wishes to express its thanks to Eugenia Grohman, of the Assembly
executive office, for her critical and incisive editing. We also wish to
-------
Preface xiii
express our appeciation to the staff of the editorial office of the
Commission on Natural Resources, Robert C. Rooney, Philippa Shep-
herd, and Estelle Miller, for their assistance in the production of this
report.
Last, I must claim a chairman's privilege to express my deep gratitude
to Dr. Singer and my colleagues on the Committee. Their knowledge and
their dedication were indispensable ingredients of the process of report
preparation. Above all, I enjoyed working with them and learned a great
deal in the process. What more can one ask of one's colleagues?
WILLIAM BAUMOL, Chairman
Committee on Appraisal of
Societal Consequences of
Transportation Noise Abatement
-------
Summary and
Recommendations
The Environmental Protection Agency (EPA) estimates (1974) that 16.5
million people live in urban areas of the United States where the outdoor
average sound levels are higher than those that will cause hearing loss in
the long run (over a 40-year period) and that an additional 61.6 million
people live in areas where the outdoor sound levels exceed those causing
annoyance and interference with outdoor activities. The three major
contributors to these noise levels are general urban traffic, freeway traffic,
and aircraft operations. Overall, it is estimated that 75 percent of the U.S.
urban and suburban population live in areas with average outdoor sound
levels above or at the border of annoyance or activity interference. This is
only the most obvious part of the transportation noise problem to which
this report is addressed.
SUMMARY OF THE REPORT
The report is organized around two major topics: the range of alternative
policy measures for transportation noise abatement and the benefits and
costs of abatement. These two topics comprise Parts I and III of the
report; Part II covers the measurement of noise, the current pattern of
transportation noise and its effects, and the projected future pattern of
transportation and the noise associated with it. The Committee's
recommendations are presented in the latter part of this summary
chapter.
-------
2 Summary
POLICY AND LEGAL ISSUES
The Committee believes that transportation noise is a serious problem; it
has health effects potentially leading to permanent hearing loss as well as
significant annoyance and interference effects. A central conclusion of
the Committee is that the abatement of noise requires a federal policy,
but one that can be integrated with state or local programs. We also
conclude that, for effectiveness and given present techniques for
monitoring and enforcement, a federal abatement policy should include
both direct controls and emissions charges.
Legal problems of authorization, responsibility, and mandate must be
considered in policy formulation. For example, the question of whether
emissions charges are to be construed as taxes, fines, or regulations is
complicated and far from settled: the consequences of a judicial decision
on this matter may have significant implications for the effectiveness of a
program of emission charges. Since the desirability of increased reliance
on a policy of emission charges is a topic of considerable concern and
since so much remains to be settled about the legal status of such a policy,
the discussion of the legal issues immediately follows the analysis of
policy.
The choice of a particular instrument of policy for noise abatement is
affected by the circumstances of the emissions. There are four types of
policy instruments available: (1) direct government activities to shield
people from noise or to protect them from its effects, e.g., sound
insulation of schools and hospitals near airports; (2) direct controls
specifying required techniques or processes, e.g., required retrofitting of
muffling devices on engines or prohibition of housing construction very
close to highways or airport runways; (3) direct quantitative controls,
e.g., noise emission limits for trucks and motorcycles; and (4) financial
incentives, e.g., subsidies for relocation of residences near airports or
charges on airplane engines that are proportional to their noise levels.
The report seeks to provide general answers to the questions: What
determines which instruments of control should be used? At what level of
severity or strength should the instruments be used?
Direct Governmental Activities
In comparison with some other areas of environmental protection, the
scope for direct governmental activities in noise abatement programs
seems to be relatively narrow. The government can erect sound barriers,
relocate roads or runways, or build public housing in areas that are
relatively less exposed to noise. But there seems to be nothing, for
-------
Summary 3
example, that plays the central role that waste treatment plants do in
improving the quality of waterways.
Technical Specifications
Direct controls that impose technical specifications are often fairly crude
and inefficient. They do not lend themselves readily to differences in
circumstances, such as the differences from area to area in the ratio of
residential to industrial buildings. However, this type of direct control has
a significant role to play, particularly when effective monitoring is
prohibitively expensive or impractical, since it is not feasible to enforce a
rule that places a limit on emissions when there is no way to determine
whether and by whom the rule is violated. In this case, the reasonable
policy instrument is the imposition of a technical requirement, such as the
installation of muffling equipment.
Quantitative Controls
Direct quantitative controls—i.e., noise emission limits—have one major
advantage over technical specifications: they allow private decision
makers to determine the most efficient way to comply, thus tending to
reduce costs. On the other hand, as with technical specifications, this
approach does not provide incentives to emitters to bring noise levels to
less than the maximum permissible amount. Direct quantitative controls
also require monitoring of emissions.
Financial Incentives
The last class of policy instruments consists of measures to make emitters
pay in proportion to* the noise emitted. By leaving it up to the emitters to
choose their own ways to reduce emissions and, hence, their payments,
they are motivated to reduce emissions by the most efficient means and at
lowest cost. However, financial incentives, like direct quantitative
controls, are practical only if effective monitoring procedures are
available, to permit assessment of the appropriate payment for each
emitter. It should be noted that monitoring is not always easy.
TRANSPORTATION NOISE: ITS MEASUREMENT, SOURCES, AND PROSPECTS
The measurement of noise is a complex endeavor. There are about 100
noise indices in current use, of two basic types: single-event pressure
levels and cumulative measures that sum noise exposure over time. In
-------
4 Summary
spite of what appears to be a bewildering maze of indices, however, we
find that there is a high intercorrelation between measures. Some
characteristics of sound that are not well described by any index of
measurement—for example, whether bursts of noise are randomly spaced
or occur at periodic regular intervals—and the high intercorrelation
between measures suggests that an attribute missing from one measure is
not likely to be captured by any of the other measures.
The effects of transportation noise in the United States are examined in
several ways in this report. First, noise from all sources is considered for
the extent to which it is a source of annoyance or complaints.
Transportation sources are by far those most heavily implicated by
neighborhood residents: motor vehicles, for example, are cited 55 percent
of the time. Second, noise sources are examined to determine the number
of people affected by each and the magnitude of that effect. Noise from
general urban traffic, from freeways, and from aircraft operations affect
the largest numbers of people. On an energy basis, the number of
kilowatt-hours per day equivalent to the noise emitted by medium and
heavy trucks and by aircraft operations is 60 percent greater than the
energy equivalent of 30 other common sources combined.
The projections for future transportation operations and mixture of
vehicles and their associated noise production indicate that increases in
transportation activities—even with each vehicle at its current noise
emission level—would have only a minor effect on noise levels. The
decibel scale is logarithmic, and doubling of the sound pressure level at
any point results in a 3-decibel (dB) increase in sound: if a single
motorcycle's noise is 80 dB, the noise from two such motorcycles is 83 dB.
Thus, if all transportation activities were doubled with existing vehicles
and facilities, only a 3-dB increase in general environmental noise levels
would result. (It would take a 10-dB increase for the sound to be
perceived as doubled). Since new vehicles, cars, trucks, and aircraft, are
quieter than those they replace, it is likely that overall transportation
noise will remain relatively constant, even with increased operations.
Conversely, cutting all transportation activities in half would result in
only a 3-dB reduction in overall noise levels. The implication is clear that
any program for abating transportation noise will have to do so by
quieting sources, insulating receivers, using barriers and the like, rather
than by simply reducing operations.
-------
Summary J
BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
Benefits of Noise Abatement
The report examines some of the health effects, including hearing,
cardiovascular, mental health, and other health benefits of the abatement
of transportation noise. The clearest and most obvious health benefit of
abatement is reduction in hearing loss. While it is difficult to implicate
transportation as a separate cause of hearing loss when other factors—
aging and industrial and general environmental sounds—are also
involved, the report concludes that a reduction in transportation noise
may contribute to a significant reduction in hearing loss.
The benefits to welfare of noise abatement constitute a broader range
of categories than the benefits to health. They include the economic
benefits of more efficient systems, of saving resources, and of productivity
increases from less noise. These occur directly when noise ceases to
interfere with work and indirectly when motivation and morale are
improved in a quieter workplace. The report summarizes the known
effects and probable abatement benefits—and the limited information—
on the social effects of noise and on annoyance and quality of life.
Monetary Measures of the Benefits of Abatement:
Property-Value Analysis
While the informal weighing of benefits and costs has probably always
been conducted by policy makers, the analytic economic technique of
cost-benefit analysis requires that costs and benefits be described in
commensurate units. Since the costs of noise abatement programs are
usually estimated in monetary terms and the benefits usually are not, a
model for the monetary estimation of the benefits of noise abatement is
needed. The one used is based on statistical evaluation of the conse-
quences of noise for real estate prices (the property-value model), which
indicates how much individuals are willing to pay to avoid noise. The
report examines the use of the property-value model in general and in a
number of specific studies and also discusses a number of related issues.
For example, noise at a particular location often comes from several
sources—trains, cars, airplanes, etc.—but studies of particular abatement
options are usually addressed to one source at a time and rarely consider
their relation to the abatement options for other sources, which would be
necessary to determine the cost-effectiveness of an abatement program
for noise from all sources.
-------
6 Summary
Costs of Noise Abatement
In considering the costs of abatement, the report concentrates on the two
major contributors of noise: commercial aircraft and motor vehicles. For
aircraft, such factors as the retirement rate of the current, relatively noisy
fleet, the introduction of quieter new aircraft, the development of new
technology, and the probable effect of increasing operations are analyzed.
For motor vehicles, six categories are considered: autos, motorcycles,
buses, and light, medium, and heavy trucks. For motor vehicles, factors
such as production costs, fuel economy, maintenance costs, and mode of
operation are analyzed.
The cost estimates are based on abatement programs involving
reduction of emissions and, to a lesser extent, those involving shielding of
recipients. The costs are subdivided by the degree of abatement desired
and the speed with which the abatement is to be carried out. More
stringent reductions in noise levels entail higher costs, and the relation-
ship between the degree of reduction and its cost is not linear. On the
contrary, the report concludes that costs accelerate as the amount of
noise abatement increases.
Cost-Benefit Analysis: Some Illustrations
The report concludes with some cost-benefit calculations. These calcula-
tions do not provide any basis for evaluation of the desirability of any of
the specific abatement proposals—such as retrofitting and two-segment
landings—currently under discussion. Their purpose is modest—to
illustrate the current state of the art and to indicate the limited degree to
which they can assist the decision process.
RECOMMENDATIONS
The Committee's recommendations are grouped into five categories:
emission charges and direct controls, monitoring, federal policy and
coordination, research, and cost-benefit considerations. The groupings
are merely for the reader's convenience and some recommendations
could have been placed, equally appropriately, in a different group.
Following each recommendation, the chapter that contains the support-
ing discussion and data is indicated. For many of the recommendations,
there is pertinent information in more than one chapter; in these cases,
more than one chapter is indicated, with the primary one listed first.
-------
Summary 7
EMISSION CHARGES AND DIRECT CONTROLS
1. Some regulatory mechanisms are relatively self-enforcing; others
require considerable enforcement activity to be effective. Provision of
adequate funds and enforcement personnel is necessary for any abate-
ment program, but is crucial if regulations of the latter type are to have
anything beyond moral force. Enforcement funds adequate to assure
general compliance should be provided on a continuing basis (Chapters
1,3).
2. In the choice between direct controls and emission charges, noise
abatement policy has, until now, deprived itself of what many analysts
consider to be a valuable and powerful tool by its exclusive reliance on
the former. If effective monitoring of the sounds emitted by an individual
source is practical, the use of charges may offer substantial benefits in
effectiveness, efficiency, and reliability. Accordingly, we recommend that
a substantial role should be considered for emission charges in cases for
which monitoring of individual sources is practical: for example, in the
control of airport noise. This approach is also a promising instrument for
the control of noise emitted by trucks and major construction projects,
and study of the use of emission charges for the control of noise emitted
by these sources should be undertaken without delay. On the other hand,
largely because of the difficulty of monitoring emissions from the
individual sources that contribute to the overall urban noise level and to
noise along highways, it is preferable, at least in the immediate future, for
programs designed to deal with these important noise problems to rely on
direct controls (Chapter 1).
3. It would be highly desirable to carry out one or more carefully
designed and monitored experiments to test and to document the
effectiveness of a system of emission charges. This experiment should, if
necessary, be authorized by explicit Congressional action and should
permit the system of charges to be confined for experimental purposes
(with suitable compensation, if necessary) to some limited geographic
areas, say to some particular airports (Chapters 1 and 2).
MONITORING
4. The effectiveness of noise abatement programs that rely on emission
charges depends on the availability of reliable and economical methods
of source monitoring. Consequently, it is a matter of priority to provide
means for the financing of research on practical source monitoring
techniques, particularly for cases in which sound is contributed by a large
-------
8 Summary
number of very mobile sources whose emissions vary with time and with
mode of operation (Chapters 1 and 3).
5. Single-number noise indices (such as day-night sound level and
equivalent sound level) appear to be effective in monitoring the overall
noise level at a given time and location and are useful for an assessment
of its effects. Once computed, however, these indices do not help in
identifying the sources of individual intrusive events. As a result, some
other measurement procedures will have to be used if the system of
monitoring is to permit an effective set of direct controls based on source
emission standards (emission quotas) or a system of emission charges;
each of these regulatory procedures requires information on emissions by
the individual regulated source (Chapters 3 and 1).
6. The Committee recommends that some provision be made for
monitoring the effects of regulation. This includes the gathering by a
central agency of current information on the enforcement of regulation at
federal, state, and municipal levels as well as information on fines,
emission charges, compensation awards, easement purchases, and other
types of regulatory effects. This information should then be organized
and made available to the public. The Committee believes that such data
are essential for the evaluation of the efficacy of noise regulation and
abatement policies. Publication of the results, using some medium such as
the Federal Register, is recommended (Chapter 1).
FEDERAL POLICY AND COORDINATION
7. In the choice between federal policy and local option in noise
control, there are grounds for favoring a federal program with as much
scope for variation by geographic location as effective administration
permits. Without a program designed and administered under federal
supervision, there may, except in a few isolated localities, be no effective
noise abatement. However, a federal program should, as far as is possible,
avoid the imposition of uniform and inflexible standards that disregard
local differences in needs and preferences, and it should, wherever
practical, offer opportunities for genuine local choice by permitting local
design of programs meeting federal regulations and subject to federal
review (Chapter I).
8. The Committee recommends that special provision be made to
disseminate among local control agencies information about the design,
administration, and effectiveness of various noise abatement procedures.
States and municipalities are unlikely to possess the resources or
personnel for research and design relating to abatement procedures. If,
for example, emission charges are adopted by local governments,
-------
Summary 9
considerable care will be required in structuring and in clearly explaining
the charges to minimize the danger of legal difficulties. The Committee
suggests that sample regulations for procedures such as emission charges,
as well as techniques for monitoring, data processing, setting of rates, and
the like, be distributed to state and municipal agencies (Chapter 2).
RESEARCH
9. Responsibility for research on noise is divided among several
agencies, among them the Department of Transportation, the Federal
Aviation Administration, and the Department of Housing and Urban
Development. Since evaluation of each of their noise abatement
programs depends upon the specification of each source's contribution to
the total noise, we recommend that EPA, as assigned in the Noise Control
Act, Section 4(c), coordinate governmental noise research. In order to
specify the relevant components and to permit effective assessment of the
entire set of programs, however, a mechanism for the establishment of
priorities and for program control is needed (Chapter 7).
10. Although there is a large body of knowledge on the effects of noise
it is, of course, by no means complete. However, the absence of definitive
information on noise effects is not in itself a sufficient basis to reject a
proposed regulation designed to avoid the chance of such effects. In
deciding whether there is justification for intervention in an area in which
there is strong reason to suspect detrimental effects, but the presence of
such effects is not fully proven, one should consider both the probability
that the effects will occur and the seriousness of the effects if they do
occur (Chapter 6).
11. The evaluation of the benefits of transportation noise abatement
programs, given the present state of knowledge, relies heavily on
statistical evidence showing the effects of noise on real estate values. This
is virtually the only source of systematic evidence that yields a monetary
figure constituting an overall evaluation of the benefits of noise
abatement that is directly commensurate with the costs of abatement.
This approach is fundamentally valid, though it is subject to a number of
sources of error and bias for which explicit correction must be made, to
whatever extent the available evidence permits. At present, these errors
are significant and reduce confidence in the results. Consequently, it is
important that research be carried out to help in the design and testing of
alternative or complementary methods of evaluation of the benefits of
abatement, with survey research approaches perhaps constituting the
most promising of these alternative methods (Chapters 7 and 6).
12. Additional research on the effects of noise on public health and
-------
10 Summary
welfare will apparently cost very little relative to current expenditures on
noise control and can provide significant additions to the information on
which future noise policy can be based. EPA is the only federal agency
that now is assigned responsibility for protection of the general public's
health and welfare from the effects of noise. EPA should consequently be
provided the funds needed to conduct additional research on the effects
of noise on public health and welfare, with emphasis on hearing loss,
other physiological effects, annoyance, and other social consequences
(Chapter 6).
13. As indicated in EPA's 1972 Report to Congress (U.S. Congress,
Senate, 1972), there are a number of sources of noise about whose extent,
intensity, and duration little is known. If all the major sources of noise at
a particular point and time were known, any proposed abatement
program for one source could be evaluated after ascertaining the prior
effects of all the more cost-effective alternatives for abating the other
sources. We recommend investigation of the characteristics—extent,
intensity, and duration—of the noise of currently unregulated sources in
order to facilitate these evaluations (Chapters 7 and 8).
COST-BENEFIT CONSIDERATIONS
14. In designing programs or regulations for noise abatement, it is
essential that costs be taken fully into account. At the local level, there
seems to be some propensity to adopt noise control regulations regardless
of the resources that must be used in carrying them out. It must never be
forgotten that resources used for noise abatement become unavailable for
the construction of hospitals, schools, improved housing, or for other
programs of high priority for the social welfare. Costs must never be
disregarded in the design of noise abatement programs (Chapter 8).
15. The Committee recommends that any evaluation of the costs and
benefits of a proposal for noise abatement should explicitly consider their
distributive effects. It is not sufficient to determine that, in the aggregate,
benefits exceed costs or vice versa; rather, the distribution of costs and
benefits among different groups must also be evaluated. The identity of
the groups that will bear the cost of the abatement programs and of those
that will reap their benefits is a crucial issue. An extreme example is a
program that, despite a favorable benefit-cost relationship, would reduce
only the noise heard by the wealthy and would levy all costs on the poor.
Analyses of the distributive consequences of proposed policies
sufficiently detailed to indicate possible inequities should be a part of
cost-benefit analysis (Chapters 1 and 9).
16. In a benefit-cost analysis, it is not legitimate to treat the size of the
-------
Summary 11
benefit-cost ratio as an index of the desirability of the program under
consideration. A program with a 2.6 benefit-cost ratio is not necessarily
superior to one whose benefit-cost ratio is 1.9. Rather, the appropriate
criterion is the selection of a combination of programs, the discounted
present value of whose expected net benefits (benefits minus costs) is as
great as possible (Chapters 9 and 7).
-------
I
POLICY AND
LEGAL ISSUES
-------
The Choice of
Policy Instruments
INTRODUCTION
The central objective of this report is to be helpful in the design and
execution of abatement policy for transportation noise. There are two
fundamental issues: What is the appropriate level of severity (strength) of
the abatement measures to be undertaken? What instruments of control
should be used in carrying out these measures?
On the first issue there is obviously a great deal of choice. At one
extreme, one could undertake absolutely no restriction of sound
emissions, letting anyone produce an unlimited amount of transportation
noise, without hindrance. At the other extreme, transportation noise
could be reduced to zero by bringing all transportation to a standstill.
Clearly, neither of these extremes is desirable or even practicable. The
issue, then, is to find what intermediate point best serves the public
interest.
The essence of the issue is that increased restriction of sound emission
is not free. It imposes a very real cost on the community. By this we mean
that the cost is not merely a matter of dollars, to which one may assign
secondary importance in comparison to the effects of noise on health and
human stress; rather, the costs take the form of inhibition of other vital
activities. An obvious one is transportation activity itself, as has just been
noted. As control of noise becomes increasingly severe, the cost of
transportation can also be expected to grow, meaning that the economy
will find itself paying more for this vital service and possibly obtaining
15
-------
16 POLICY AND LEGAL ISSUES
less of it. Another cost of increased restriction of sound emissions—which
is a bit less obvious but not less important—is the use of resources in the
process, in building quieter engines, in insulating dwellings against noise,
and so forth. All resources used in this way become unavailable for other
social purposes—building hospitals and schools, eliminating slums, etc.
Thus there is a real and unavoidable trade-off that is implicit in any
decision to strengthen restrictions against noise.
If policy makers ignore this trade-off, they may impose a degree of
reduction in noise that, while desirable in itself, imposes social costs that
are greater than the gains. Or, even if a given program does produce a
positive net benefit, it may be that a slightly less severe program will yield
net benefits that are higher still, all effects considered. In either case,
because the trade-off in terms of costs and benefits has not been
considered properly, society will find its interests poorly served.
It may be noted that, characteristically in environmental programs, the
terms of the trade-off become increasingly unfavorable as the program
grows increasingly severe. One finds typically that, say, a 10-percent
reduction in emissions can be achieved at negligible cost and a second
decrease of 10 percent is apt to cost very little more, but that by the time
one gets to an 85-percent reduction (in total) yet another 10-percent
decrease is prohibitively expensive, while going from a 90- to a 100-
percent reduction (i.e., total elimination of the emission) is for all
practical purposes impossible.
In sum, because of the trade-off between costs and benefits, it is not
true that a stricter (more effective) noise abatement program is always to
be preferred to one that is less restrictive. The objective is to find the point
at which further tightening of the restrictions is no longer beneficial from
the point of view of the public interest.
In trying to help policy makers achieve that objective, this chapter
examines four major questions: Is there a need for restrictive interven-
tion? If so, what criteria should be used in deciding on the degree of noise
restriction to be achieved? What role should be left to local option as
against uniform national policies? (By "local" here we mean state and
regional as well as municipal, strictly defined.) What instruments of
control should be used: that is, what role should be played by direct
controls rather than financial incentives or some other approach?
IS INTERVENTION TO REDUCE
NOISE JUSTIFIED?
It has been argued that there really is no defensible justification for
governmental regulation of noise. First, it can be argued that aside from
-------
Choice of Policy Instruments 17
outright physiological damage, the undesirability of noise is so much a
matter of personal preference and cultural conditioning that the decision
to require reduction in noise emissions amounts to the arbitrary
imposition on others of the preferences of the group that prefers quiet. If
someone likes noisy motorcycles or loud music, what right have others to
require a reduction in the noise to which that person is exposed? Second,
it is sometimes argued that differences in the level of noise in different
areas really give people all the choice they need. If they have the
necessary incomes, they can live in neighborhoods that are noisy or quiet
as they prefer. Indeed, since rents and land values are reduced by noise,
as has been documented amply, the market mechanism, if it is working
properly, automatically provides financial compensation to those who are
willing to live in noisy neighborhoods. Therefore, why is it appropriate to
intervene and force people to accept noise levels lower than those they are
willing to live with, in the process undoubtedly forcing their rents
upward?
The answer is that, for noise, the market mechanism does not give
people what they really want. There is an inherent bias in the pricing
arrangements that forces people to accept levels of noise higher than they
themselves would select taking into account all of the pertinent costs and
benefits.
Under the market mechanism, there will be overexpansion (in terms of
benefits and costs) of any activity for which the user escapes payment in
whole or in part. If individuals do not pay for the water they use, for
example, they tend to let taps run freely. A firm that does not pay for
water will rarely if ever recirculate it if recirculation incurs any costs. The
market mechanism does not work because the water-using activities
impose a cost on the community, but the user of the water does not bear
the full share of that cost.
Exactly the same issue arises in noise generation. Suppose, to make the
argument more specific, that we can measure the social cost of noise
precisely—say, that every run of a noisy truck through the center of the
city causes $50 in noise damage. Obviously, it makes a great deal of
difference to the economics of trucking if the firm that supplies the
transportation is forced to pay that $50 or if the cost is borne by those
who suffer the damage. If the truck firm is forced to pay the full cost of
operation, including the $50 cost of noise damage, trucking prices will be
raised, the demand for truck transportation is likely to be reduced, and
the demand for substitute means of freight transportation and for quiet
trucks will be stimulated. Ah1 these effects will result in a quieter city, not
because someone has decided that this should be so, but because truckers
must pay the costs that the noise of their activities generates. In other
-------
18 POLICY AND LEGAL ISSUES
words, taking it to be a demonstrated fact that noise does involve some
social cost, it is clear that if those who cause the cost are forced to pay for
it, there will be a quieter environment.
This analysis is not undermined by the argument that individuals can
now select their ambient noise level through choice of residential
communities. If those who generate noise do not pay its social cost, the
level of noise everywhere will be increased. Thus, with higher noise levels,
even people who now live in noisy neighborhoods will generally have less
quiet than they would have had otherwise. People in (the now scarcer)
quiet neighborhoods will have to pay rents that are higher than they
would have paid otherwise. People will pay for quiet with money they
would use to pay for other things if there were less noise overall.
Everyone is likely to lose in the process, except for the noise emitters who
are permitted to escape the costs of their activities or the buyers of their
products who also escape the social costs of their consumption. The
choice of type of neighborhood in which to reside, to the extent that there
really is freedom in this choice, merely permits people to divide up the
burden of the excessive noise; it does not cause that excessive noise to
diminish.
This standard economic analysis implies that there will be damage to
the interests of society, as measured in terms of the preferences of
individuals themselves, whenever those who carry out some activity do
not themselves bear its costs but shift them to others. So long as the social
costs of noise are not borne by those who generate it, noise levels will
necessarily be excessive from the viewpoint of the affected public, and
some noise abatement measures will definitely serve the public interest.
CRITERIA FOR DECIDING HOW MUCH
NOISE REDUCTION IS APPROPRIATE
While it follows that some decrease in noise will generate greater benefits
than costs, the critical issue for policy is the degree to which it is
appropriate to restrict noise emissions. The "best" policy, by definition,
involves a balancing of the benefits of noise reduction against the social
costs that must be incurred in achieving these reductions. It also involves
consideration of who receives the benefits and who pays the costs.
Resources used in producing retrofitting devices become unavailable for
the construction of schools or hospitals or housing, and this is part of the
true social cost of any abatement problem. The balancing of costs and
benefits of noise abatement is equivalent to the allocation of resources
among competing uses, all of which offer benefits to society.
Thus, the optimal degree of noise abatement can only be decided after
-------
Choice of Policy Instruments 19
one determines to an acceptable degree of accuracy the magnitudes of all
the pertinent costs and benefits. As the discussion in later chapters shows,
the state-of-the-art is still far from the point at which it can yield clear-cut
figures.
VOLUNTARY VS. INVOLUNTARY SUBJECTION TO SOUND
We can, however, enunciate several general propositions about the sorts
of restrictions on noise levels that are indefensible in principle.
For this purpose, one must distinguish between voluntary and
involuntary subjection to noise. A person who attends a rock concert or
sits close to the tympany section of an orchestra in a performance of a
Wagner opera may be subjected to higher noise levels than someone who
lives 500 yards from an operating sledgehammer. But the concertgoer
chooses voluntarily to be subjected to the sound level while the victim of
construction noise does not.
There is a general principle that applies to this distinction: whenever all
of the individuals who hear a noise choose to subject themselves to it
voluntarily, the generation of that noise is deemed to cause no net social
costs. That is, there is no cost that the noise generator shifts to others; one
cannot use the analysis of the preceding section to argue that excessive
amounts of noise will be generated.
Even in such cases, however, policy makers may decide to impose some
restrictions on sound levels in order to protect those affected by it. Just as
the law discourages cigarette smoking, prohibits the use of narcotics, and
requires the installation of safety belts in automobiles, it may be
considered appropriate to protect people from the hearing loss that
frequent attendance at rock concerts may cause. But these restrictions are
either (1) designed to protect society (i.e., those who do not hear the
noise) from costs that the hearing of noise by the others entails (e.g.,
public hospital care of people who go deaf from rock concerts), or (2)
they are an act of paternalism in which the government in effect decides
that it knows better than the concert goers what is good for them. This
applies only to choice that is purely voluntary.
PHYSIOLOGICAL AND PSYCHOLOGICAL DAMAGE
VS. ANNOYANCE EFFECTS
A second distinction relevant to the appropriate level of abatement
measures is the difference between noise whose effects are primarily
annoying and noises that produce demonstrable physiological or
psychological damage. It is not necessarily true, however, that the
-------
20 POLICY AND LEGAL ISSUES
physiological and psychological damage must always be considered more
serious and more unacceptable than annoyance. Mild and perhaps
temporary hearing loss may be considered less serious than the
persistence of noise that makes life extremely unpleasant, though neither
physical nor psychological damage can be traced to it.
Indeed, the appropriate policy measures in the presence of either
physical and psychological damage or annoyance is always a matter of
balance between costs and benefits, and when all of the required
information is available, exactly the same principles apply in both cases.
A significant difference may seem to arise only if facts are uncertain-
potential harm has been identified but not proven. We reject the notion
that the absence of evidence of more than potential harm should preclude
all regulatory action. Conclusive proof cannot be expected if research is
required to show that delayed effects can be attributed to earlier causes
and if researchers are not allowed to conduct controlled laboratory
experiments that may subject people to dangerous sound emissions.
Damage that is unproven but for which there is good reason to suspect
must be considered in reaching a regulatory decision, taking into account
both the probability of the harm and the magnitude of the threatened
damage. The larger the product of the probability and the magnitude, the
greater the abatement costs it may be rational to incur. If the magnitude
of the threatened damage is sufficiently large, the probability itself need
not be one or very close to one in order for regulatory measures to be
justified.
LACK OF THRESHOLDS FOR NOISE DAMAGE
It would be helpful, in dealing with problems of physiological damage, if
there were one or more identifiable threshold levels of sound at which
harmful effects to most people first become serious or become increasing-
ly so. For example, if it could be shown that any sound less than 90
decibels produces no hearing loss but that any increase in sound level
beyond 90 decibels brings with it frequent and protracted hearing loss to
many people, this would immediately indicate what sound levels
probably should be prohibited.
Unfortunately, the evidence indicates that matters are by no means so
simple. First, since sound is multivariate in character, the damage done
by sound is probably affected by pitch, duration, variability, and a
number of other characteristics in addition to its intensity. Second, the
magnitude of the resulting damage will vary with individuals, their age,
their physical constitution, and so on. Finally, there seems to be no
conclusive evidence that the damage effects of sound vary in a sequence
-------
Choice of Policy Instruments 21
of steps rather than increasing more or less gradually with sound energy
levels. In short, the existence of convenient thresholds is, at least on the
basis of the evidence currently available, doubtful at best. Accordingly,
the assumption that convenient thresholds applicable to the entire
population exist is likely not to be helpful in the formulation of policy
and, at worst, can lead to the setting of inflexible standards whose rigidity
cannot be justified by the facts.
NATIONAL CRITERIA VS. LOCAL OPTION
A third issue, which, like the others we have been discussing, affects all
environmental policy, is the degree to which it is desirable to permit each
community to set its own standards. Should the federal government
determine goals that apply uniformly throughout the country, or should
each community decide for itself the appropriate trade-off between noise
and abatement cost?
In favor of local option are the significant differences in local
conditions and the differences in the preferences of local populations. In
the center of a densely populated metropolitan area, one simply cannot
hope to achieve the degree of quiet of a remote country area. Different
ethnic groups may differ in their attitude toward noise. People engaged in
different activities may differ in their sensitivity to noise; a sound that is
practically unnoticeable in a factory may severely disturb a hospital or a
wilderness area. Moreover, even if people have the same preferences, the
cost of noise abatement will cause variations in choices from one income
group to another, just as it affects choice of vacation spot and type of
clothing. The availability of a variety of communities that differ in levels
of noise, with offsetting differences in rents and tax rates will, all other
things equal, broaden the range of available choices. This, along with
general distrust of an omnipotent central government, is the basic case
for local control of noise emissions. There are, however, several counter
arguments of comparable persuasiveness.
The first rests on a denial that there really exists the freedom of
informed choice required by the preceding argument. If zoning require-
ments, social pressures, and a variety of other impediments effectively
prevent the poor or families with small children or members of minority
groups from moving into neighborhoods whose quiet/rent level combina-
tion they really prefer, then the assertion that a greater variety of
situations broadens the range of choice loses its validity. The problem is
compounded by lack of information about the full effects of noise. In
short, lack of freedom to move may effectively undermine the broadening
of choices that variation in noise levels from one community to another is
-------
22 POLICY AND LEGAL ISSUES
said to make possible. As a result, some sort of uniformity in a national
program, with all of its inflexibility, may prove to be the lesser evil.
There is a second reason for federal control of noise abatement.
Competition for industry and jobs among local governments may all but
prevent any effective control of noise generated by economic activities
whose products are sold in a national market. Since noise abatement
measures are costly, firms that operate in an area whose noise program is
strong are afraid that they will find themselves at a competitive
disadvantage compared to firms located in a jurisdiction whose program
is weak or nonexistent. An airport whose noise standards are weak may
perhaps be able to lure business from another airport with an equally
convenient location but with strong abatement requirements. Any local
government is afraid of driving its industry and its job opportunities into
the arms of another and so none may be willing to make the first move
towards the adoption of measures for the control of noise. Since each area
will be reluctant to make itself unattractive to industry, if the matter is left
to local option we may end up, for all practical purposes, with no
abatement at all.
This problem has certainly proved a serious stumbling block for local
management of controls over air and water pollution. However, the
likelihood of industries fleeing from areas with strong noise measures may
perhaps be smaller than in the case of air and water programs because a
high proportion of noise is generated by activities that are fairly
immobile. Construction and transportation may not relocate as readily as
a paper mill or a chemical plant because the services provided by the
former must to a considerable degree be produced in the area where they
are consumed.
A third disadvantage of local control is that it may increase the cost of
gathering the information needed to design a noise regulation program.
Local controls require acquisition of the relevant information by all the
local governments, rather than just the federal government. While some
of this duplication may be avoided by having the information generated
at one place and provided to agencies at other levels of government, this
entails communication costs and at least some review by local govern-
ments. On the other hand, some of the requisite information may apply
only to the locality in question, and here the cost to the local authority
may be lower than that to a federal agency.
A fourth argument for federal regulation arises from the mobility of
some major sources of transportation noise, such as large trucks and
other sound emitters that move in interstate commerce. If controls on
noise emissions of such sources differ from state to state, the sources will
-------
Choice of Policy Instruments 23
be required to comply with the controls of the strictest state. Such
controls will then effectively become national controls. It may be
inappropriate to have such stringent controls imposed everywhere
throughout the country. In these cases, national controls may be
preferable to local controls and, in any event, there may be Constitutional
questions about the legal right of individual states and localities to
exercise such controls.
The conclusion from all this is that the choice is by no means open and
shut. Neither local option nor complete federal control is completely
unobjectionable. As a compromise, one can seek the adoption of a federal
program with as much flexibility built into it as effective administration
can permit.
There are two ways in which flexibility can be built into a federal
program. First, goals can be varied systematically on the basis of a
formula that sets standards that depend on a number of variables, such as
density of population, density of industrial and commercial activity, etc.
Alternatively, particularly where uniformity is not crucial, the choice of
criteria, perhaps within specified bounds, can be left to local option. In
this case, however, it is essential that there be some attention to
procedures that offer an effective voice to all residents in any area, not
merely to a few groups with particular economic or political influence.
POLICY INSTRUMENTS: DIRECT CONTROLS
AND FINANCIAL INCENTIVES
DIRECT CONTROLS
Most environmental programs that are now in effect use one of two types
of direct control as their main policy instrument. The first type of direct
control can be described as process or equipment specification: for
example, the requirement that railroad tracks be welded or the prohibi-
tion of certain types of traffic in the neighborhood of hospitals. This type
of direct control specifies in detail some action or process that is either
prohibited or required.
The second and somewhat more flexible type of direct control can be
called an assigned emission limitation or performance standard, which
imposes a quantitative ceiling on emissions by any given source, leaving it
to the emitters to decide how to satisfy the standard. So long as the
emissions do not exceed the assigned ceiling, the emitter is taken to have
complied with the regulation. (The distinction between these two types of
direct control, which is significant for policy, is discussed below.)
-------
24 POLICY AND LEGAL ISSUES
FINANCIAL INCENTIVES
While legislators and administrators have generally shown a predilection
for direct controls, most economists have argued that another approach
to regulation of environmental damage is generally superior. This is the
use of financial incentives in the form of emissions charges that require
the generators of environmental damage to bear the social costs of their
activities. The principle is straightforward: since the ability of the
emitters to escape the social costs of their emissions is a central cause of
activity that is socially undesirable because of its effects on the
environment, the way to deal with the problem is to make those
responsible for those costs bear their burden.
It should be emphasized that, as with direct controls, the basic purpose
of a system of noise emission charges is reduction in noise levels, not
punishment of the emitters of noise. The idea is to achieve reductions in
noise by making it attractive financially for emitters to take appropriate
abatement measures or, rather, by making it financially unattractive for
them to fail to do so. Any payments by them or any receipts by the public
treasury are incidental; a system of emission charges will be successful
only if it succeeds in inducing a substantial reduction in emissions.
It should be noted that a system of charges, if set at appropriate levels,
can always achieve its purpose. As an early Supreme Court stated, "The
power to tax constitutes the power to destroy . . . ." Emissions charges
are not taxes, and it is not their object to destroy any economic activities,
but a non-token charge can always be set at a level that achieves whatever
degree of noise abatement is desired. The issue is not whether a system of
charges can reduce noise, but whether they are in any circumstances the
most effective way to do the job.
The levels of such charges can be determined in one of two ways: they
can be made equal to the best available estimate of the monetary value of
the social damage caused by a unit increase in emissions (which, if
satisfactory information is available, is considered by economists to be
the best approach); or they can be formulated in terms of a set of target
standards for source emissions or environmental sound level, with the
charges selected, on the basis of the statistical evidence, sufficiently high
to induce the reductions in emissions necessary to achieve those
standards. These target standards can be determined either on the basis
of some evidence of harm or on the basis of some evaluation of the
requirements of effective control.
It should be noted that, to be effective, a system of charges must apply
to governmental as well as to private emitters. If the truck fleet run by a
state or a federal agency is responsible for unacceptable amounts of
-------
Choice of Policy Instruments 25
noise, an appropriate payment should be taken from its budget as an
inducement to take appropriate abatement measures. Exemption of any
class of emitters from any type of environmental program is obviously
likely to impede its effectiveness and efficiency although there is neither a
legal nor a logical barrier to the establishment of an emissions charge
plan with one or more classes of emitters excluded.
THE CHOICE BETWEEN DIRECT CONTROLS
AND FINANCIAL INCENTIVES
The authors of this report take a position intermediate between the two
views that direct controls should never be used and the belief that they
are the solution to all noise problems. We conclude that policy makers
have gone much too far hi the universality of their rejection of financial
incentives and have denied themselves a set of potentially powerful and
efficient tools that can make valuable contributions to environmental
policy. On the other hand, a large number and variety of emissions,
particularly in the case of noise, are best dealt with by direct controls
rather than by emissions charges, at least given the present state of
knowledge and technology. In the remainder of this chapter we will
discuss the virtues of each of these basic approaches and suggest in broad
terms which types of sources of noise are best controlled by which
method.
In confining our discussion to the three types of instruments—the two
types of direct control and the use of charges for noise emissions as a
financial inducement for abatement, we do not intend to imply that these
are the only instruments that have been proposed or discussed. A variety
of other tools, such as subsidies for insulation or construction of sound
barriers and the auctioning of sound emission permits, have also been
used or suggested. However, we believe that in practice the major
contenders for a primary role in noise abatement policy are the
instruments discussed here, and we will therefore do no more than refer
to the other policy instruments.
Desirable Attributes of Emission Charges
The grounds on which economists have argued the superiority of a system
of charges for environmental damage over reliance on direct regulation
and enforcement through resort to the judicial system for fines or orders
in each individual case have been presented so often that a brief summary
will suffice for this report. First, economic incentives in the form of
charges for environmental damage will prove more effective in both the
-------
26 POLICY AND LEGAL ISSUES
long run and the short. Although they are not, legally speaking, a tax and
are not collected through the tax system, charges will be collected on a
regular basis just as taxes are. Of course, resort to the courts will be
necessary to compel payment by those who do not comply with the
system, so enforcement can at least initially turn out to be as burdensome
as enforcement under direct controls, though it is hard to believe that it
will continue to be so once the charges have been tested in the courts and
have become routine.
Second, emission charges (like taxes), once effectively in operation,
tend to be self perpetuating. Direct control systems depend more on
repeated initiation of enforcement actions by the regulator. Therefore, a
charge system should continue to work even if an environmental issue
disappears temporarily from the headlines.
Third, economists believe that a system of emission charges promotes
efficiency in an environmental program. That is, it induces a pattern of
abatement by individual emitters who comply with a program that
improves the environment at as low a cost as is practicably achievable. It
provides the largest financial incentive for abatement measures to those
emitters who can abate most cheaply and efficiently and can therefore
avoid the charge at lowest cost to themselves. This contrasts with most
programs of direct controls, which usually try to apportion the task of
abatement among emitters in a manner that is considered fair, rather
than attempting to assign them on the basis of the relative costs of the
abatement measures by different emitters.
Fourth, regulation through a system of charges for environmental
damage involves minimal interference in the freedom of choice of
individuals, not dictating how they must operate or what technical
processes they must adopt. Rather, it uses the price system to stimulate
their effort and ingenuity to achieve the objective of environmental
protection. This also avoids locking firms into one technology dictated by
regulation, which may later prove to be inefficient for particular firms or
plants.
Fifth, and perhaps most important, a program of charges does not
transform normal economic behavior into a matter to be dealt with by
criminal penalties. Although high levels of noise emission may be
dangerous, there is nothing inherently criminal about them. Rather, it is a
part of everyday economic activity, which may, however, involve a
misuse of society's resources. Just as a trucking firm is expected to pay for
its labor and its fuel, it is appropriate for it to pay for the insulation and
medical costs its activity imposes on others. Emission charges force
manufacturers to take cognizance of the social cost of their products (and
emissions) along with the cost of production.
-------
Choice of Policy Instruments 27
Sixth, a system of emission charges can potentially provide some
revenue to the government and hence perhaps reduce to some degree its
need to raise revenue by other means. This is in marked contrast to other
instruments of noise control policy, which generally require substantial
increases in government expenditures. However, this point should not be
overemphasized. Emission charges are not intended as a revenue-raising
device, and to the extent that they are successful in reducing noise
emissions, the payment of charges by emitters will be reduced.
Two other issues about a system of emissions charges require
comment. It is often asked whether such payments by emitters will
ultimately fall proportionately more heavily on the rich or the poor. The
answer is that no one really knows. For example, it may be surmised that
if emission charges were paid by airlines, the bulk of the cost would be
borne by rich people, who fly more often than poor people. But part of
the cost would be borne by air freight, which may or may not involve
products bought preponderantly by more affluent consumers, and part of
the cost may be covered by lower wages for airlines' cleaning personnel.
In short, many of the ramifications are so indirect and remote that we
cannot say with any degree of confidence who is likely to end up paying
the largest proportion of emissions charges. What spotty evidence there is
constitutes some reason for concern that the poor may indeed pay more
than their share (see, for example, Freeman 1972, Dorfman 1975, 1976,
and Baumol and Gates1) but this may well be as true of the costs of
environmental programs involving direct controls.
A second question relating to a system of emission charges is whether it
will work in view of the apparent ability of the emitters simply to pass the
charges on to consumers in the form of higher prices and go on emitting.
The answer is that, except where the firm in question is completely
removed from competitive pressures, it cannot pass all of these charges
on to its consumers. The most conclusive evidence that this is so is the
virtually universal and bitter opposition of emitters to a system of
charges, an opposition that can most readily be explained by the fact that
these charges will cost them some money.
It is of course true that the charges can be expected to produce some
rise in the prices of the emitters' products. But this, too, is part of the
abatement mechanism of a program of financial incentives. For the rise
of relative prices of commodities that cause emissions will help to shift
consumer demand toward commodities whose production is less damag-
1Baumol, W.J. and W.E. Gates: Economics, Environment and the Quality of Life.
Englewood Cliffs, NJ: Prentice-Hall, Inc., forthcoming.
-------
28 POLICY AND LEGAL ISSUES
ing to the environment, and that, surely, is one of the aims of such a
program.
Circumstances Favoring the Use of Direct Controls
Despite the arguments for the virtues of a system of emission charges or
financial incentives, they do not constitute a panacea for all environmen-
tal problems. There are many circumstances, perhaps encompassing the
preponderance of major sources of noise, in which emission charges will
prove impractical or not be as effective as direct controls. Specifically,
there are at least six general cases that favor direct controls. In four of
them, direct controls are inherently superior, while the other two arp
matters of adaptation to political and institutional realities.
First, if the effects of an emission are judged to be so serious that it
should be prohibited altogether, then there is clearly nothing to be gained
by setting up the elaborate machinery involved in imposing charges that
would have to be so prohibitively high that they would never be collected.
While it is true that charges can, in principle, work even in these cases,
there is little point in using this indirect route. It is not clear that this
advantage of direct controls is likely to be of major significance for noise
abatement, except perhaps for the protection of workers in extremely
noisy factories or the prohibition of operations by particularly noisy
aircraft.
Second, if there is a likelihood of unanticipated environmental
emergencies, particularly if their consequences are extremely serious, a
system of charges may prove to require too much time for its adoption or
modification and be too slow in eliciting its effects. When there is an air
pollution emergency that threatens to cause an unacceptable increase in
both mortality and morbidity, the easiest and most effective policy
approach may be a ban on the use of private passenger vehicles except for
emergency purposes and a ban on the use of incinerators (perhaps with
certain specified exceptions). This may not be a case that is important for
noise abatement since it is hard to imagine serious and unanticipated
noise emergencies. Physiological damage from noise is usually a result of
protracted and repeated exposure to sound; hence, the fact that direct
controls lend themselves to modification on short notice seems to offer
little or no advantage in noise abatement.
Third, and probably the most important case in which direct noise
abatement controls offer advantages over financial inducements, is if
monitoring of emitting sources is impractical because it is unacceptably
inaccurate or costly; a program of charges will not work if it lacks the
information on which to base its assessment of fees. Charges that are only
-------
Choice of Policy Instruments 29
haphazardly related to individual emission levels will fail in their basic
objective—the provision of a reward (the reduction in payments) that
increases with every reduction in emission levels.
It is important to note that where monitoring is impractical, the
assigned emissions limitation (performance standards) approach to direct
control is ruled out as effectively as a program of emission charges. One
cannot enforce a fixed maximum decibel level on motorcycles if one has
no way of measuring how much noise any given motorcycle emits once it
leaves the showroom. Where effective monitoring is not practical, process
or equipment specification (such as required retrofitting or welding of
rails) remains the only reasonable option.
After reviewing the available material on noise monitoring techniques,
we conclude that some sources can be monitored with sufficient
effectiveness and economy to constitute no significant impediment to the
use of charges as an implement of regulation. In other cases, source
monitoring is far more difficult, and direct controls of the equipment-
specification type may be the only available option.
The level of noise emission of standardized commercial vehicles, such
as aircraft or trucks, can be evaluated in different circumstances by
testing of the prototype model or by periodic retesting, vehicle by vehicle,
although both of these approaches may run into some legal problems.
While the character of their noise emissions will vary with speed and
route, sample studies and records may provide tolerably approximate
information. Strategically placed microphones at construction sites or
along railroad tracks may be able to provide most of the information that
can reasonably be desired for a program of charges, though there are
some techniques for partial evasion of such monitoring procedures. Small
construction projects of short duration and general highway and urban
noise levels, however, constitute difficult monitoring problems, and,
taken together, are major sources of noise emission.
Fourth, financial incentives work very imperfectly if there are ways the
emitter can escape the burden of charges. This may be true of emissions
generated by government activity unless the charges cut directly into the
budget of the agency that operates the vehicles. If political arrangements
permit the agency to increase the budget by an amount sufficient to cover
the charges, its motivation will be undermined, and only direct controls
will be able to induce an improvement in performance.
The same may be true, at least partly, of a regulated public utility in
private industry, such as an airline whose profits are effectively con-
strained, if it is permitted to raise its prices to cover any emission charges.
There are two important reservations here: first, regulatory lag is likely to
delay any such price adjustment and, meanwhile, the burden of the
-------
30 POLICY AND LEGAL ISSUES
charges will fall on the emitter. Perhaps more important, any rise in prices
granted by the regulator will discourage consumption of the emitting
products to some degree and will reduce emissions correspondingly. In
any event, it is clear that, in either of these circumstances, the case for a
program of financial inducements is weaker than it is when more of the
burden of these charges falls on the emitter.
Fifth, in light of the legal issues (see next chapter), it is possible that a
program of emission charges will initially involve more complex, costly,
and time-consuming challenges in the courts than a program of direct
controls. If so, this should be counted as an item that favors the latter.
However, it is by no means clear that matters work out that way on
balance. Experiences such as the Reserve Mining case, which has so far
been in the courts for seven years, indicate that direct controls are by no
means immune from legal costs and delays.2
Finally, it is no doubt true that direct controls are today more
acceptable politically than are emissions charges. One may conclude that,
even if charges are otherwise superior, it is better to settle for the second-
best program of direct controls than to hold out for the best and end up
with none at all. We are not in a position to judge this matter. We may
conjecture that the opposition to emission charges by local governments
may weaken if their financial difficulties continue to grow (as there is
good reason to expect), for a program of direct controls is likely to add
costs for enforcement while a system of charges will, incidentally, provide
additional revenue.
TOWARD A CHOICE OF POLICY INSTRUMENTS FOR NOISE ABATEMENT
As we have seen, monitoring is one of the key considerations affecting the
choice among the three major options: direct controls by process or
equipment specification, direct controls by assigned emission limits, and
emission charges. If monitoring of emissions by individual sources is
cheap, accurate, and effective, there is a great deal to be said for the use of
emission charges, while if effective source monitoring is impractical, only
process or equipment specification will work. This suggests that the
approach that most nearly resembles a compromise — direct controls by
assignment of quantitative emission limitations on individual emitters —
Reserve Mining Company, jointly owned by Republic Steel and Armco Steel, dumps
tons of taconite tailings (asbestos-containing wastes) into Lake Superior, a chief source of
drinking water for Duluth, Minnesota, and many other communities. The grave health
effects of the ingestion of asbestos fibers appear to be well documented, but the courts have
been unable, in seven years of litigation, to ban the dumping of these wastes.
-------
Choice of Policy Instruments 31
may rarely be the preferable choice. Either the more extreme form of
direct control or the use of financial incentives is likely to be superior.
There is, however, a second form of intermediate arrangement. This is
the use of financial incentives along with assigned emission limitations,
which the emitter is prohibited from exceeding in any event. If charges
are sufficiently high to be effective, the limitations are likely to serve
merely as standby ceilings that are in fact rarely, if ever, reached.
Nevertheless, such an arrangement may prove comforting to those who
are skeptical about the reliability of financial incentives.
On the critical issue of monitoring, we may note that noise abatement
programs differ in two significant ways from other environmental
programs. First, despite its many problems (discussed in later chapters),
noise monitoring is probably more accurate, more straightforward, and
more easily automated than monitoring for almost any other major
emission. This would appear to suggest that a system of emission charges
would be particularly promising for noise control. Commercial jet aircraft
are probably the best candidates for emission charges since they operate
from a small set of airports and are capable of being monitored with
sufficient accuracy. However, in many cases of urban or highway noise,
the sound is contributed by a very large number of highly mobile sources.
It is extremely difficult in such cases, despite the effectiveness of
monitoring equipment, to attribute a specific component of the total
sound to a particular automobile, truck, or motorcycle. While in such
situations it is possible to monitor the general level of environmental
noise quite accurately, and even, perhaps, to distinguish the contribution
of particular classes of emitters (truck tires versus motorcycles), it is hard
to determine, for example, which vehicle muffler was defective and how
much it contributed to the overall noise; this would be required by an
effective program of emissions charges. All these considerations are
reflected in our recommendations on appropriate instruments of control.
REFERENCES
Freeman, A.M., Ill (1972) Distribution of Environmental Quality. Pages 243-280,
Environmental Quality Analysis: Theory and Method in the Social Sciences. Edited by
A.V. Kneese and B.T. Bower, Baltimore: Johns Hopkins Press.
Dorfman, N.S. and A. Snow (1975) Who Will Pay for Pollution Control?—The Distribution
by Income of the Burden of the National Environmental Protection Program, 1972-1980.
National Tax Journal 28(1): 101-115.
Dorfman, R. (1976) Incidence of the Benefits and Costs of Environmental Programs.
Presented at the Meetings of the American Economic Association, Atlantic City, N.J.,
September 1976. American Economic Review 67(1):333-340.
-------
2
Legal
Issues
INTRODUCTION
This chapter examines some legal issues that must be faced in deciding
how (and whether) to impose governmental control on transportation
noise. Legal issues must be considered so that the legal protection to
various parties is provided while unnecessary legal challenges are
avoided. The chapter deals with a few of the legal issues affecting noise
regulation, focusing on problems that have received only limited
attention in the past and that are particularly relevant to the findings
discussed in this report.
First, it considers the legal problems raised by emission charges as
distinct from other forms of governmental regulation of noise. While this
type of regulation is emphasized by economists, it has received very little
judicial consideration. While economists have frequently compared
emission charges to taxes, this is not a felicitous comparison from the
legal point of view. That is, if the courts treat emission charges as taxes,
they are likely to find them invalid. Furthermore, if emission charges are
treated as criminal penalties, the charge system will be so burdened by
procedural requirements that the advantages posited for a charge system
will be lost. The most advantageous legal characterization of emission
charges would be as some sort of civil fine or, better yet, as a unique
regulatory device.
Second, the chapter considers legal problems that are raised by
regulation that is designed primarily to prevent annoyance, nuisance,
32
-------
Legal Issues 33
etc., so-called aesthetic harm. Most people dislike noise—that is, they
view noise as an aesthetic harm—so aesthetic objections are clearly
identifiable as a reason for noise regulation. While we know of no past
examples of federal regulation explicitly designed to prevent aesthetic
harm, we believe the federal government should be able to regulate for
aesthetic purposes. It should be noted that the power of state and local
governments to regulate on aesthetic grounds is firmly established.
Any regulatory scheme involves a myriad of legal problems. These are
generally familiar to lawyers versed in administrative law and are not
included in this chapter, which is not designed as a definitive legal study.
It is intended more modestly to highlight a few legal issues that may have
a significant effect on shaping the regulation of transportation noise and
on which information is not readily available.
A SYSTEM OF EMISSION CHARGES
We begin with a description of how a system of emission charges might
work.
An administrative agency would set a schedule of charges after
considering the levels and effects of noise emissions from the noise
sources or categories of sources to be subject to the charges. The charges
could be set in several different ways. The charge on any given noise
emission could be related to the value of the harm caused by that level of
emission: the charge could be made equal to the best available estimate
of the monetary value of the damage caused by the noise. Alternatively, a
generally acceptable level of noise could be determined. (This would not
have to be a level at which there is no harm from the noise, but rather a
level at which the harm would be accepted without charging any costs to
the noise emitter.) A charge could then be made for noise emissions in
excess of that level. The charge for any increase in emissions would be an
amount equal to the incremental cost of the harm caused by the emission
increase. Charges set in either of these manners are harm-related charges.
Charges can also be set on a control-related basis. An acceptable base
level of noise emissions would be determined, and charges for any
emissions in excess of that level would be set at a level sufficiently high to
induce a reduction in emissions to the base level.
Notice of the charge schedules would be given to emitters, who would
have an opportunity to challenge them before the agency and the courts.
Once the schedules were settled, the agency would make periodic
determinations of emission levels from each source or category of
sources, and charges would be assessed based on those levels and the
charge schedules. Before any emitter could be required to pay the
-------
34 POLICY AND LEGAL ISSUES
assessed charge, it would have an opportunity to challenge the assessment
either before the agency or before a court.
For example, if the agency found that the emissions from a certain type
of truck were 95 dB(A) (decibels, A-weighted sound level, see Chapter 3),
and a truck fleet owner claimed that they were only 85 dB(A), the owner
would be able to challenge the agency finding. The challenge might be
brought before a court or the administrative agency. If it were brought
before the administrative agency, it would be subject to limited review by
a court. In either case, if the challenge were unsuccessful, the emitter
would have to pay the assessment. The agency could seek the aid of a
court to compel payment, if necessary.
LEGAL CONSTRAINTS ON A SYSTEM OF EMISSION CHARGES
There is no true system of emission charges now in operation in the
United States. If a system of emission charges for noise regulation is
instituted, courts will probably have to determine whether the charges are
valid and what legal limitations apply to them. In making this determina-
tion, courts may view emission charges as a unique and new type of
regulatory device, or they may treat them as analogous to existing types
of monetary assessments, applying to emission charges those legal rules
that apply to such other assessments. Statements of legislatures and
administrative agencies may guide, but they will not bind the courts in
deciding how to characterize emission charges.
This section surveys the other devices that emission charges may be
taken to resemble and the legal constraints on their use that would follow.
It also discusses whether specific statutory authorization for emission
charges, as distinct from other types of regulatory programs, is required
before emission charges may be imposed by a regulatory agency.
Taxes
Perhaps the most familiar type of government assessment is a tax. It is
difficult to define a tax in a way that distinguishes it from a license fee, a
fine, or other monetary assessments. The best we can say here is that a tax
is a type of revenue-raising device to which a certain set of legal
constraints usually applies. If emission charges are treated as taxes, these
constraints would apply to the charges and may limit their usefulness.
There would at least be some uncertainty about how the charges could be
set and applied.
One of these uncertainties is whether an administrative agency would
be permitted to set a tax. The system of separation of powers prohibits a
-------
Legal Issues 35
legislature from delegating too much power to an administrative agency,
which is part of the executive branch of government. A statute delegating
so much power to an agency that it violates the separation of powers
doctrine would be held invalid by a court, and the regulations passed
under such a statute would also be invalid. While federal courts in recent
decades have not invalidated statutes on delegation grounds, a recent
case (National Cable Television Association, Inc. v. United States 1974)
suggests that the threat to a statute is still alive when the power to tax is
concerned. According to this case, a statute may be invalid if it delegates
to an agency the power to set taxes without sufficient guidelines on how
that power must be exercised. The cable television case may be read to
suggest that an agency cannot be given the power to set taxes at all (but
see Federal Energy Administration v.Algonquin SNG, Inc. 1976). A similar
doctrine is recognized by some states. In fact, some state courts may
apply the doctrine against emission charges even if federal courts do not
do so. If so, federal agencies could set emission charges, but state agencies
in some states would not be permitted to do so.
This could be a significant impediment to use of emission charges,
because it is desirable to have an agency rather than a legislature set such
charges. Setting emission charges involves many technical determinations
about medicine, economics, sociology, geography, etc., and agencies are
generally more adept than legislatures at handling complex technical
problems. Moreover, agencies are more likely than legislatures to have
the flexibility needed to adapt to unforeseen problems in designing a new
type of system. In addition, charges will probably vary over time as a
result of changes in prices, in the volume of total emitter activity, in
control technology, etc., and legislatures are less well suited than agencies
to monitor such changes and to adjust charges to them. If the setting of
emission charges by an agency is held invalid on delegation grounds, the
advantages of expertise and flexibility may be lost. Since the delegation
problem is more likely to arise if the charges are characterized as taxes,
such a characterization would be unfortunate.
One way to minimize a potential delegation problem would be to have
the legislature specify in a statute what factors the agency must consider
in setting charges. Alternatively, the delegation problem would be
circumvented were an agency to recommend charges to the legislature,
with such charges then enacted by statute. But this technique is likely to
be ponderous, particularly at anything other than the local level. In
addition, it is more likely to produce charges set in part on the basis of
particular political trade-offs that are beyond, and even contrary to, the
economic rationale for the charges.
Emission charges, if construed as taxes, would be a form of indirect
-------
36 POLICY AND LEGAL ISSUES
taxes. Indirect taxes levied by the federal government must be uniform
throughout the United States. Similarly, in most states, taxes levied by the
state must be uniform throughout that state. Under certain conditions,
some differences in tax rates may not violate the uniformity requirement:
similar activities must be taxed at the same rate wherever they are found,
but activities that are deemed different may be subjected to different
rates. Thus, emitters of different levels or kinds of noise could be charged
at different rates. However, it is unclear whether one emitter of a given
noise could be charged more than another emitter of the same noise just
because the first emitter is located in a more populous area where there
are more hearers and greater external costs associated with the noise
emissions. Yet the economic rationale for harm-related emission charges
that the charge be set at the level of the external cost associated with the
emissions—would require that higher charges be assessed to the emitter
in the more populous area. (Similarly, it is unclear whether equal emitters
could be charged different amounts because of different control costs;
this may be desirable under a control-related charge scheme.)
A federal system of emission charges construed as taxes may face a
further problem. State and local governments are immune from federal
taxation under certain circumstances. This immunity from taxation may
be more extensive than state and local immunity from direct federal
regulation. Thus, if emission charges are characterized as taxes, it might
be difficult for the federal government to apply the charges to state and
local governments. However, the tax immunity does not now appear to
be very broad, and even if emission charges are construed as taxes, they
probably could be assessed against states and localities as long as they
neither discriminate against states and localities as noise emitters nor
unduly interfere with the sovereign functions of the taxed governments.
These rules would probably permit the federal government to apply
emission charges to most state-owned vehicles even if the courts chose to
treat the charges as taxes, although some immunity question may arise if
almost all of the types of vehicles charged are state owned. Again, a
parallel problem exists with respect to state and local systems of emission
charges, where courts may question the power of one unit of state
government to tax another unit.
Finally, if emission charges are construed as taxes, the statutory
authorization must originate in the popular chambers of federal and state
legislatures, which, on the federal level, is the House of Representatives.
This point could pose a problem if Congress were to pass a Senate-
originated bill involving emission charges that is later interpreted as a
taxation measure by the courts.
The foregoing discussion does not argue that emission charges will be
-------
Legal Issues 37
treated as taxes, but only that charges might be so treated and outlines
the uncertain consequences. It is mainly because both emission charges
and taxes raise revenue that a court-drawn analogy between the two is
foreseeable. However, there are also reasons for not treating emission
charges as taxes.
Emission Charges Distinguished from Taxes
First, emission charges can be distinguished from most, if not all, taxes.
The primary purpose of emission charges is not raising revenue, which is
usually seen as the primary purpose of a tax, but rather to encourage the
proper level of emission reduction at a minimal administrative cost.
Second, there is no need to place extra constraints on the government's
authority to impose emission charges by calling them taxes. As discussed
above, it appears that there may be stricter restraints on the government's
authority to tax than on the government's authority to regulate. To the
extent that this is so, it can probably be ascribed to the view that the
authority to tax is stronger than the authority to regulate. At least so far
as emission charges are concerned, we think this is an inaccurate
perception.
Emission charges involve no greater interference with noise emitters
than would direct controls, either in the form of emission limitations or of
equipment specifications. In fact, one of the justifications for emission
charges is that they involve less interference with the emitter. Each
emitter is free to decide what degree of noise abatement to achieve and
how to achieve it, subject only to the constraint of paying the cost of the
unabated noise through the emission charge. Since emission charges thus
interfere with emitters even less than do more traditional types of
regulations, the legal safeguards applied to limit governmental power in
the case of regulations should be sufficient; such additional safeguards as
may apply to the taxing power are unnecessary.
Fines
A fine is a charge levied on a party as a penalty for violating a legal
standard. There are two types of fines: criminal and civil. The distinction
between the two is sometimes unclear, but can be important; more
procedural protections must be given to noise emitters if the charges are
treated as criminal fines than if they are treated as civil fines.
However emission charges are characterized, courts are likely to find
that each emitter must have the opportunity for some sort of hearing
before being required to pay the charges. The alleged emitter must have
-------
38 POLICY AND LEGAL ISSUES
the opportunity to challenge both the validity of the charge scheme and
the accuracy of the specific charge assessed against it. However, the
procedural protections that must be afforded to a source at the hearing
may vary greatly depending on how the charge is characterized.
If emission charges are treated as criminal fines, the law would afford
many procedural safeguards to the emitter at a hearing. For example,
individual court trials with a judge presiding would be required before a
charge could be enforced against a noise emitter, the emitter would have
a right to a jury trial, and emission levels would have to be proved beyond
a reasonable doubt instead of only by a preponderance of the evidence. It
is also likely, but not certain, that each emitter could challenge the
validity of the charge scheme and the charge schedule at the time of the
enforcement action. These requirements would substantially increase the
cost of administering a system of emission charges. They would also
hamper its operation by decreasing the likelihood that an emitter would
have to pay a charge. This, in turn, would probably decrease voluntary
compliance because an emitter would be encouraged not to pay a charge
if it found that sanctions for not paying were unlikely.
It is not clear what procedural requirements would apply to emission
charges if they are treated as civil fines. Most likely, each emitter would
still have a right to an individual hearing on the level of its emissions.
These hearings could probably be held before an agency rather than a
court, with some formalities eliminated. For example, there may be no
jury trial or formal rules of evidence, but an emitter would probably
retain the right to be represented by counsel and to have all evidence that
is to be used against it presented at the hearing. Courts would most likely
review the agency proceedings but not rehear the evidence. Also, it is
more likely, but still not certain, that challenges to the charge schedule
could be limited to the time the schedule is promulgated and not
permitted at the time of enforcement. These procedures reduce the cost
and uncertainty associated with enforcing fines.
Another legal impediment to civil fines that might be applied to
emission charges could arise at the state and local level. State courts may
view imposition of civil fines as a judicial function and prohibit charge
assessments by a state administrative agency as an improper usurpation
of judicial power. While there is no similar usurpation problem at the
federal level, state courts are free to assert their own views on the powers
of the various branches of state government when state assessments are
involved. Some state courts have allowed administrative agencies to
impose civil fines, but other states may not do so.
Having examined the consequences of treating emission charges as
fines, we turn to the question of whether they should be so treated. There
-------
Legal Issues 39
are a number of significant distinctions between emission charges and
criminal fines that have led writers on the subject to argue that the two
should not be treated alike. Assessment of emission charges do not
involve a collective judgment on the character of the emitter or on the
social desirability of its conduct, while criminal fines indicate collective
condemnation of the criminal's character and conduct. Emission charges
are not designed to force adherence to some fixed standard, as are
criminal fines. In addition, it is appropriate to deny the procedural
protections associated with criminal penalties to noise emitters. In the
case of criminal fines, it is so important to protect a person from wrongful
condemnation or incarceration that it is worthwhile to require protective
procedures involving high costs and a significant chance of erroneous
acquittal of the accused. Where neither the element of collective
condemnation nor potential incarceration exists, the high cost and
inefficiency of the criminal justice system is not justified.
The argument for distinguishing emission charges from civil fines is not
as clear, partly because the characteristics of civil fines are not well
defined. Some distinctions can be drawn. Civil fines are sanctions for
violation of some norm, but there are no norms involved in a system of
emission charges intended to force a noise emitter to bear the full costs of
its emissions. On the other hand, when the charges are designed
specifically to encourage noise reduction to some predetermined target
level, noise emission charges look more like civil fines.
There is another distinction from civil fines. In the case of a fine, a
party who violates a standard and pays the fine is thought to have
committed a "wrong" in having exceeded or violated the standard and is
not privileged to continue the violation. In the case of an emission charge,
no concept of "wrong" applies, and an emitter who regularly pays the
charge is privileged to exceed any target level.
Regulations
Not all the potential legal constraints on emission charges result from the
chance that they will be treated by analogy to other charge devices. Even
if a court recognizes emission charges as a unique regulatory device, it
must decide whether the agency that tries to regulate noise through
emission charges, rather than through some type of direct controls, has
the statutory authority to do so. May emission charges be imposed under
a statute that only provides authority to "regulate" noise, or is specific
statutory authority for the charges required?
It would be safest to set emission charges under a statute specifically
authorizing them. Without such specific authorization, there is precedent
-------
40 POLICY AND LEGAL ISSUES
indicating that a federal agency could impose emission charges for noise
under a statute authorizing regulation only if the legislative history (i.e.,
committee reports and Congressional floor debates) indicated that
Congress had intended that emission charges be among the regulatory
techniques that might be used.
There remains the question of whether an agency could use emission
charges as a regulatory technique when such charges are not specifically
authorized by either the statutory language or its legislative history. There
is no precedent holding that such charges would not be permissible.
Whether such charges would be allowed is probably tied in part to the
question of how a court would characterize the charges. Charges that are
characterized as taxes are most likely to be held beyond an agency's
authority; charges treated as a unique regulatory device, least likely.
Moreover, the more closely emission charges are shown to be reasonably
related to a statutory goal, the more likely they are to be held valid
without specific statutory authorization. Thus, if the statutory goal, as
stated by the legislature, is reduction of noise emissions, an agency
promulgating emission charges may be called on to show that the charges
are calculated to reduce noise levels. (It would no doubt be helpful if the
agency were to explain why emission charges were preferred over other
control devices.) On the other hand, if the statutory goal is attainment of
"the optimal level" of noise emissions, the agency would not necessarily
have to show that emission charges are calculated to reduce noise, but
only that the charges are calculated to optimize levels of noise emission.
In summary, there is some risk that agency action imposing emission
charges that are not specifically authorized by statute or legislative
history would be found invalid. This risk can probably be diminished by
careful agency explanation of the rationale behind the charges as a form
of regulation. When legislation is being drafted, it would be wise to
include explicit authorization for emission charges.
LEGAL CONSTRAINTS ON GOVERNMENTAL REGULATION
OF AESTHETIC HARM
To a significant extent, the harm done by noise appears to be aesthetic.
Aesthetic harm, as the term is used here, is not limited to infringement on
one's preferences, but encompasses all harms not correlated with
demonstrable physical or psychological harm. This corresponds in part to
what is also called annoyance (see Chapter 6). In addition, however, it
may well encompass harmful physical or psychological effects that
cannot be immediately demonstrated, for the effects of noise exposure are
seldom sudden or immediate. While we can rarely prove that noise causes
-------
Legal Issues 41
specific physical or psychological harm, we know that many people
object to noise. Because controls on noise, perhaps more than controls on
other types of environmental pollutants, rely on aesthetic grounds for
their justification, we must consider the legal position of measures
designed to abate aesthetic damage.
Any scheme of governmental control involves the imposition of costs
on at least some of the parties subject to such control. Simple fairness as
well as legal notions of due process generally require that two factors be
present before control costs may be imposed on a private party by the
government. First, there must be a legally cognizable harm. Second, there
must be a determination that the party who is charged with the cost is the
one who caused the harm. The first of these factors presents more of a
problem when dealing with noise emissions than when dealing with many
other pollutants. The problem is whether the harm caused by noise is
legally cognizable. What constraints are there on legal cognizance of
aesthetic harm?
One of the two sources of law on this issue is law on court-imposed
sanctions on actions by private individuals for aesthetic interference. The
second source results from some governmental regulatory schemes
designed to alleviate aesthetic harm or to enhance aesthetic values.
The main source of private law is the law of private nuisance, which
provides a remedy to one person against a neighbor who causes an
unreasonable interference with the use and enjoyment of the person's
land. We may look at how courts have reacted to claims that noise is an
unreasonable interference to determine whether noise is regarded as a
legally cognizable harm. In general, we find that it is so regarded if the
noise is substantial. On the other hand, courts are reluctant to recognize
some aesthetic interferences, particularly when the degree of interference
is difficult to measure. This measurement issue arises with regard to noise
but is less problematic than for some other aesthetic values, such as visual
ugliness. There are reasonably good techniques for measuring the amount
of noise, if not the discomfort it causes.
There is also widespread acceptance of aesthetic interference as a
legally cognizable harm that may be subjected to governmental regula-
tion. Even if courts are themselves hesitant to measure aesthetic qualities
or to balance them against other values, they will usually defer to
attempts by legislatures or administrative agencies to do so. Legislatures
earn this deference by right of their political mandate; agencies earn it by
their expertise. Thus, local zoning regulations based on aesthetic values
alone, that is, schemes designed to protect natural beauty, have been
upheld. On the federal level, there are numerous regulatory statutes
designed to protect the public health and welfare, and public welfare
-------
42 POLICY AND LEGAL ISSUES
almost certainly includes aesthetic values. While we have found no
federal laws that provide for regulation of private activity on aesthetic
grounds alone, we find no reason why statutes or regulations of this sort
would not be valid. The power of Congress to decide what societal values
should be protected is broad. Moreover, even where the harm appears to
be mainly aesthetic, the chance that there may also be physical harm may
be used as the basis for regulation. Congress may, and indeed has, found
that the potential health hazard from noise is reason for regulating noise
emissions.
Other subtle stumbling blocks may exist in the area of regulations
based on aesthetics. A statute or regulation that is qualitative, rather than
quantitative, may be found invalid as impermissibly vague. More
importantly, qualitative ordinances create enforcement problems, partic-
ularly in proving the existence of noise exceeding the stated standard. But
these are potential pitfalls for the drafter of the regulatory provision and
do not detract from the general proposition that regulation for aesthetic
purposes is acceptable.
REFERENCES
Federal Energy Administration v. Algonquin SNG, Inc. (1976) 426 U.S. 548, 96, S.Ct. 2295.
National Cable Television Association, Inc. v. United States (1974) 415 U.S. 336.
BIBLIOGRAPHY
Anderson, F., A. Kneese, R. Stevenson, and S. Taylor. Economic Charges: Economic,
Technical, Legal and Political Aspects. July, 1975 draft. (Forthcoming with possible title
change by Resources for the Future.)
Irwin, W.A. and R.A. Liroff (1974) Economic Disincentives for Pollution Control: Legal,
Political and Administrative Dimensions. Prepared for the Office of Research and
Development by the Environmental Law Institute, EPA-600/5-74-026. Washington,
D.C.: U.S. Environmental Protection Agency; PB-239 340/3BE. Springfield, Va.:
National Technical Information Service.
-------
II
TRANSPORTATION
NOISE:
ITS MEASUREMENT,
SOURCES,
AND PROSPECTS
-------
3
Noise
Indices
INTRODUCTION
This chapter describes the different measures used to assess human
exposure to sound and to show their interrelations. Because of the
multivariate character of sound, a number of different elements usually
enter into the calculation of a measure. Such a combined measure is
called an index.
Since no single number can accurately reflect the behavior of a
multiplicity of variables, there can be no "ideal" noise index. Any index is
a combination of the variables representing the relevant attributes of the
sound, such as intensity, frequency spectrum, duration, and intermitten-
cy. Moreover, the combination most appropriate for one purpose will
give more weight to certain attributes than will the combination for some
other purpose. Thus it is quite natural that a number of different noise
indices are currently in use, and it is futile to search for any single index
capable of serving all the different purposes for which indices are used.
It is not our aim here to provide details sufficient to permit the reader
to calculate a given index from the material provided, but, rather, to
describe enough about the different indices to indicate the nature of their
construction and to offer a comparison of their objectives. Besides
describing the indices that are used most widely, this chapter indicates
the costs of obtaining the required data and calculating a given index so
that the total cost of a monitoring program can be assessed.
45
-------
46 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
NOISE
Noise is often defined as unwanted sound. Such a definition makes it
clear that the designation of a given sound as noise is a matter of human
evaluation. Obviously, people will differ in their evaluations and even the
same person may change an evaluation from one time or circumstance to
another. Some sounds also have special meanings, such as a baby's cry to
its mother. No index attempts to measure such individual differences in
the evaluation of sound. Rather, measures of noisiness are intended to
reflect the fact that at an intensity sufficiently high most people will
display an antipathy to sound. It is this common judgment that has
served as the basis for the design of noise indices.
DECIBELS
All of the basic noise measures use decibels as a unit. The decibel (dB) is
simply a logarithmic transformation of the basic measure of sound
intensity (energy). Such a logarithmic transformation is convenient
because it makes it easier to deal with the enormous range of intensities
that people can hear. Its main characteristic is that a given increase in the
logarithm corresponds to a given ratio in the basic number rather than a
given absolute change. In a fairly quiet location, a given rise in decibel
level will mean a fairly small rise in absolute noise level, say measured in
watts/cm2, because a given ratio of a small number is still a small
number, while under noisy conditions an equal rise in decibel level can
indicate a relatively large increase in absolute noise.
Zero decibels on a scale of sound pressure level is nearly the threshold
of sound at 3000 cycles per second—that is, it is the softest sound that can
just be heard when the frequency is near the highest note of a piano. Sixty
decibels is moderate speech heard at 1 meter, and 90 dB is the sound one
hears on a subway platform when the train arrives. At 110 and 120 dB the
sound is so intense that most people start to complain about pain or tickle
in their ears.
Doubling the intensity of sound pressure is equivalent to an increase of
3 dB. However, if one asks people to double the loudness of a given
sound, there will be a range of judgments, from about 5 dB to 15 dB, but
the mean setting will probably be about 10 dB. Thus, changing a sound
from 40 dB to 80 dB does not double the loudness, it doubles it 4 times—a
16-fold increase in loudness. Incidentally, 40 dB and 80 dB correspond,
respectively, roughly to a whisper and a shout, heard at about 1 meter.
-------
Noise Indices 47
NOISE EXPOSURE INDICES
A striking feature of the study of noise control is the variety of different
indices and units of measurement that are used. Loudness level (LL),
perceived noise level (PNL), and speech interference level (SIL), are some
of the quantities proposed to measure the magnitude of noise. In
addition, the noise number index (NNI), noise pollution level (NPL),
noise exposure forecast (NEF), and many other indices demonstrate the
different procedures used to assess noise exposure in different situations.
It is natural to ask why specialists have not settled on a single method for
the assessment of noise magnitude or exposure. There is no simple answer
to that question, other than the obvious one that different measures were
suggested by different groups to satisfy different requirements.
Differences in the treatment of the sheer energy content of a sound are
relatively slight and do not change much from one source to another.
Significant variations in the indices result from differences in the uses for
which they are intended. For example, a noise index appropriate for a
turbine generator probably does not need to assign an important role to a
measure of intermittency. After all, the turbine should, under normal
circumstances, run continuously. However, intermittency is a primary
characteristic of other types of industrial and construction noise: the
sound of a pile driver, for example, is seldom steady and the components
of an index used to evaluate it should include the duration of the
exposure and the number of noise events. Aircraft flyovers also fall into
this class and, over the years, a number of special components have been
added to the basic measure of intensity in order to provide a better
assessment of exposure to noise from these sources.
In our judgment, the major differences among indices result from
differences among the elements or components used in making up the
composite index. These components are often evaluated differently in
different indices. The inclusion or exclusion of different elements in a
particular noise index is probably the natural result of the specific
application for which the measure was devised.
There are many reasons for desiring a single noise index: regulation
would be facilitated because standards could be based on a single,
comparable, scale; instrumentation could be standardized and the cost of
monitoring would be reduced; and informing the public about noise
would be easier. But effective noise abatement need not await the
adoption of a single noise index because the differences among the
measures and indices are relatively small compared with other uncertain-
ties. Often the differences among different measures is only 1 or 2 dB.
This is much less than the variability one would encounter if one asked a
-------
48 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
group of subjects to adjust two sounds to be equally "noisy." Such
variability can be at least five times greater. But while such differences in
the indices are relatively small compared to the variations in subjective
responses, this does not prevent heated controversy. Differences of 1 or 2
dB become important when limits or noise ceilings are proposed and the
economic implications of these small differences are not trivial.
This discussion is intended to identify and compare the elements of the
different noise indices and to provide approximations that are useful in
comparing one noise index with another.
NOISE MAGNITUDE
Central to any noise index is an assessment of the magnitude of the
sound. Sound affects people by its threat of hearing damage, through its
capacity to annoy, or through other effects on behavior. But any of these
effects can be produced only by the physical process that generates the
sound.
Physically, sound can be denned as a mechanical disturbance
propagated in an elastic medium. The elastic medium is the air or
atmosphere and the mechanical disturbance can be thought of as a
variation in pressure. At any point in space there is an average level of
pressure, the atmospheric pressure, and sound is a relatively rapid
fluctuation in this average pressure. This fluctuation travels through space
as a wave.
For most purposes, an adequate physical description of sound is
provided by the variation in pressure measured by a microphone placed
near a subject's ear. Initially we will deal with this aspect of sound,
ignoring other variables such as duration, repetition rate, etc. There are
two general classes of measures of the magnitude of a sound: direct
measures and derived measures.
Direct Measures
Because of the physical structure of our ears, some sounds are easier to
hear than others even when they are equal in energy. We can most easily
hear energy in the frequency region near 3000 cycles per second. The unit
of frequency, cycles per second, is designated as Hertz (Hz), after the
physicist. Sound whose predominant energy lies at frequencies below 100
Hz or above 10,000 Hz may require a million times more energy to sound
as loud as a 3000-Hz component.
Various functions have been proposed to provide weights for each
frequency in the spectrum in order to reflect the ear's differential
-------
Noise Indices
49
m
TJ
LU
-20 -
-30 -
50 100 200 500 1000 2000 5000 10000
FREQUENCY (Hz)
SOURCE: Pearsons (1973).
FIGURE 3.1 Four common frequency weightings used in direct measurement of sound.
sensitivity to frequency. These direct measures apply a weighting to the
overall spectrum of the sound; that is, they selectively weight pressure
variations at different frequencies. The weighted pressures are then
squared and combined to obtain a single number. The logarithm of this
quantity is proportional to its decibel value and is called a sound level.
The weighting network is usually specified by an adjective preceding
sound level, for example, the A-weighted sound level or simply sound
level A.
Figure 3.1 shows four common weights used in direct measurement of
sound. The weights are somewhat different and thus a given sound's A-
weighted level will be different from its C-weighted level, but both will be
expressed in decibel units. These different weighting structures have been
constructed because each serves some purposes better than the others.
Often these units are labeled as dB(A), dB(C), etc. This is technically
incorrect because it is not the decibel, but the sound level that is A, B, or
C weighted, but the practice occurs and it need not lead to confusion if its
meaning is understood.
(Sound level, using A, B, or C weights, is usually measured by a meter
satisfying requirements of American National Standard Specification for
-------
50 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
Sound Level Meters SI.4-1971. The meter has two dynamic characteris-
tics. One, called the fast setting, integrates sound over an interval of
about 250 milliseconds. The second, called slow, integrates over a longer
period of time, about 2000 milliseconds.)
Derived Measures
Derived measures were first proposed because in some circumstances it
was felt that a better correlation with people's assessment of the
magnitude of a sound can be obtained by using a somewhat more
complicated combination rule to aggregate the sound pressure levels at
the different frequencies.
Derived measures of sound are obtained from physical measurement of
the sound pressure level in successive frequency bands over the entire
spectrum, for example, in an octave or a third-octave band. From these
various sound pressure levels, some combination rule is used to derive an
overall estimate of a level. Two common methods are used to estimate
loudness levels: one is Stevens' Mark VI loudness level (SLL), the other is
Zwicker's loudness level (ZLL).
Stevens' method converts the sound pressure level in each band to an
equivalent loudness. The various loudnesses are combined via a
nonlinear combination rule that gives maximum weight to the loudest
band. Zwicker's method is essentially geometric and attempts to calculate
the total loudness of the sound by integrating the loudnesses over all
frequencies. It is a nonlinear combination, however, since effects of
masking (the fact that one sound can render another inaudible) are
explicitly considered in combining adjacent bands. Both have been used
in international standards for noise, International Standardization
Organization (ISO) R592.
Another measure of noise magnitude is perceived noise level (PNL), a
method devised by Kryter and similar in form to Stevens' loudness level.
It is also employed in an international standard, for aircraft noise, ISO
R507. The vast majority of noise indices employ one of the preceding as
the basic measure of noise magnitude. Because sound is often annoying
simply because it interferes with speech, another derived measure, called
speech interference level (SIL), has also been used on some occasions. It
is an arithmetic average of sound pressure levels calculated in four octave
bands with center frequencies of 500, 1000, 2000, and 4000 Hz.
Finally, Stevens (1972) has designed a revised version of his loudness
level calculation called perceived level of noise (Mark VII), which makes
use of revised equal loudness scales. It has been little used as yet.
-------
Noise Indices 51
SOUNDS THAT VARY
While magnitude is an important characteristic of steady sounds, many
sounds are not steady. The airplane flyover is a prime example. A passing
car, lawn-mower noise, and many sounds created by machinery are
intermittent or vary in intensity as time passes. There are three measures
of noise magnitude that assess the level of a sound whose magnitude is
not constant:
1. Maximum sound level (SLAM)—the greatest sound level during a
designated time interval or event. We use the term to mean the greatest
A-weighted sound level of an event recorded on the fast setting of a
sound level meter. The measured quantity in decibels is denoted L,nax.
2. Average or Equivalent Sound Level (SLAQ)—a sound level typical
of the sound levels at a certain place in a stated time period. Technically,
average sound level in decibels is a mean-square, A-weighted sound
pressure level over the stated time period, unless some other frequency
weighting is specified. Average sound level differs from sound level in that
average sound level gives equal emphasis to all sounds within the stated
averaging period. The measured quantity in decibels is denoted Leq.
3. Sound Exposure Level (SEL)—the level of sound accumulated
during a given time period or event. The SEL is particularly appropriate
for a discrete event such as the passage of an airplane, a railroad train, or
a truck. It is not an average, but a kind of sum. In contrast with average
sound level, which may tend to stay relatively constant even though the
sound fluctuates, sound exposure level increases continuously with the
passing of time. Technically, sound exposure level, in decibels, is the level
of the time integral of the square of the A-weighted sound pressure. The
measured quantity in decibels is denoted LAE. For a very steady sound,
the maximum and average sound level (averaged over the time the sound
is steady) will be nearly the same. For the same sound, the total energy
will double for each doubling of duration and, hence, if the sound is a
discrete event, the sound exposure level will increase 3 dB (3 dB per each
doubling of duration).
Many discrete sounds are so much larger than the background level
that the equivalent or total exposure level can be calculated on the basis
of the discrete events. For example, if the N discrete events are nearly
equal in level and last for T seconds at this constant level, then the sound
exposure for each discrete event is simply:
LAE = Lmax+101ogT;
-------
52 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
If one wished to compute the equivalent sound level for N equal sounds
in a 24-hour period, then, since 10 log 86,400=49.6 dB (86,400 sec in a
24-hour period):
L2\ = SEL+10 log N-49.6 dB.
Interestingly, this approximation can be used for airports with
moderate traffic loads because the noise produced by each landing and
takeoff is so much larger than the background level that the Leq is
dominated by these discrete events even if the averaging period is as long
as 24 hours.
CORRECTION FACTORS
Besides the average or integrated magnitude of a sound, the disturbance
created by some sound is affected by a number of other variables such as
the time of day or the presence of audible tones. These variables, referred
to as correction factors and indicated by the letter F and a number, are
used to amend the noise magnitude to arrive at an index that more nearly
reflects the objectionable quality of a particular sound exposure. (These
corrections are made by adding or subtracting decibels from the original
assessment of magnitude. Thus the correction terms are additive, but
because of the logarithmic character of the decibel measure, they
correspond to multiplication of the base quantities.) This section first lists
the various correction factors, then presents a table indicating which
indices take account of which factors.
There are nine commonly used correction factors:
F1. Duration of the Sound—the length of time during which the sound
is emitted.
F2. Frequency of Occurrence of Noise—a correction that indicates the
number of noise events that occur in a specified length of time, such as
the number of aircraft flyovers during a 24-hour period. This sometimes is
evaluated in terms of percentage of time that a source operates in a given
period.
F3. Discrete Frequency Components in Noise—a correction for the
presence of audible pure-tone components in noise: i.e., distinctive
pitches that are apparent in the source.
F4. Impulsive Nature of Noise—a correction for noise that is
composed of discrete impulses, such as the noise produced by an air
hammer.
F5. Background Noise—the average noise level when the source is not
operating. Some measures of noise magnitude, such as Leq or SEL,
-------
Noise Indices 53
automatically reflect the background sound level. Some indices require
one to calculate explicitly the background level, with the source removed,
and then to add a correction based on the increment caused by
reintroduction of the source in question.
F6. Variability of Noise—a measure of how much the noise fluctuates
over a given time period.
F7. Time of Day—a correction for the time of day in which noise
occurs. Typically, indices impose a penalty for night-time as opposed to
day-time occurrences.
F8. Time of Year—a correction for the season in which the noise
occurs. An index may impose a penalty for a summer exposure as
opposed to a winter exposure because building windows are left open in
the summer.
F9. Previous Exposure of the Community to Noise—a correction that
assumes that communities with previous exposure to noise levels that
approximate the new noise level will be less likely to protest the added
noise, provided that the total noise level is below some maximum value.
Table 3.1 lists these correction factors and indicates which indices take
which factors into account. The first column in Table 3.1 lists the index,
the second column gives the common abbreviation for that index, and the
third column gives the symbol used to denote the quantity. The division
of the table into two parts notes that either A-weighted sound level or
PNL is used as the means of assessing noise magnitude.
RELATIONS AMONG INDICES AND MEASURES OF NOISE MAGNITUDE
Given the dozen or more available noise indices, it is essential to
understand how any one of these indices is related to any other one. Since
the indices are different, the measures they yield for a given source will be
different. The correlation among the different measures is surprisingly
strong (about 0.95 [Young 1964]), and one can often approximate the
relation among the indices by a simple additive constant. Thus, for a
variety of noise sources, sound level A plus 13 dB is approximately equal
to PNL. In fact, the relation depends on the distribution of energy over
frequency for each source. In most cases, the source is an airplane, since
this has been the most frequent noise source measured.
Rather than attempt to relate every index to every other, we will use A-
weighted sound level (LA) or the average sound level (Leq) as our
common basis for comparison. Table 3.2 shows the approximation
between the various measures of noise magnitude and the corresponding
A-weighted level.
-------
TABLE 3.1 Correction Factors Used by Various Noise Indices
Noise Index Abbreviation Symbol
Based on A Level
Single Event
Sound exposure level SEL LAE
Multiple Events
Community noise equivalent level CNEL —
Day-night average sound level1 DNL Ljn
Equivalent sound level1 EQL Leq
c
o
a
a
Fl.
X
x or
X
X
CJ
g
g
o
O
"3
|
3
F2.
X
c
o
1
F3.
J2
O.
F4.
t
•a
1
I"
^A
1
F5.
X
1
•g
>
F6.
I
^
1
F7.
X
X
13
§
ad
t/i
F8.
.1
n
8*
w
CO
p
O
1
F9.
-------
Mean annoyance level Q —
Noise pollution level2 NPL -
Noisiness index NI -
Total noise load1 B
Traffic noise index2 TNI
Based on Perceived Noise Level
Single Event
Effective perceived noise level EPNL -
Tone corrected perceived noise PNLT -
level1
Multiple Events
Composite noise rating CNR -
(Airport-FAA-DOD)
Composite noise rating (community) CNR
Isopsophic index2 N -
Noise and number index NNI -
Noise exposure forecast NEF -
X
X
X
x or
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
i from Pearsons and Bennett (1974).
2Data from Pearsons and Bennett (1974) and Shultz (1972).
-------
56 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
TABLE 3.2 Approximate Relations Among Measures
of Noise Magnitude
Source Measure
Aircraft LB = LA + 3a
LC = LA + 3a
LD = LA+ 6°
SLL = LA + 9b
ZLL = LA + 14C
PNL = LA + 13<*
Mixture
(Manufacturing, Neighborhood,
Vehicle and Aircraft) SLL = LA + 10e
ZLL = LA+lJ
PNL = LA+13e
Broadband Flat Noise LB = LA - 1°
LC = LA-ia
Lp =LA + 8ft
"From U.S. EPA (1974).
*From Young and Peterson (1969).
^Indirect via SLL + S dB = ZLL (see Schultz 1972).
dFrom Peterson and Gross (1972) and Schultz (1972).
eFrom Botsford (1969) and Parkin (1964).
•^From Botsford (1969).
^Frorn Parkin (1964).
^Computed from the area under the weighting curve (see Figure 3.1).
One other set of approximations is extremely useful:
Ldn ~CNEL~NEF + 35 dBaCNR-35 dB
The last relation is less precise and this is indicated by the double
approximation sign.
EVALUATING A SINGLE NOISE SOURCE
Practically the entire discussion has been devoted to measurement at a
given place—the noise near an airport, urban streets, or a given
neighborhood. Attempts to regulate will undoubtedly continue to impose
limits on the noise exposure as monitored in these places. But one can
also measure single sources and attempt to control the total sound in a
given place by monitoring each of the contributing sources.
In some cases this can be reasonably successful and one can actually
-------
Noise Indices 57
predict the noise level at a given place from the individual sources. The
noise near airports is a prime example. If one knows what planes are
using the airport and how often each type is landing and taking off, one
can calculate the contribution of each type of plane to the total noise
exposure with a fair degree of accuracy.
One cannot do as well on a given street corner because all cars are not
alike, and even the same model of automobile may have a quiet or noisy
muffler. In addition, buildings reflect sound and cause reverberation and
may make a car in one lane more noisy than the same car in a different
lane. The testing of vehicle noise is therefore usually carried out in an
open environment, with no structures or buildings nearby.
Trucks are a source of vehicular noise that present special measure-
ment problems. Two trucks of a manufacturer may differ widely in the
noise they generate because of differences in tires, mufflers, cooling fans,
and transmissions. One can, however, predict with reasonable accuracy
the noise a truck will generate—in a given circumstance—if one knows
the various components that the truck uses. Work has begun on a volume
that is tantamount to a noise catalogue that permits one to evaluate the
various options that the truck may select and combine these into a
reasonably accurate prediction of the total noise it will generate per mile
of operation under "typical" driving conditions.
COST ESTIMATES FOR MONITORING
A noise abatement plan necessarily requires a continuing program of
monitoring either of the sound levels at some place or of the sound levels
of some sources to assure compliance. The cost of such monitoring
programs will vary depending on which index is employed.
Table 3.3 lists some of the indices and indicates the equipment one
could use in calculating the particular index. The minimum and
maximum cost for such equipment is also listed. The total cost is difficult
to estimate with any precision because it depends on many circumstanc-
es. For example, whether the index is calculated once or will be
calculated periodically over some longer period, such as a month or a
year, will influence the level of investment in automatic equipment. Large
investments in automatic equipment along with more frequent use will, of
course, yield lower costs per calculation. We have somewhat arbitrarily
designated the total costs as "high," "moderate," and "low." These
correspond to the indicated ranges of dollar cost when we estimate the
cost for a single non-recurrent measurement. (The labor costs are not
more than a few person-hours for any index.) Other estimates for the
-------
58 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
TABLE 3.3 Instruments Required and Cost Estimates for Various Indices
PNLT
liPNL
Sl:L
CNR
NICF
CNI'L
Leg
l-dn
N
0
NNI
Nl
li
TNI
NFL
Minimum Cost ($)
Maximum Cost {$)
300
1,600
5,000
30,000
son
3,000
3,300 3,100
4,000 3,400
x - required
0 - optional
High > $10,000/singk
I0,000> Med > 5,000
Low < 5,000
nonrecurrent measurement
8,000
20,000
iiigh
High
Med
High
High
High
Low-Med
Low-Med
High
High
High
High
Med
Low-Med
High
costs of noise monitoring can be found in a report by Wyle Laboratories
(1976).
REFERENCES
Botsford, J.H. (1969) Using sound levels to gauge human response to noise. Sound and
Vibration 3(10): 16-28.
Parkin, P.H. (1964) On the accuracy of simple weighting networks for loudness estimates of
some urban noise. Journal of Sound and Vibration 2(l):86-88.
Pearsons, K.S. (1973) Systems of noise measurement. Pages 7-24, Proceedings of the
International Congress on Noise as a Public Health Problem Held at Dubrovnik
(Yugoslavia), on May 13-18, 1973. Office of Noise Abatement Control, EPA-550/9-73-
008. Arlington, Va.: U.S. Environmental Protection Agency; PB-241 060/3BE.
Springfield, Va.: National Technical Information Service.
Pearsons, K.S. and R.L. Bennett (1974) Handbook of Noise Ratings. Springfield, Va.:
National Technical Information Service.
Peterson, A.P. and E.E. Gross, Jr. (1972) Handbook of Noise Measurement. 7th edition.
Concord, Mass.: General Radio Company.
Schultz, T.J. (1972) Community Noise Ratings. London: Applied Science Publishers, Ltd.
-------
Noise Indices 59
Stevens, S.S. (1972) Perceived level of noise by Mark VII and decibels (E), Journal of the
Acoustical Society of America 51,575-601.
U.S. Environmental Protection Agency (1974) Information on Levels of Environmental
Noise Requisite to Protect Public Health and Welfare with an Adequate Margin of
Safety. Washington, D.C.: U.S. Environmental Protection Agency; PB 239 429 4BE.
Springfield, Va.: National Technical Information Service.
Wyle Laboratories (1976) Community Noise Monitoring—A Manual for Implementation.
Wyle Research Report 76-8. Prepared for Office of Noise Abatement and Control, U.S.
Environmental Protection Agency. Draft document.
Young, R.W. (1964) Single number criteria for room noise. Journal of the Acoustical Society
of America 36(2):289-295.
Young, R.W. and A. Peterson (1969) On estimating noisiness of aircraft sounds. Journal of
the Acoustical Society of America 45(4):834-838.
BIBLIOGRAPHY
Harris, CM., ed. (1957) Handbook of Noise Control. New York: McGraw-Hill.
National Bureau of Standards (1971) Fundamentals of Noise: Measurement, Rating
Schemes, and Standards. EPA-NTID 300.15. Washington, D.C.: U.S. Environmental
Protection Agency; PB-206 727. Springfield, Va.: National Technical Information
Service.
U.S. Environmental Protection Agency (1971) Community Noise. Prepared by the Office of
Noise Abatement and Control in Cooperation with Wyle Labs, Rockville, Md., EPA-
NTID 300.3. Washington, D.C.: U.S. Environmental Protection Agency; PB-207 124.
Springfield, Va.: National Technical Information Service.
-------
4
Noise from
Transportation
Sources
NATURE AND EXTENT OF THE IMPACT OF NOISE
The noise from transportation and other sources has significant effects
upon a large segment of the U.S. population. These effects are of two
basic types (U.S. EPA 1974):
• hazardous exposures potentially leading to permanent hearing loss,
and
• interference with human activity such as speech communication and
sleep and various forms of annoyance.
It is estimated (U.S. EPA 1972; Bolt Beranek and Newman, Inc. 1976;
Galloway et al. 1974) that as many as 20 million people are exposed to
noises of duration and intensity sufficient to cause a permanent reduction
in their ability to hear. Of these, approximately 9 million are production
workers in industry, 1 million are operators of transportation equipment,
2 million are passengers, and 8 million are operators or passengers of
recreational equipment and other equipment for personal use.
Noise is the most frequently cited cause of annoyance in neighbor-
hoods. In a 1973 national survey (U.S. Bureau of the Census 1975) of
housing conditions, street noise was cited by 34 percent of the 60,000
respondents as a "condition existing in this neighborhood"; 60 percent of
those reporting its presence felt that the street noise was "disturbing,
harmful, or dangerous"; and 18 percent of those reporting the condition
60
-------
Noise from Transportation Sources
61
35
£ 30
m
Q
O 25
CL
w
111
DC
<
o
o
DC
111
a.
20
15
10
Condition Reported
I I Disturbing, Harmful,
I I Dangerous
Would Like to Move
CONDITIONS PEOPLE HAVE ON THEIR STREET
SOURCE: U.S. Bureau of the Census (1975)
FIGURE 4.1 Results of 1973 survey on neighborhood conditions.
felt that "it is so objectionable" that they would "like to move." In
addition, 20 percent of the respondents listed airplane noise among the
conditions characterizing their neighborhood, of whom 34 percent were
disturbed by it and 6 percent wished to move because of it. Figure 4.1
shows the data from this survey.
When the respondents in the survey were asked which attributes they
considered "disturbing, harmful, or dangerous," street noise was the one
most often cited; heavy street traffic was second. (This is indicated by the
heavily cross-hatched middle section of the bars in Figure 4.1.) These
noises were considered more bothersome than the other attributes
reported. Airplane noise, though cited third most often as an attribute of
the neighborhood, was cited as less "disturbing, harmful, or dangerous"
than crime, street lighting, street repair, trash and junk, and odors, but
-------
62 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
more so than abandoned structures, rundown housing, and commercial
activity.
Extrapolating from these data, it has been estimated that more than 41
million Americans find street noise disturbing and more than 12 million
people would like to move because of it. Further, more than 14 million
Americans find aircraft noise disturbing, and 2.6 million would like to
move away from it.
A 1970 survey (Bolt Beranek and Newman, Inc. 1971b) conducted for
the Automobile Manufacturers Association found motor vehicles to be
the most frequently cited source of annoying noise: 72 percent of the 1200
respondents in this survey classified their neighborhoods as noisy, and 55
percent of them cited motor vehicles as the primary cause. Thus, in this
survey, which was conducted in urban areas remote from airports,
approximately 40 percent of those surveyed were annoyed primarily by
the noise of motor vehicles. This result agrees substantially with the 1975
housing survey cited above (allowing for the fact that the housing survey
used a sample of all neighborhoods, urban and rural, near to and far from
airports), in which 34 percent of those questioned cited street noise as a
"condition" in their neighborhood. Other results of this 1970 survey are
reported in Table 4.1.
In 1974, a survey was conducted of 2000 individuals living in 24 urban
TABLE 4.1 Percent Contribution of Each
Source Indicated by Respondents Classifying
Their Neighborhood as Noisy (72% of 1200
Respondents)
Source Percentage
Motor vehicles 55
Aircraft 15
Voices 12
Radio and TV sets 2
Home maintenance equipment 2
Construction 1
Industrial 1
Other noises 6
Not ascertained 8
SOURCE: U.S. EPA (1973a).
-------
Noise from Transportation Sources 63
TABLE 4.2 24 Site Survey Noise Sources
Ranked by Percent of Urban Population
Highly Annoyed at Sites Remote from
Freeways and Airports
Rank
1
2
3
4
5
6
7
8
9
10
11
Source
Motorcycles
Large trucks
Autos
Construction
Sport cars
Helicopters
Constant traffic
Airplanes
Small trucks
Buses
Power garden equipment
% H.A.
11.7
6.9
6.5
5.8
5.4
4.0
3.9
3.4
3.1
2.8
1.9
SOURCE: Fidel!, S. (1977) Analysis of the National
Urban Noise Survey. Bolt Beranek and Newman Re-
port 341 2. Draft submitted to U.S. EPA.
neighborhoods.1 The neighborhoods were selected to be representative of
the distribution of the total urban population and its census tract density,
but the areas surveyed were deliberately chosen to be relatively remote
from both airports and freeways. This survey offers additional evidence
about the types of noise sources found annoying.
Each respondent was asked, "Over the past year, have you heard
. . . [source] in your neighborhood?" The sources listed included three
types of human and animal sounds and 11 categories of sounds of
mechanical origin, and the questionnaire allowed the respondent to add
to the list. Again, the noise of motor vehicles was at the top of the list,
with 12.6 percent of the respondents reporting that it was "highly
annoying." When motor vehicles were subdivided into different catego-
ries, motorcycles ranked first in annoyance, followed by large trucks,
autos, sports cars, constant traffic, and buses. Table 4.2 summarizes the
rank order in terms of annoyance for noise sources in urban areas remote
from freeways and airports. Table 4.3 indicates the other noise sources
specified by respondents. It should be noted that many of these sources
are characterized as intrusive, although they do not occur with sufficient
frequency to affect cumulative noise levels.
Tidell, S. (1977) Analysis of the National Urban Noise Survey. Bolt Beranek and Newman
Report 3412. Draft submitted to U.S. EPA.
-------
64 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
The results of the three questions on people and animal sounds showed
that: 4.9 percent are highly annoyed by "other people's radio or TV";
12.1 percent are highly annoyed by noise from "pet animals"; and 6.8
percent are highly annoyed by "people's voices."
The percentage of people annoyed by radio or TV and by pet animals
showed little variation with population density, but the annoyance with
other people's voices was directly related to population density. Very
little annoyance was indicated at low population densities, but a
considerable amount of annoyance was reported in densely populated
urban areas. These survey results do not differentiate between voices of
people outdoors and voices of people in adjoining apartments, heard
through common walls. They suggest that if the level of traffic noise in
urban areas were reduced by a large amount, it might be replaced as a
source of annoyance by people's voices intruding from outside the
dwelling.
In summary, the public's perception of environmental noise in
residential areas in terms of annoyance is rather well documented. Motor
vehicles are the source of noise most often cited, with aircraft noise
ranking second in the surveys that sample the entire population. Noise
produced by other individuals and animals is third, followed by noise
TABLE 4.3 24-Site Survey Other Sources Rated Highly Annoying
Rank
1
2
3
4
5
6
7
8
9
10
11
Source
Sirens
Fiie trucks
Ice cream trucks
Trash pickup
Gun shots
Trains
Burglar alarms
Auto horns
Chain saws
Hot rods-drag racing
Defective mufflers
Defective pump
Reefer truck
Air conditioner
Model airplanes
Cement mix truck
Welding equipment
No. of Sites
8
7
5
4
4
4
2
3
3
2
1
1
1
1
1
1
1
Total Mentions
14
12
6
4
4
4
4
3
3
2
1
1
1
1
1
1
1
SOURCE: Fidell, S. (1977) Analysis of the National Urban Noise Survey. Bolt Beranek
and Newman Report 341 2. Draft submitted to U.S. EPA.
-------
Noise from Transportation Sources 65
from construction activities, use of power garden equipment, and other
miscellaneous sources.
The surveys have included few questions that permit an analysis of the
public's perception of the relative importance of other effects of noise,
such as hearing loss and property damage. Such questions should clearly
be included in future surveys.
GOVERNMENT REGULATION OF NOISE
The significance of the effects of noise on the public has been recognized
by Congress, many state legislatures, and various city councils. Several
agencies, including the Department of Housing and Urban Development,
the Department of Interior, the Department of Transportation, and the
Environmental Protection Agency have been given authority by Con-
gress to control noise within specific areas appropriate to each agency's
overall charter.
There are many regulations, both in this country and abroad, designed
to control noise emission of various products (U.S. EPA 1972, 1973b). In
this country, aircraft noise is partly regulated by the Federal Aviation
Administration (FAA) in Federal Aviation Regulation Part 36, and
similar regulations have been adopted by many other nations within the
International Civil Aviation Organization (ICAO). Motor vehicles,
including trucks, buses, automobiles, and motorcycles, are regulated in
terms of noise standards in several states and a few cities, as well as in
many other nations. In addition, 43 states require that all vehicles on
highways be equipped with mufflers. Snowmobiles, motor boats, and
other recreational vehicles have been regulated by several states and
cities, and property maintenance equipment, such as power lawn mowers,
have been regulated by cities.
Equipment associated with construction noise has been regulated in
several cities and in some other countries. In the United States, there are
regulations limiting the total noise emanating from construction sites as
well as the hours during which construction activity can proceed. Such
regulations are issued by the General Services Administration for federal
projects, by at least one state, and by many cities and foreign countries.
External industrial noise has been regulated by many cities and towns,
and, recently, power plant noise has been regulated by a few states. Many
other products are regulated by local authorities, specifying the times and
places that they may be used.
Although these regulations do not provide a basis for a quantitative
evaluation of the major sources of noise, their existence does, in effect,
constitute a listing of the products whose noise emissions are considered
-------
66 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
serious. This list is quite consistent with the conclusions of the surveys
described in the previous section.
THE NOISE ENVIRONMENT
Most of the available data on levels of environmental noise in residential
areas were obtained outdoors. Such data are useful in characterizing
neighborhood noise, in evaluating the noise emitted by identifiable
sources, and in relating the measured values to those calculated for
planning purposes by theoretical noise distribution models. For these
purposes, the outdoor noise data have proved more useful than
information about indoor noise levels because the indoor noise levels are
affected by variations from building to building in the amount of
reduction in sound from outdoor levels. This variation among dwelling
units is a consequence of differences in type of construction, interior
furnishings, orientation of rooms relative to the noise, and the manner in
which the dwelling unit is ventilated.
The range of outdoor sound levels in the United States, using the
Day/Night Sound Levels (Ldn) index, is very large. The quiet end of the
spectrum is from 20 to 45 dB for a quiet wilderness area. (This may be
compared with recent estimates of the noise from rainfall, depending
upon geographical location and other local factors.2 But not all
wilderness areas are quiet: a measurement approximately 25 feet from a
mountain waterfall of a small canyon stream in Wyoming gave an Ldn of
approximately 85 dB [Garland et al. 1973].) At the other end of the
spectrum, sound levels of 80-90 dB are found in the most noisy urban
areas, and still higher levels are found within the property boundaries of
some governmental, industrial, and commercial areas not accessible to
the public. The measured variation in day/night average sound levels
outside dwelling units, for example, ranges from 44 dB on a farm to 89 dB
outside an apartment located next to a freeway. Some examples of these
data are summarized in Figure 4.2.
The sound levels inside dwellings are produced by noise generated
both outside and inside the dwelling, the latter being composed of noise
produced directly by human activity, appliances, and heating and
ventilating equipment. In a 100-site EPA survey of urban noise
(Galloway et al. 1974), the inside Day/Night Sound Level averaged 60.4
2Keast, D. (1976) Summer Acoustic Environment of the Jamesport and Shoreham Sites.
Bolt Beranek and Newman, Inc. for Environmental Engineering, Long Island Lighting
Company. BBN Report 2656. (Unpublished)
-------
Noise from Transportation Sources
67
QUALITATIVE
DESCRIPTIONS
DAY-NIGHT
SOUND LEVEL
DECIBELS
_90_
City Noise
(Downtown Major
Metropolis)
OUTDOOR LOCATIONS
I[I_ Los Angeles — 3rd Floor Apartment Next to
- Freeway
Los Angeles — % Mile from Touch Down at
Major Airport
QQ
I Los Angeles — Downtown With Some Con-
struction Activity
Harlem — 2nd Floor Apartment
-70—
-|~~ Boston — Row Housing on Major Avenue
Small Town and :+-
Watts — 8 Miles from Touch Down
at Major Airport
_\ Newport - 3.5 Miles from Takeoff at
Small Airport
Los Angeles — Old Residential Area
EE\
Fillmore — Small Town Cul-de-Sac
_50_V
iet Suburban _|_T\ San Diego — Wooded Residential
California — Tomato Field on Farm
-40-
SOURCE: U.S. EPA (1974)
FIGURE 4.2 Examples of outdoor day/night sound level differentials in dB (re 20
micropascals) measured at various locations.
-------
68 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
dB with a standard deviation of 5.9 dB while the outdoor Day/Night
Sound Level averaged 58.8 dB with a standard deviation of 3.6 dB. In the
survey, continuous measurements were made during a 24-hour period in
12 houses (excluding areas where significant amounts of noise were
contributed by freeways and aircraft). This implies that the sounds in the
homes were composed to a considerable degree of internally generated
noise.
Indoor sound levels vary significantly among homes, as indicated by
the data shown in Figure 4.3. The hourly equivalent sound levels reach an
average minimum of approximately 36 dB during the hours between 1
a.m. and 6 a.m. This minimum level is probably governed by outdoor
noise in the majority of situations. During the daytime, the hourly
equivalent sound levels have a range of more than 30 dB, depending on
the type of activity. Thus, during waking hours, outdoor noise sets a lower
bound on indoor noise. Where the outdoor Day/Night Sound Level is
CD
I
o
L_
O
E
o
CM
CD
80
70
60
50
40
30
r
. :
• . . s ' ' : i :
.... :;•.••.;•
. • • . « .
'• * . ! \
.
' I > .
•: • ; •
f -. .
... . i .
••'• %
,-•»:
! .'
• •
I'.
•
•
•
' J i
201
0 3
MIDNIGHT
12
15
18
21
NOON
HOURS OF DAY
24
MIDNIGHT
SOURCE: Galloway etal. (1974)
FIGURE 4.3 Noise inside living areas of 12 homes-values of hourly equivalent sound
level as a function of hour of day.
-------
Noise from Transportation Sources 69
less than 65 dB, this lower bound is significantly below the average level
of internally generated noise, since the average noise reduction for houses
with windows open (2 square feet of opening) is approximately 15 dB and
with windows closed is 25 dB (Society for Automotive Engineers, Inc.
1971).
ESTIMATE OF THE CURRENT NATIONAL EFFECTS OF NOISE
The number of people in the country affected by environmental noise
produced by mechanical equipment can be discussed in terms of six
major sources of noise: urban traffic; aircraft operations; freeway traffic;
construction; rail line operations; and equipment with operators and
passengers. These source categories have been selected for convenience in
quantification of the noise emissions and the number of persons affected
by them in varying degrees. The sounds in a given area do not necessarily
come from only one of these sources. For example, in some areas, urban
traffic, freeway traffic, and aircraft operations each contribute more than
55 dB. Consequently, it is incorrect to estimate the number of people
affected nationally by adding together the number of people affected by
each source. It should also be noted that although most sources generally
fall into only one of the categories, there are some exceptions. The most
notable example is trucks, which contribute to noise from urban traffic,
freeway traffic, and construction.
The estimated cumulative numbers of people living in urban areas in
which the Ldn is estimated to exceed various values are summarized in
Table 4.4. Some of these estimates have been reported previously (U.S.
EPA 1974); the remainder were later obtained for EPA.3
Urban Traffic
By and large, the data confirm expectations, indicating that urban traffic
is by far the most significant contributor of noise levels of intermediate
intensity (Ldn levels of 55 dB), followed by airports and construction,
which have about an equal role. However, more intense noise levels stem
primarily from freeways and aircraft noise.
3Eldred, K. McK. and T.J. Schultz (1975) Comparison of Alternative Strategies for
Identification and Regulation of Major Sources of Noise, February. Unpublished draft for
the EPA.
-------
70 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
TABLE 4.4 Estimated Cumulative Numbers1 of People in Millions Who Live
in Urban Areas Within Which the Yearly Average Day-Night Sound Level
Exceeds Various Values
Day-Night Sound Level in dB re 20 Micro Pascals
General Source
of Noise
Urban Traffic2
Aircraft Oper.
Freeway Traffic
Construction
Rail Line Ops
80
0.1
0.2
0.3
-
-
75
1.3
1.5
0.7
-
-
70
6.9
3.4
1.3
0.5
-
65
24.3
7.5
2.2
2.4
0.3
60
59.0
16.0
3.6
8.7
0.9
55
93.4
24.5
5.6
26.2
2.0
In addition, approximately 11.5 million people may be exposed to levels in excess of
Leq(8) = 7S ^B when operating various types of equipment.
This estimate accounts for approximately 134 million people who live in incorporated
urban areas. It does not include approximately 16 million people who live in unin-
corporated urban areas, nor the 40 million who live in "rural areas" not on farms but
who may be exposed to highway noise.
SOURCE: U.S. EPA (1974) and chapter-footnote 3.
The estimates of people affected by urban traffic are based on a survey
conducted for EPA in the summer of 1973 (Galloway et al. 1974). The
survey measured the 24-hour pattern of outdoor noise at 100 sites in 14
cities, including at least one city in each of the 10 EPA regions. These
data, supplemented by data from previous measurements at 30 other
sites, were correlated with census tract population density in order to
obtain a general relationship between Ldn and population size. This
relationship was then used (Galloway et al. 1974), together with census
data giving population in incorporated urban areas as a function of
population density, to derive the national estimate in Table 4.4.
Aircraft Operations
The estimates of the number of people affected by aircraft operations
have evolved during a series of studies over the past few years (U.S. EPA
1972, 1973a, 1974; U.S. Congress, Senate 1973; U.S. DOT 1971; Bishop
and Simpson 1970; Bartel et al. 1974). The CARD study (U.S. DOT
1971) estimated that 1500 square miles were exposed to levels in excess of
Ldn of 65 dB. (The aircraft noise estimates were originally calculated in
terms of Noise Exposure Forecast [NEF], and have been transformed
into Ldn with the aid of the approximate relationship: Ldn = NEF + 35
[U.S. EPA 1974].) This estimate was confirmed in the Title IV Report
-------
Noise from Transportation Sources 71
(U.S. EPA 1972) by an independent assessment of the calculated
contours for 27 airports (Bishop and Simpson 1970), supplemented by
surveys of additional contours at several other airports. The estimate that
7.5 million people are affected by aircraft noise exceeding 65 dB (Table
4.4) was obtained by multiplying the CARD figure of 1500 square miles
by the national average urban population density of 5000 people per
square mile. The applicability of this figure for population density in
urban areas near airports has recently been confirmed by a DOT study of
23 airports (Bartel et al. 1974). The estimates of number of people
affected by maximal levels other than 65 dB were extrapolated using
relationships derived in a study for the President's Aviation Advisory
Commission (Bolt Beranek and Newman, Inc. 1972) and have already
been reported partially in several EPA documents (U.S. EPA 1973a,
1974; U.S. Congress, Senate 1973).
Freeway Traffic
The estimates of the number of people affected by noise from freeway
traffic are a revision of earlier estimates (U.S. EPA 1972, 1973a, 1974)
based on Ldn contours for a typical urban freeway, calculated in accord
with the new model of freeway noise constructed for the Highway
Research Board.4 The number of people estimated to live within various
Ldn contours was calculated on the basis of the 8000 miles of urban
freeway in the United States and the average urban population density of
5000 people per square mile.
Construction
The estimates of the number of people affected by construction noise are
based on the analysis of construction site noise in the Title IV Report
(U.S. EPA 1972, Bolt Beranek and Newman, Inc. 1971a), together with
an updated data base recently accumulated for EPA.5 The analysis deals
with several types of construction sites, the mix of sources and the
duration of operations, the surrounding population densities, and other
factors appropriate to each type of site.
4Kugler, B.A., D.E. Commins, andW.J. Galloway (1974) Design Guide for Highway Noise
Prediction and Control. Bolt Beranek and Newman, Inc. report for National Cooperative
Highway Research Program Project 3-7/3.
5Patterson, W.M., R.A. Ely, and S.M. Swanson(1974) Regulation of Construction Activity
Noise. Bolt Beranek and Newman, Inc. for Office of Noise Abatement and Control, U.S.
Environmental Protection Agency. BBN Report 2887. EPA Contract 68-01-1547.
-------
TABLE 4.5 Summary of Approximate Impact for Operators and Passengers in Nonoccupational Situations
A-Weighted 8 Hour Equiv-
Sound Level alent Sound
(dB) Estimated An- Level jn dB
nual
Fvr»n_
Source Avg. Max. sure (Hours) Avg.
Snowmobile 108 112 200
Motorcycle 95 110 250
Motorboat (>45 HP) 95 105 100
Chain saw 100 110 20
General aviation
aircraft 90 103 100
Light utility
helicopter 94 100 20
Trucks, personal use 85 100 180
Subways 80 93 400
City buses 82 90 250
Commercial pro-
pellor aircraft 88 100 50
Lawn care
(int. comb.) 87 95 50
School bus 82 86 125
Home shop tools 85 98 30
Highway bus 82 90 50
Automobile 68 90 300
Avg. is median of group of available measures on
Year of 8 hour days has 2920 hours.
96
84
80
78
75
72
73
71
71
70
69
68
65
64
58
various products.
Max.
100
99
90
88
88
78
88
84
79
82
81
72
78
72
80
Fractional impact based on 10 percent dB in excess of identified level of Leq
Noise impact units calculated for average sound 1
Average Frac-
tional Impact
Re Leq (8) = 75
Avg
Level
2.1
.9
.5
.3
0
0
0
0
0
0
0
0
0
0
0
(8) = 7SdB.
Max.
Level
2.5
2.4
1.5
1.3
1.3
.3
1.3
.9
.4
.7
.6
0
.3
0
.5
level. Actual impact may be greater depending upon
Approximate
Number of Noise Impact Units
People Ext
(Million)
1.6
3.0
4.4
2.5
.3
.05
5.0
2.15
11.0
5.0
23.0
24.0
13.0
2.0
100.0
«
correlation
>osed Re LeQ (8) ~ 75
(Million)
3.4
2.7
2.2
.8
0
0
0
0
0
0
0
0
0
0
0
of distribution of individual
annual exposures with sound level of sources.
SOURCE: Principal data source is U.S. EPA (1972).
-------
Noise from Transportation Sources 73
Rail Lines
The estimates of the number of people affected by the noise of rail line
operations are based on the noise levels for locomotives and freight cars
calculated by EPA.6 These levels are used to derive Ldn contours for the
estimated average urban main-line operation of 6 trains per 24 hours (2 at
night) of a national average train that is assumed to consist of 2
locomotives and 40 loaded freight cars traveling at a speed of 33 miles per
hour. The number of people estimated to live between various Ldn
contours is based on the assumption that there are 8000 miles of main
lines in urban areas and that the population density near railroad tracks is
2500 people per square mile, one-half the urban average.7
In addition to the sources of noise listed in Table 4.4 there are many
sources of noise such as snowmobiles, chain saws, lawn mowers, and the
like that can produce sound levels sufficient to threaten hearing loss if
exposure is sufficiently long. The estimated noise characteristics, average
annual exposure, and number of people exposed to many of these sources
is summarized in Table 4.5.
EQUIVALENT NOISE IMPACT
The total effect of environmental noise can be described in terms of two
variables: extensity and intensity. Extensity of effect is measured by the
number of people affected. Intensity, or severity, is measured in terms of
the level of the environmental noise. The relationship between these
elements is portrayed in Table 4.4 in which the number of people are
tabulated as a function of noise level.
For various analytical purposes, it is desirable to obtain a single
number indicating the total noise effect in a specific situation. Such a
number permits one to describe the effect of some increment in emissions
in terms of the percentage changes in the index from its initial value,
rather than having to use a multiplicity of numbers to characterize each
situation.
This has recently led to the design of a measurement procedure called
the equivalent noise impact (ENI) analysis. This method characterizes
the intensity of the effect of a sound by what is referred to as its fractional
impact (FI), which is determined by multiplying a constant by the
number of decibels by which the level of environmental noise exceeds the
"Bender, E.K., R. Ely, M. Rudd, S. Swanson, and G. Fax (1973) Contribution to
Background Document for Rail Carrier Noise Regulations. Submitted to Environmental
Protection Agency, 5 December. (Unpublished)
7See note 6 above.
-------
74 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
TABLE 4.6 Summary of the Estimated Noise Impact Expressed as a
Percentage of the Current National Total Impact for Various General
Sources of Noise, and for Noise Reduction of 5, 10, and 15 dB
Effect of Average Noise Reduction of
General Source of Noise
Urban traffic
Aircraft operations
Freeway traffic
Construction
Rail line operations
Equipment with operator/passenger
Total percentage of current impact
% Reduction from current impact
Current
57.0
16.8
4.45
5.9
0.98
14.8
100
0
5dB
25.5
8.5
2.60
0.89
0.33
6.26
45
55
10 dB
8.57
3.62
1.42
0.016
0.066
2.89
17
83
15 dB
2.14
1.48
0.67
_
-
1.58
6
94
appropriate base level given in the EPA "Levels Document" (U.S. EPA
1974). The three levels that are significant for this discussion are:
1. A Day/Night Sound Level (Ldn) of 55 dB, for outdoor noise in
residential areas with outdoor spaces (a level that produces activity
interference);
2. A yearly average sound level for 8 hours (Leq [8]) of 75 dB, for
individual exposure to noise (a level that threatens hearing loss); and
3. A yearly average Day/Night Sound Level (Ldn) of 45 dB, for noises
generated within residences.
The FI constants used in this report are: 0.05 for effects that involve
activity interference, annoyance, and so on, and 0.10 for effects that
involve direct risk of hearing damage.
Partial effects (FIjP;) are evaluated by multiplying the number of
people exposed to each level of environmental noise by the FI;
corresponding to that noise level. The total ENI is then determined by
summing the individual partial effects on all the people affected.
To facilitate comparison of alternative regulatory targets, the current
total national noise level, as calculated by ENI analysis, has been chosen
on a base and its value set equal to 100 percent. The percent of the total
that is now contributed by each major type of source is indicated in Table
4.6, in the column labeled "current." These contribution figures are
generally consistent with those reported in the previous section: they
indicate urban traffic as the single most important source, outweighing all
-------
Noise from Transportation Sources 75
others together. It is followed by aircraft operation, which contributes
some 17 percent of the total. However, in this calculation, construction
falls far below aircraft operation as a contributor of noise. Indeed, third
place is now taken by equipment with an operator/passenger, such as
motorcycles.
Table 4.6 also provides evaluations of the consequences of reductions
of 5, 10, or 15 dB in the average noise emitted by each source. For
example, it indicates that a reduction of 5 dB from all pertinent sources
would reduce the total effect to 45 percent of its current level; that is, it
would produce a reduction of 55 percent. Similarly, a reduction of 10 dB
would reduce the total impact to only 17 percent of its current value, a
reduction of 83 percent.
It should be noted that the method used in this calcuation is based on
the available correlations between cumulative noise levels and annoy-
ance. These measures appear to correspond reasonably well to the
evidence on annoyance from general sources of noise (for instance,
airports and highways), but they may not give sufficient weight to the
annoyance resulting from infrequent intrusive sounds such as those
caused by motorcycles, power lawn mowers, and barking dogs. There-
fore, it would be preferable if the approach were modified to take explicit
account of infrequent intrusive sounds before the results are used in the
design of any comprehensive noise abatement program.
INDIVIDUAL SOURCES OF NOISE
An evaluation of the prospective effects of any program of noise control
must consider the contribution of each individual type of noise source.
For example, an analysis of a program of urban noise control must
consider the contribution of trucks, buses, automobiles, and motorcycles.
This must take into account their proportions in urban traffic and their
mode of operation in an urban setting. Similarly, evaluation of a program
of freeway noise control must take account of the differences in the mix
and the mode of operation of various types of vehicles.
One way to rank the various sources in order of their noise emissions is
to estimate their total daily A-weighted sound energy for the relevant
mode of operation. Although this estimate is necessarily crude, it
provides an indication of the order of the sources of noise. The A-
weighted daily total sound energy of a group of sources may be
calculated by multiplying the A-weighted sound level emitted by each
source by the number of hours it operates daily and adding the
contributions of all sources. An approximate calculation of this type was
provided in an EPA Report to Congress (U.S. EPA 1972) for many of the
-------
76 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
TABLE 4.7 Sources of A-Weighted Daily Total Sound Energy Greater than
20 kW hr/day (Excluding Industrial Plants, Building Air Conditioners, Warning
Devices)
Sources kW-Hr/Day
Medium and heavy highway trucks 5,800
Aircraft (nonmilitary) 4,860
Locomotives 1,330
Sports cars 1,150
Passenger automobiles 800
Light duty trucks 570
Motorcycles (off road) 500
Motorcycles (highway) 325
Construction trucks 296
Snowmobiles 160
Air compressors 147
Concrete mixers 111
Jack hammers 84
Scrapers 79
Dozers 78
Pavers 75
Generators 65
Lawn mowers 63
Garden tractors 63
Pile drivers 62
Rock drills 53
Inboard motor boats 52
Construction pumps 47
Outboard motor boats 42
Chain saws 40
Snow blowers 40
Pneumatic tools 36
Backhoe tractors 33
Derrick cranes 28
Railroad freight cars 25
Graders 25
Buses (city & school) 20
SOURCE: Eldred, K. and W. Patterson (1973). Rationale for the Identification of Major
Sources of Noise, Bolt, Beranek and Newman, Inc., BBN Report No. 2636, September.
Draft submitted to U.S. EPA.
sources of noise considered in the report. The calculation was based on
the average A-weigh ted sound level at a fixed distance from the source, its
estimated average daily operating time, and the number of sources
estimated to be operating in the United States in 1970.
These estimates have been extended to include all of the sources for
which data are available (U.S. EPA 1972). The results are given in Table
4.7 for sources whose daily total sound output exceeds 20 kilowatt hours
-------
Noise from Transportation Sources 77
(KWh) of energy. The results show that transportation sources—road,
air, and rail—produce the most sound energy per day, followed in
approximate order by the noise of construction equipment, recreational
vehicles, property maintenance equipment, and home appliances. This
order is entirely consistent with the priorities that can be inferred from
the noise regulations previously promulgated by various levels of
government, both here and abroad.
To illustrate better the relationship between sound and mechanical
energy, a graph of the daily sound and mechanical energy generated by
various sources is displayed in Figure 4.4. The diagonal lines indicate the
acoustic efficiency (n), that is, the fraction of mechanical energy that is
transmitted as acoustic energy. This fraction ranges from less than one
billionth (refrigerators) to an amount exceeding one thousandth (model
airplane engines). The majority of sources have an efficiency of about one
millionth. This is true, in particular, of highway vehicles and construction
equipment powered by internal combustion engines. Sources correspond-
ing to points above this line for highway vehicles, n = KT6, are not
usually equipped with noise reduction devices such as mufflers or
enclosures. Thus, the efficiency fraction conveys two types of informa-
tion: the lower the fraction for a source, the more attention has been
devoted to silencing or muffling it; the higher the fraction, the greater the
likelihood that the source will generate noise. A notable example of
sound reduction is provided by muffled power plants, which have
reduced their sound energy output by a factor of about 40,000. Without
such sound reduction, the total daily sound energy from power plants is
estimated to be 3962 KWh/day, which would rank them as the third
largest source of noise.
CONCLUSION
This chapter has examined the evidence on the contribution of a number
of different sources of noise and has described methods that can be used
in evaluating such sources. It has shown that a number of different
studies consistently rank urban traffic noise as the major contributor of
annoying sound with aircraft serving as a significant second source.
Most important, the chapter has confirmed the pervasive character of
sound and the large number of people affected by it. Over 40 million
residents of the United States seem to be disturbed by urban traffic noise
and some 14 million by airplane traffic noise. More than 12 million seem
to be annoyed sufficiently by sound levels in their neighborhoods to
report that they are contemplating moving. Thus, noise would seem
clearly to be imposing a very real and very substantial cost upon
American society.
-------
78 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
> 100,000
o
DC
UJ
z
Q 10,000
z
D
o
CO
LL
° 1000
o
cc
UJ
o.
w 100
cc
D
0
I
| 10
1
^
Q 1
UJ
to 0 1
/ /
- /^
/^ ^x
/ X /
/
s
— / / -^Highway^
X / / Veh icles ^
/ / /
/ / / '
f * s s
• / »Rail / /
—
-------
Noise from Transportation Sources 79
Bishop, D.E. and M.A. Simpson (1970) Noise Exposure Forecast Contours for 1967, 1970
and 1975 Operations at Selected Airports. Bolt Beranek and Newman, Inc. for the
Federal Aviation Administration. BBN Report 1863. AD-712 646. Springfield, Va.:
National Technical Information Service.
Bolt Beranek and Newman, Inc. (1971a) Noise from Construction Equipment and
Operations, Building Equipment, and Home Appliances, EPA-NTID 300.1. Washing-
ton, D.C.: U.S. Environmental Protection Agency; PB-207 717. Springfield, Va.:
National Technical Information Service.
Bolt Beranek and Newman, Inc. (1971b) A Study of Annoyance from Motor Vehicle Noise.
Prepared for Vehicle Noise Committee, Automobile Manufacturers Association. BBN
Report 2112. Detroit, Mich.: Motor Vehicle Manufacturers of the United States.
Bolt Beranek and Newman, Inc. (1972) Aircraft Noise Analyses for the Existing Air Carrier
System. Bolt Beranek and Newman, BBN Report 2218, PB-215 611/5. Springfield, Va.:
National Technical Information Service.
Bolt Beranek and Newman, Inc. (1976) Economic Impact Analysis of Proposed Noise
Control Regulation. Prepared in cooperation with Abt Associates, Inc. for the Technical
Data Center, Occupational Safety and Health Administration, U.S. Department of
Labor. BBN Report 3246. Washington, D.C.: U.S. Occupational Safety and Health
Administration; PB-258 841. Springfield, Va.: National Technical Information Service.
Galloway, W.J., K. McK. Eldred, and M.A. Simpson (1974) Population Distribution of the
United States as a Function of Outdoor Noise Level. Office of Noise Abatement and
Control. Prepared by Bolt Beranek and Newman, Inc. for the Office of Noise Abatement
and Control, EPA-550/9-74-009. Arlington, Va.: U.S. Environmental Protection
Agency; PB-235 022/1. Springfield, Va.: National Technical Information Service.
Garland, W.L., S.J. Hanna, and D.R. Lamb (1973) Ambient Noise, Wind and Air
Attentuation in Wyoming. Pages 127-132, Proceedings, National Conference on Noise
Control Engineering, October 15-17, 1973, Washington, D.C. Edited by D.R. Tree.
Poughkeepsie, N.Y.: Institute of Noise Control Engineering (Noise/News).
Society of Automotive Engineers, Inc. (1971) House Noise—Reduction Measurements for
Use in Studies of Aircraft Flyover Noise. SAE Committee A-21, Aircraft Noise
Measurement, AIR 1081. Warrendale, Pa.: Society of Automotive Engineers.
U.S. Bureau of the Census (1975) Annual Housing Survey; 1973, United States and
Regions: Part B. Indicators of Housing and Neighborhood Quality. Prepared in
cooperation with Department of Housing and Urban Development. Series H-150-73.
Washington, D.C.: U.S. Government Printing Office.
U.S. Congress, Senate (1973) Report on Aircraft-Airport Noise. Report of the Administra-
tor of the Environmental Protection Agency. Committee on Public Works, Committee
Serial 93-8.93rd Congress, 1st Session.
U.S. Department of Transportation (1971) Joint DOT-NASA Civil Aviation Research and
Development Policy Study—Report. Prepared in cooperation with National Aeronautics
and Space Administration. N71-30506. Springfield, Va.: National Technical Information
Service.
U.S. Environmental Protection Agency (1972) The Economics of Clean Air. Annual Report
of the Administrator of the Environmental Protection Agency to the Congress of the
United States in compliance with Public Law 91-604, the Clean Air Act, as Amended.
Washington, D.C.: U.S. Government Printing Office.
U.S. Environmental Protection Agency (1973a) Impact Characterization of Noise, Includ-
ing Implications of Identifying and Achieving Levels of Cumulative Noise Exposure.
Office of Noise Abatement and Control. EPA-NTID 73.4. Washington, D.C.: U.S.
Environmental Protection Agency.
-------
80 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
U.S. Environmental Protection Agency (1973b) Noise Source Regulation in State and Local
Noise Ordinances. Office of Noise Abatement and Control. EPA-NTID73.1. Washing-
ton, D.C.: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency (1974) Information on Levels of Environmental
Noise Requisite to Protect Public Health and Welfare with an Adequate Margin of
Safety. Final Report. Office of Noise Abatement and Control, EPA-550/9-74-004.
Arlington, Va.: U.S. Environmental Protection Agency; PB-239 429/4BE. Springfield,
Va.: National Technical Information Service.
-------
5
Projections
of Transportation
Activity
INTRODUCTION
Today, transportation is one of the most pervasive sources of noise.
Projections into the future suggest significant increases in most forms of
transportion activity: by 1985 there may be more than 430 million
aircraft operations per year compared with 80 million today, 130 million
autos in use compared with 84 million today, and 28 million trucks as
compared with 17 million today. Although the use of public transporta-
tion is expected to increase, its share in overall transportation may
increase by only about 1 percent because of a concomitant growth in
vehicle miles traveled, primarily by auto. While the increase in the level of
noise emanating from each transportation source (using existing facili-
ties—equipment, roads, etc.) could be expected to rise about 3 dB for
each doubling of its transportation operations, the rise in noise levels will,
in fact, be less than 3 dB since new equipment, already affected by
current noise regulations, is quieter than that currently in use.
The magnitude of the social cost of transportion noise depends on the
amount of noise from other sources (which may mask the transportion
noise or accentuate it), the characteristics of the noise path, and the
recipient. Thus, the particular conditions under which a sound is
generated and received can determine how detrimental transportation
noise will be.
Figure 5.1 shows how path and transmission characteristics affect a
sound. Noise can be reduced by (1) reducing the amount generated or
81
-------
82 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
Point Source,
e.g., 1 vehicle
Relative Distances
from Source
Ference in SPL is same
ien 1 and 10ft as
between 10 and 100ft
100
a. Noise attenuation with distance from a point source, such as an individual,
isolated vehicle.
Line Source,
T1
1 1
I
10 '
IU ,
Relative Distances 1
from Source |
1
1
I i i
I 1
' I 1
1 i l
1 1 1
1 |
i !
i j i
1 A III
1 SPL, Sound Pressure
| at Cross Section A-A
1 'it'
i ! jx
i LXi
i ^^"^ '
i~ i '
i i i
.1
i ^^^%
l
]
.1 i I i l
Level
~L— — — -
LP —
^-^^^
e.g., STTBdm or
vehicles
Difference in SPL is sam
between 1 and 10 ft as
between 1 0 and 1 00 ft.
7 Reduction with distance
I / jg one-half that from a
* point source.
100 10 1 0 1 10 100
Location
NOTE: Not drawn to scale.
b. Noise attenuation with distance from a line source such as a stream of vehicles.
FIGURE 5.1 Noise attenuation.
-------
Projections of Transportation Activity 83
emitted (e.g., truck mufflers), (2) increasing the length of the path (e.g.,
locating an airport away from residences), or (3) creating path character-
istics that reduce transmission (e.g., tunnelling of the transit system or
installation of sound barriers or building insulation).
TRAFFIC NOISE
Today, more people are affected by highway and street vehicular noise
than by noise from any other source, and this noise grows directly as the
number of vehicle registrations, miles of road, and average speeds
increase. If crowded roadways or national speed limits lower average
speeds, and especially the very high ones, the associated noise will also
decrease.
Although programs for the reduction of noise on new facilities are
being planned and carried out by federal, state, and local agencies and
are likely to be at least moderately successful in reducing highway noise,
much of the troublesome traffic noise is generated on local residential
streets. While it is possible to design programs to reduce noise on such
streets—by rerouting traffic, controlling speed limits and traffic flow
conditions, regulating truck routes, and so on—it is difficult. The
manipulation of traffic flows as a means to reduce the effects of noise is
limited by the pervasiveness of local traffic and the comparatively small
number of routes that are both noise-tolerant and suitable for use, given
the purposes, origins, and destinations of trips. In addition, measures to
reduce traffic noise have to be balanced with their economic effects upon
the community and the desire for mobility.1
Figure 5.2 shows the cumulative distribution of traffic noise, measured
50 feet from the edge of a roadway, by type of vehicle. In addition to
variation by type of vehicle, vehicle noise can also vary drastically with
mode of operation. The difference between the noise made by a cruising
auto and an auto accelerating at open throttle can be 7 to 15 dB. Removal
or modification of noise control equipment can increase the noise level
further by some 10 or 20 dB. In addition, the type of tires or vehicle can
affect the sound level produced by 5 to 10 dB. Figure 5.3 is a nomograph
for calculating one index of expected noise level—LIO, the level exceeded
10 percent of the time—as a function of several of these factors.
At present, the greatest single source of highway noise is trucks,
primarily large vans and trailer trucks used in interstate commerce. The
average heavy truck cruising at 45 mph produces approximately 86 dB(A)
"Mosbaek, E.J., J.P. Goodrow, and W.C. Kester (1975) Policy and Techniques for Highway
Noise Valuation and Compensation. Jack Faucett Associates, Inc. report for National
Cooperative Highway Research Program Project 11-6.
-------
84 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
100r
"\ \ \
Diesel Trucks
Gasoline Powered Trucks
60
65
70 75 80
SOUND LEVEL dB(A)
Measured 50 Feet from Edge of Road
SOURCE: U.S. DOT (1973:10)
FIGURE 5.2 Cumulative distribution of highway vehicles versus noise level.
at 50 feet, although levels above 94 dB(A) are not uncommon.
Acceleration produces 5 dB(A) more than cruising, and an additional 2
dB(A) is generated on a 3-5 percent upgrade. At lower highway speeds,
where engine exhaust noise is dominant, mufflers can reduce truck noise
to approximately 78 dB(A). At higher speeds, tire/pavement noise
predominates, as shown in Figure 5.4. New rib tires make the least noise,
while pocket retreads are noisiest, and old, worn tires are noisier than new
ones for each class. Therefore, regulation of the type and maintenance of
tires can be an effective means to reduce truck noise.
Another important determinant of road noise is the spacing of vehicles.
Noise from individual vehicles diminishes at a rate of 6 dB for each
doubling of distance. However, a line of closely spaced vehicles produces
both a higher noise level and a diminution rate of only 3 dB for each
doubling of distance. For most highway and urban traffic situations, the
line source model, as illustrated in Figure 5.1b, is the appropriate one.
However, for single intrusive events, such as an individual motorcycle or
-------
Projections of Transportation Activity
"01 234 5 6 7 8 9 10 11 12
DIFFERENCE BETWEEN TWO
LEVELS TO BE ADDED
Instructions and Example of the Use of the Nomograph
2000 ^ VEHICLE
DISTANCE VOLUME
TO
OBSERVER
1. Automobiles: Extend a line from the pivot point through index below the pivot point
representing the speed of the traffic flow to vertical line A. Connect that intersection on
A (e.g. A1) with the appropriate vehicle volume and note where vertical line B is inter-
sected (e.g. B1). Connect the intersection on B with the distance to the observer, and
read the predicted LIQ from the scale. In the example shown, if automobiles are travel-
ing at 50 mph at a volume of 3000 cars per hour, an observer 200 feet from the flow
would be exposed to an L10 of 66 dBA.
2. Medium trucks: Proceed as with automobiles for the first step, but multiply the vol-
ume by 10 for use in the nomograph. Thus, the example shown also corresponds to
medium trucks traveling 50 mph at a vehicle volume of 300 trucks per hour with the
observer at 200 feet from the traffic line.
If the medium trucks and automobiles are traveling at the same speed, their volumes
can be combined. The example could be used for a vehicle volume of 2000 automobiles
and 100 medium trucks per hour to yield the 3000 vehicle volume figure.
3. Heavy trucks: The estimation of noise from heavy trucks is performed in similar
fashion except that the speed indices to be used are above the pivot point. If the L-gg
for heavy trucks is to be combined with that for medium trucks and automobiles, the
dB addition scale in the corner of the figure should be used.
SOURCE: NRC (1977) Highway Noise-A Design Guide for Prediction and Control.
Transportation Research Board, National Cooperative Highway Research Program
Report 174. Washington, D.C.: To be published by the National Academy of Sciences
in 1977.
FIGURE 5.3 Nomograph for estimating traffic noise as a function of type of vehicle,
speed of traffic, traffic density, and distance from the roadway.
-------
86 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
Pocket Retread
Truck Noise Sources
Cross Bar-D Ancillary Equipment Exhaust
Load
1__J I I I I I
SPEED MPH
Data for Single Axle Truck at 50 ft. from Centerline of Vehicle Travel
Fan
Engine
SOURCE: U.S. DOT (19731
FIGURE 5.4 Relative level of engine noise compared to levels of various types of tires.
garbage truck, the point source model, as illustrated in Figure 5.la, is
applicable.
AIRCRAFT NOISE
Aircraft noise, though it affects fewer people than traffic noise, generally
has a more intense effect on people under air routes in close proximity to
airports. Figure 5.5 portrays noise levels from the operations of a variety
of jet aircraft in current use. New facilities, using careful airport site
selection and design, compatible zoning, buffer zones, and noise barriers,
can significantly reduce the effects of aircraft noise. For existing facilities,
land use controls, buffer zones, and barriers may also be useful, but it will
probably be more costly and more difficult for them to be adopted.
Aircraft noise will be a growing problem at small airports as the use of
private jets and general aviation increases, and current growth patterns
are likely to make it difficult to implement local controls.
RAIL NOISE
Rail transit systems are another source of noise, but one for which there
has been significant progress in abatement. Welded rail can provide an
improvement of 6 dB(A) or more over bolted rail sections; rail and wheel
maintenance can yield another 5-dB improvement. Larger radius turns
-------
Projections of Transportation Activity
Sideline Noise
.35 NM
.25 NM
707,
727,
747,
737,
707-1OOB
DC-8-30
DC-8-61
Concorde
DC-8,
DC-9,
CV880
DC-10
747-100
O
o
747-200
O DC-10
707-300B
FAA Certification Limits (Part 36 FAR)
J_
I
I
I
J
100 200 300 400 500 600 700 800
MAXIMUM GROSS WEIGHT 1000 Ib.
87
z 130
0
III
H 12°
o
o: _
2S 110
>-
i UJ
fe j 100
UJ UJ
u. >
U- UJ
ui -1 on
707-1 OOB
\ 707-320B
DC-9-30 72°B0*> ^^-8-61^ Concorde
\737-100 '
- 0" fO 727-200
DC-9-10/^ ^
727-1 00>.
-
'ODC-8-62
O747
r
O DC-10, L-1011
FAA Certification Limits (Part 36 FAR)
.1. 1 1.
I I I I I
100 200 300 400 500 600 700 800
MAXIMUM GROSS WEIGHT 1000 Ib.
Take Off Noise-3.5 N. Mi. from Brake Release
§ 130
Q
UJ
> 120
UJ
O
S? 110
UJ Z
5_i 100
UJ UI
u. >
%% 90
727-200
DC-8^1 00DC^r2 H
Q • Concorde
^707-3208
727-100
737-100 ^O,
O707-100B
DC-9
0747
O DC-10, L1011
"IPJO in FAA Certification Limits (Part 36 FAR)
~\ \ I I I I I I
100 200 300 400 500 600 700 800
MAXIMUM TAKEOFF GROSS WEIGHT 1000 Ib.
SOURCE: U.S. DOT (1973)
FIGURE 5.5 Current aircraft noise levels.
-------
88 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
90|-
8-Car Train (600 ft Long)
< 80
CO
-o
01
w 70
_i
Q
Z
D
O
w 60
50
Very Long Train
2-Car Train (150 ft Long)
1
1
50 100 200 500 1000
DISTANCE FROM TRACK CENTER-LINE IN FEET
SOURCE: U.S. DOT (1973)
FIGURE 5.6 Wayside noise level for transit trains of various lengths at 40 raph.
can make a major contribution by reducing flange squeal. Finally,
improved drive systems can reduce noise 5-8 dB at high speeds.
Older systems, which often lack these amenities, are gradually being
replaced or upgraded. However, proper maintenance of new systems is
crucial to prevent significant increases in noise. Rail systems generally
present an excellent opportunity for the use of sound barriers close to the
vehicles. Noise barriers installed along the right of way will reduce noise
levels by 10-13 dB at 50 feet and by 7-8 dB at up to 500 feet.
Figure 5.6 illustrates the behavior of train-generated noise levels as a
function of distance from the centerline. The increase in noise levels
resulting from multiplicity of cars is accentuated with distance. Thus,
noise from a one-car train will attenuate at a rate of 6 dB with a doubling
of distance, but only at a rate of 3 dB with doubling of distance if the
train is long.
-------
Projections of Transportation Activity 89
FUTURE TRANSPORTATION NOISE
GENERAL CONSIDERATIONS
Three important issues must be dealt with in any general discussion of the
noise problems that can be expected from transportation: (1) one must
determine which transportation activities (or modes) and locational or
environmental situations are most likely, initially, to constitute noise
problems; (2) one must estimate the extent to which changes in patterns
of transportation use will alter the number and magnitude of these
problems; and (3) one must consider any changes in transportation
technology, system operations, etc., that are likely to alter the nature or
magnitude of these problems.
LIKELY SOURCES OF NOISE PROBLEMS
With reference to the first issue, it is difficult to specify with certainty
what circumstances will create noise problems because so much depends
on the subjective feelings of recipients rather than on any directly
observable physical or psychological danger to recipients. (There are only
a few exceptions—such as in extremely noisy subways, and in the vicinity
of airports.) In general, however, one can say:
While there are many important sources of intrusive noise, transportation vehicle
noise tends to dominate most residential areas. In fact, the cumulative effect of
the increase in noise intrusion by transportation vehicles is, to a large extent,
responsible for the current general concern with noise (U.S. Congress, Senate
1972).
Some of the ways of measuring the effects of transportation noise on
the community are discussed in Chapter 4. Obviously, more noise energy
will be produced by those modes of transportation that generate higher
noise levels, are more numerous, or operate for more hours. Table 4.7 in
Chapter 4 summarizes this information; Table 5.1 presents it in
somewhat different form. Another measure is the contribution to the
residential noise level—nonidentifiable community background noise—
which is summarized in Table 4.6 in Chapter 4. Still another measure can
be made by estimating the noise levels produced by a single intruding
event for each kind of transportation aircraft and vehicle. This informa-
tion is given in Table 5.2, together with information on the size of the fleet
in 1970.
-------
90 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
TABLE 5.1 Relative Noise Energy by Modes of Transportation
Major Category1
Specific Mode
Aircraft
4-engine turbo fan
2 and 3 engine turbo fan
General aviation
Helicopters
Sub-total
Highway Vehicles
Medium and heavy trucks
Sports cars, imports and compacts
Passenger cars (standard)
Light trucks and pickups
Motorcycles
City and school buses
Highway buses
Sub-total
Railroad Vehicles
Locomotives
Freight trains
High-speed intercity passenger trains
Standard passenger trains
Sub-total
Urban Rail
Rail rapid transit
Pre-WWII trolleys
Post-WWII trolleys
Sub-total
TOTAL
Noise Energy
(Kilowatt-hours/day)
3,800
730
125
25
4,680
5,000
1,000
800
500
500
20
12
7,832
1,200
25
8
1
1,234
6
1
<0.1
7
13,750
Percent of Total3
27.6
5.3
0.9
0.2
34.0
36.4
7.1
5.8
3.6
3.6
0.1
0.1
57.0
8.7
0.2
0.1
<0.1
9.0
<0.1
<0.1
<0.1
0.1
100.1
2Top ten categories that each generate at least 12S kWh per day.
Rounded to nearest unit.
Totals may not add due to rounding.
NOTE: This table duplicates some of the information presented in Table 4.7; it is
based on an earlier, slightly different set of data.
SOURCE: U.S. Congress, Senate (1972), pp. 2-47 to 2-80.
-------
Projections of Transportation Activity
TABLE 5.2 Typical Noise Levels by Kind of Vehicle and Aircraft
91
Major Categoiy
Specific Vehicle or Aircraft
Aircraft
2-3 engine turbo-fan
4 engine turbo-fan
4 engine turbo-fan (wide body)
3 engine turbo-fan (wide body)
V/STOL
Light helicopters
Medium helicopters
Heavy helicopters
STOL aircraft
General Aviation
Small engine prop
Multi-engine prop
Executive jet
Highway Vehicles
Automobiles
Standard
Sports, imports, compacts
Trucks
Light
Medium
Heavy
Buses
City and School
Highway
Motorcycles
Railroad Vehicles
Locomotives
Freight cars
Passenger cars
Urban Rail
Rail rapid transit
Trolley
Range of Typical
Noise Level
(dBA at 50 feet)
85-1 OO1
94-1 05 *
92-1 03 '
84-95 l
65-S62
76-8S2
S2-922
833
67-901
70-931
81-971
64-76
70-87
70-85
80-89
85-95
70-85
75-87
64-95
88-98
80-94
80-90
82-95
68-80
Fleet Size in 1970
(Vehicles)
1,174
815
79
—
2,900
320
40
—
110,500
17,500
900
87,000,000
19,000,000
400,000
NA4
27,000
NA4
1,000
NA4
NA4
At 1,000 feet.
At 500 feet.
Proposed limit.
4 Not available
SOURCE: U.S. Congress, Senate (1972)
-------
92 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
EXPECTED TRENDS IN TRANSPORTATION USE
The second issue that must be dealt with in formulating projections of
transportation noise is calculated growth trends for transportation
activities. The U.S. Department of Transportation has produced trans-
portation activity projections up to 1980, based on expected economic
growth as estimated by the Interagency Economic Growth Project
(IEGP). The determinants of activity are assumed to be population,
income, location patterns, price of transportation, and quality of
transportation. The DOT and IEGP assumptions about future values of
these parameters can be summarized as follows:
The population of the U.S. is expected to grow from 205 million in 1970 to 228
million in 1980 and 255 million in 1990. Income is expected to grow at an average
annual rate of 4.3 percent per year through 1980 and 4.1 percent per year through
1990. Location patterns will continue to favor the use of the private automobile in
intracity travel, while the decline of agriculture and the self-sufficiency of large
metropolitan areas will tend to dampen some aspects of intercity freight growth.
Continued depletion of our supply of fossil fuel and the costs of satisfying
ecological considerations will exert upward pressure on transportation prices. The
quality of transportation between now and 1990 is expected to improve in all
modes; it will be determined in large part by public policy decisions (U.S.
Transportation Department 1972).
The DOT estimates of future transportation activity based on these
assumptions are reported in Table 5.3 for passenger travel and in Table
5.4 for freight transportation. These tables also include estimated ranges
of activity in 1980, based on a revised version of the original DOT data.
Figure 5.7 is a graphic representation of indices of GNP, population, and
aggregate freight and passenger transportation expenditures. The growth
estimates are based on data from 1972 and earlier and results in estimates
near the upper bound for most transportation activities.
PROJECTED PATTERNS OF TRANSPORTATION USE AND NOISE
One conclusion implied by the discussion in Chapter 4 is that an
abatement program for transportation noise should focus on airports,
freeways, and high-speed arterials (especially those with significant
amounts of truck traffic). That conclusion and the projected patterns of
transportation activity described in the preceding section permit us to
draw a number of conclusions about projections of transportation noise.
-------
Projections of Transportation Activity 93
Aviation
The number of domestic passenger-miles is expected to exceed, and
perhaps even double, its current level by 1980 and to double again by
1990. The increase in the number of operations will reflect this growth.
International travel will triple, and then double, in the same periods. In
the original estimates, hours flown in general aviation were expected to
double by 1980 then to increase by 60 percent in the following decade;
the subsequent revision of these estimates drastically reduced the rate of
increase to something above 40 percent by 1980. We can only conclude
that the importance of passenger aviation as a contributing factor to noise
will increase in the, future.
Freight aviation will also become more important. The number of ton-
miles carried by domestic air freight is expected to increase by a factor of
3.5 by 1980 and by an additional factor of 2.5 by 1990.
The number of airports is expected to remain about the same. If the
number of flights increases proportionately to passengers, the noise
problem will increase but by a smaller proportion because the increase in
average aircraft size will probably not be by so much as to absorb
completely the additional passengers. Generally, it is unlikely that any
other carriers will be able to divert any substantial amount of air traffic,
although highly improved ground transportation may be able to do so
along high density routes such as the Northeast Corridor.
Automobile Travel
Total passenger auto vehicle-miles are forecast to increase by between 48
and 107 percent by 1980 and by an additional 32 to 46 percent by 1990.
Although there is wide variation in these estimates and it is likely that the
actual values will lie toward the lower end of these ranges, the increase
will still be significant.
Auto travel may be held in check by a sharp rise in the cost of driving,
such as may result from an increase in the price of fuel or increased taxes
(which may be levied for any number of reasons—from the desire to add
to general revenue to specific purposes such as road maintenance or
public transit subsidy). There is little information on the effectiveness of
price increases in reducing the number or length of trips or in inducing
car-pooling or the use of public transit.
The U.S. Federal Highway Administration (1974) in DOT has
produced some estimates in studying various influences that may affect
their estimates of highway travel.
-------
TABLE 5.3 Trends and Projections: Passenger Travel 1980
Component
GNP
Population
Aviation
Domestic
International
General
Railroads
Auto
Business
Personal
Bus
Urban Transit
Unit of Measure
$billions 1969 constant
millions
billions pax -miles
billions pax-miles
millions hours flown
billions pax-miles
billion VMTs
billion VMTs
billion pax-miles
billion pax
1965
787.1
194.6
57.9
12.6
15.4
17.6
113.5
609.7
23.8
6.8
1970
936.4
205.2
110.2
25.4
25.1
10.8
138.2
748.3
25.4
6.1
19801
a
1481.0
227.5
258.5
79.8
53.9
8.6
231.2
1082
27.0
7.5
b
231.4-278.1
67.2- 89.1
34.92
8.2- 8.6
212.6-247.9
1108-1271
23.3- 29.2
7.0- 7.7
1990
2095.9
254.7
523.2
180.4
83.2
10.2
323.4
1439.7
27.8
9.1
1 , non h r K • • • »• D- I. *• K • • •
range of final demand growth is from 3.5-5.0 percent per year.
4.3 percent annual growth rate of final demand.
SOURCE: U.S. Transportation Department (1972) and Jack Faucett Associates (1973)
-------
TABLE 5.4 Trends and Projections: Freight Transport
Component
GNP
Population
Aviation
Domestic
International
Railroads
Truck
For hire
Intercity
Local
Private
Intercity
Local
Total Private
Intercity
Local
Domestic water
Pipeline
Unit of Measure
Sbillions 1969 constant
millions
billion ton-miles
billion ton-miles
billion ton-miles
billion ton-miles
billion ton-miles
billion ton-miles
billion ton-miles
billion VMTs
billion VMTs
billion VMTs
billion ton-miles
billion ton-miles
1965
787.1
194.6
2.0
704.5
154.0
7.9
110.8
63.9
15.8
18.9
506.3
339.0
1970
936.4
205.2
3.9
740.0
195.6
9.7
132.0
74.3
18.8
21.1
586.3
403.1
1980
a
1481.0
227.5
14.0
966.6
325.2
15.8
212.1
117.8
30.6
35.3
810.5
614.0
b
13- 15.5
890-1050
299.3-322.6
14.5- 16.8
194.1-225.9
108.0-125.2
51.2- 62.1
28.0- 32.6
32.3- 37.5
741-874
569.4-665.1
1990
2095.9
254.7
33.37
1223.1
458.7
21.5
299.7
165.1
42.2
48.0
1041.7
851.8
See note , T S.3
SOURCE: U.S. Transportation Department (1972) and Jack Faucett Associates (1973)
-------
96 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
600
500
400
300
200
100
Freight
GNP
_ Passenger
Population
I
I
1947
1958
1965 1970 1975 1980 1985 1990
YEAR
SOURCE: U.S. Transportation Department (1972:93)
FIGURE 5.7 Index of growth (1947-100).
Price of gasoline. Price elasticity of demand for gasoline is estimated to
be about -0.278. If this point estimate is applicable to large changes, a
100-percent increase in gasoline price would result in a 7- to 14-percent
reduction in vehicle miles traveled.
Characteristics of trip (by type). The various types of trips may exhibit
the following relationships:
(1) A 20-percent decrease in auto shopping would result in a 1.5-
percent decrease in vehicle miles traveled.
(2) A 20-percent decrease in social and recreational trips would result
in a 6.5-percent decrease in vehicle miles traveled.
(3) A 20-percent decrease in work trips would result in a 7-percent
reduction in vehicle miles traveled. However, work and related business
trips (which total 42 percent of vehicle miles traveled) are assumed to be
nondiscretionary and relatively inelastic.
Auto Occupancy. An increase of 50 percent in auto occupancy for work
trips to central business districts (CBD) would result in a 1-percent
decrease in vehicle miles traveled in metropolitan areas. A similar
occupancy increase in non-CBD work trips would result in a 12-percent
reduction in metropolitan vehicle miles traveled.
-------
Projections of Transportation Activity 97
Transit Improvement. Express bus operations to central business
districts (such as those on Shirley Highway, from suburban Virginia to
downtown Washington, D.C.) would reduce metropolitan vehicle miles
traveled less than half of 1 percent.
Bicycle and Walking Trips. Since 62.5 percent of all trips (which
account for 16 percent of all vehicle miles traveled) are less than 5 miles
long, it has been proposed that many of these are candidates for diversion
to walking or bicycle. If all trips under 2.5 miles were diverted, the total
vehicle miles traveled would decrease 6 percent.
Public Transit
It is also possible that public transit can be made more attractive and that
large numbers of persons will consequently be diverted from auto
commuting. Current trends do not go in this direction, for many reasons,
most importantly because: (1) most public transit routes in larger
metropolitan areas are radial and oriented to the central business district;
(2) a small fraction of all trips in such areas (typically less than 10
percent) are oriented to the central business district; and (3) a consider-
able fraction of such trips already involve the use of public transit. While
transit service for short, dispersed trips can be improved, it is unclear
whether this would lure any significant number of motorists. While there
is much debate about the wisdom of national and local transit policies, it
is not clear whether their directions will change.
Perhaps more can be achieved by a change in transportation policies
relating to government investment, pricing, and other operating charac-
teristics. A computer model2 has recently been used to analyze a set of
alternatives and yielded the following results:
(1) Changes in the allocation of investment between highway and
public transit might produce a variation in the number of auto passengers
as a percentage of total trips (the "11 modal split") by as much as 50
percent, starting from a 1972 base.
(2) A 20-percent reduction in total planned 1972-1990 investment
would reduce modal split less than 3 percent.
(3) Increases in auto occupancy might increase auto passenger trips
because highway congestion might be decreased.
Of course, any conclusions depend on the validity of the premises of the
computer model from which they stem.
Werner, E., Assessing National Urban Transportation Policy Alternatives, paper prepared
for presentation at 47th National Operation Research Society of America Meeting, April-
May, 1976. (Unpublished)
-------
98 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
Truck Traffic
Highway freight traffic is expected to grow commensurately with
passenger traffic. Intercity ton-miles are expected to grow 64 percent by
1980 and 41 percent from 1980 to 1990. The number of freight ton-miles
on local highways is expected to grow by 60 percent by 1980 and by 40
percent from 1980 to 1990. Therefore, the highway freight carried by
trucks and the noise produced by it is likely to become increasingly
important nationally, and even more so in urban areas.
The future of truck traffic noise problems is quite uncertain. The major
question is raised by the possibility of the use of trailers on flat cars
(TOFC), known as rail piggyback service, which could replace much
intercity trucking. But TOFC has not grown significantly, and most
studies indicate that regulatory and institutional constraints (e.g., trucker
work rules) now make TOFC economically unattractive.
POTENTIAL TECHNOLOGICAL CHANGES
Since the estimates just described are now a few years old, the obvious
starting point for a reappraisal is the technological advances that have
occurred since then. Foremost is the supersonic transport plane (the
SST). Recent FAA tests indicate it generates more noise than convention-
al sub-sonic aircraft. Should it gain any wider acceptance than now
appears likely, it will contribute to noise levels. In contrast, automobiles
in 1977 are less noisy than their predecessors and the newer, high by-pass
engine jet planes are quieter than the older jet planes.
One class of changes that may prove important are those that reduce
travel. For example, a recent study for DOT (Krzyczkowski and
Henneman 1974) suggests that the use of telecommunications as a
substitute for travel, changes in land use patterns, and rescheduling of
work activity can supplement economic disincentives to travel. The study
concludes that in the short term (1-3 years), a 3-percent reduction hi
urban vehicle miles traveled may be achievable through rescheduling of
work and travel. By 2015, a reduction of one-seventh in vehicle miles
traveled (168 million per day) will be possible if the appropriate
communication substitute technology (essentially video-phone and
information transfer equipment) is perfected.
The DOT commissioned a study (Golding et al. 1970) similar to the
one being undertaken in this chapter. The study, conducted in 1970, was
a cursory survey of technological developments that are in the planning
phase, in initial prototypes, or in the final stages of experimental test and
design. The survey concluded that "technological changes and improve-
ments in transportation will be evolutionary and gradual."
-------
Projections of Transportation Activity 99
UNCERTAINTIES IN ESTIMATES
OF TRANSPORTATION ACTIVITY
The use and heavy reliance on Department of Transportation projections
of future transportation activity by mode and region within the United
States is not meant to be an unqualified endorsement of those projections.
Their use was predicated on the fact that they appear to be among the
best of the projections that are available from public sources and that
contain a full description of the methods used. There are many other
projections of transportation activity, made with many different assump-
tions and methods; some of these are public to varying degrees, but
without full disclosure of the procedures. These other projections were
not included in this study because of the lack of detailed information on
the assumptions or precise methods used.
Projection of future activities in transportation or other sectors of the
economy is much more of an art than a science. There are numerous
factors that will undoubtedly influence future transportation activity that
cannot be taken into account in the projections or that can only be taken
into account by modifications of the projections by human judgment.
The effects of many of these factors—such as the long-term influence of
the recent increases in fuel prices, in which both land use patterns and
travel patterns can be modified, the adjustments presumably being in the
direction of a reduced amount of travel—can only be guessed at. Future
activity may also be affected by specific steps, such as rationing, taken to
ameliorate the adverse effects of future embargos on oil imports. On the
other hand, the development and successful marketing of small automo-
biles that might use considerably less oil-based fuel or might be propelled
by energy sources not based on oil (such as the battery-powered electric
car in which electricity is generated by coal or hydroelectric means) might
considerably relax any constraints on travel, although an increase in such
automobile use would probably be offset by the lowered noise emissions
of smaller or electrically powered vehicles.
Government policies with respect to land development patterns can
have a very significant effect on the total amount and character of
transportation activities, as the suburbanization of population and
employment in the past few decades has revealed. Policies that may have
some very significant influence on transportation include the banning of
automobiles from the central parts of cities, the adoption of traffic control
technology that would significantly expand the capacity of the existing
system of streets and highways, the fostering of the development of
integrated intermodal transportation firms for the movement of freight,
and the pricing of passenger transportation and (in particular, urban
transportation) that would require the recovery of full costs from users.
-------
100 TRANSPORTATION NOISE: MEASUREMENT, SOURCES, AND PROSPECTS
ENVIRONMENTAL IMPACT STATEMENTS
On the basis of the National Environmental Policy Act of 1970 and its
amendments, any project involving the expenditure of federal funds must
be planned so as to minimize the damaging effects of noise and other
unwanted and undesirable consequences. The specific provisions of the
Act are:
These requirements have been implemented by the U.S. Department of
Transportation in various ways. An environmental impact statement (EIS) must
be prepared for each project which is likely to have any negative environmental
effects. It must include a specification of these effects, documentation supporting
the plan or design recommended as most reasonable, and an evaluation of the
positive and negative impacts of the various alternatives considered. The
preparation of the EIS is in addition to other requirements which call for
comprehensive evaluation of the alternative means of achieving the transport
objectives at all levels—plans, systems, and projects. The comprehensive
evaluations include consideration of negative environmental effects, including the
identification of noise as a potential problem and its effects on users of the
transport system, on employees, and on the surrounding areas.
While these requirements specify that noise and other damaging effects
of transportation projects are to be considered, they do not provide clear,
operational guidelines for choice among the available plans or designs.
They provide no guidelines on the magnitude of the expense that should
be incurred to reduce detrimental effects or on the relative priority to be
given to the achievement of other objectives in the attempt to abate noise.
In practice, these decisions are left to the planners and engineers involved
in the project, with the influence of political leaders and the public.
Perhaps this assumes implicitly that their experience serves as a
reasonable proxy for benefit-cost comparisons.
The expectation that noise will produce serious environmental damage
has resulted in the termination of many transportation projects already in
their final planning and design stages, as well as of a few in which initial
construction had begun. This suggests that the earlier environmental
review process and the associated decision-making calculus are not
leading to decisions that the public and elected officials find acceptable,
possibly because of the lack of an operational means to evaluate the
benefits and costs of alternative courses of action.
CONCLUSION
While the effect of policy changes or technological changes is necessarily
uncertain, it nevertheless is the judgment of this Committee that in the
-------
Projections of Transportation Activity 101
next 5 to 10 years it is unlikely that any changes in transportation
activities will be so great as to alter significantly the major identifiable
sources of noise problems in transportation. Even if automobile or air
travel were to be reduced hi this period, for example, it is unlikely that the
reduction would be so great as to make noise problems from these
sources insignificant. In other words, these sources of noise are likely to
remain problems in the future unless specific actions are taken either to
reduce the noise or to ameliorate its adverse effects.
While these trends will perhaps be altered by changes in technology or
major public policy shifts and in the location of noise-sensitive activities,
those changes will not occur by themselves. They will result from
conscious decisions based on consideration of the available alternatives
and their benefits and costs—a subject discussed later in this report.
Finally, if there are significant changes in transportation activity or
patterns, new sources of noise may emerge that may also require
treatment hi a manner similar to that which is suggested hi this report.
Since the recommendations for dealing with noise problems are general,
applying to all modes and contexts, we do not feel it is necessary to
attempt to speculate on possible additional sources of transportation
noise.
REFERENCES
Jack Faucett Associates, Inc. (1973) Transportation Projections: 1970-1980, Revised
Version: 1973. Washington, D.C.: U.S. Department of Transportation.
Golding, E.I., W.D. Velona, and B. Poole (1970) Technological Forcasts: 1975-2000.
Washington, D.C.: U.S. Department of Transportation.
Krzyczkowski, R. and S.S. Henneman (1974) Reducing the Need for Travel, report to U.S.
DOT by the Interplan Corporation. PB-234 665/8GA. Springfield, Va.: National
Technical Information Service.
U.S. Congress, Senate (1972) Report to the President and Congress on Noise. Senate
document 92-63.92nd Congress, 2nd Session.
U.S. Department of Transportation (1973) Transportation Noise and Its Control. DOT-P-
5630.1. Washington, D.C.: U.S. Department of Transportation; PB-218 572/6.
Springfield, Va.: National Technical Information Service.
U.S. Federal Highway Administration (1974) Highway Travel Forecasts. Washington,
D.C.: U.S. Department of Transportation.
U.S. Transportation Department (1972) 1972 National Transportation Report. Present
Status, Future Alternatives. Washington, D.C.: U.S. Transportation Department.
-------
TIT
BENEFITS AND
COSTS OF
TRANSPORTATION
NOISE ABATEMENT
-------
6
Benefits
of Noise
Abatement
INTRODUCTION
In the Noise Control Act of 1972, Congress directed the Environmental
Protection Agency to consider the consequences of noise for the public
health and welfare. This chapter discusses the different types of health
and welfare benefits that would accrue from transportation noise
abatement. No attempt will be made to quantify these: some benefits are
basically qualitative, the magnitudes of other effects are not known, and
others have quantitative effects too indirect to permit effective quan-
tification. The seriousness of most of the effects of noise, and, hence, the
benefits from its reduction, seem clear, but the quantitative evaluation of
each type of benefit must await further research.
In general, there are several kinds of effects that noise produces: direct
effects on the auditory system; indirect effects on other health, social, and
economic variables such as productivity; and effects on annoyance and
the quality of life. It is important to note that programs that reduce the
effects of noise in one domain, even if successful, may not diminish its
effects in the others. For example, policies that minimize direct effects of
noise, such as damage to the auditory system, may not be as successful in
reducing annoyance effects.
105
-------
106 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
HEALTH BENEFITS
REDUCTION OF HEARING LOSS
One of the consequences of a reduction in transportation noise is very
likely to be a reduction in the amount of hearing loss. There are several
general reviews of the effects of noise upon hearing (Kryter 1970, Burns
1973, Miller 1974). They conclude that repeated or long-term exposure to
noise of high intensity results in hearing loss, at first temporary, and
ultimately permanent.
There is still debate in professional circles about the maximum levels of
environmental noise that can be considered safe, but EPA (1974) has
suggested ceilings to protect the public health and welfare. These ceilings
are designed to protect the most susceptible groups in the population and
incorporate a margin of safety. Table 6.1 presents a summary of these
suggested ceilings.
Current transportation patterns generate noise of considerable magni-
tude, in some cases over 85 dB(A), enough to cause permanent hearing
loss with prolonged exposure. As reported in Chapter 4, a large part of the
U.S. population is exposed to the noise; EPA (1974) estimates that 16.5
million people live in urban areas of the United States where the outdoor
average sound levels (primarily generated by transportation sources) are
higher than those that will cause hearing loss in the long run—24 hours a
day over a 40-year period. (An estimated additional 61.6 million people
live in areas where the outdoor sound levels exceed those levels that
interfere with outdoor activities and produce annoyance.)
When people are exposed to noise of high intensity for long periods of
time, their ability to hear is impaired. One common occurrence is a
temporary shift of thresholds: people are less able to detect quiet sounds
after the noise exposure. The more intense the noise and the longer the
exposure, the more severe the shifts are and the longer it will take for a
recovery of normal hearing. The frequency of occurrence of temporary
threshold shifts and the recovery time from them are predictors of
hearing loss—noise-induced permanent threshhold shifts.
As EPA (1971) indicates (see Figure 4.2 above), in some urban
locations ambient daytime noise levels are more than 80 dB(A) over 12-
hour periods. These are well above the levels considered damaging to
hearing with prolonged exposure (cf. Kryter 1970) and clearly implicate
urban noise, particularly from transportation, as a causal agent in hearing
loss. The degree of damage to individuals will, of course, vary with the
amount of time they spend outdoors and the adequacy of the noise
insulation by which they are protected. While transportation noise clearly
-------
Benefits of Noise Abatement 107
TABLE 6.1 Summary of Noise Levels Identified as Requisite to Protect
Public Health and Welfare with an Adequate Margin of Safety
Effect Level1 Area
Hearing loss Leq(24) < 70 dB All areas
Outdoor activity L(jn < 55 dB Outdoors in residential areas and farms
interference and and other outdoor areas where people
annoyance spend widely varying amounts of time and
other places in which quiet is a basis for
use.
Leq(24) * 55 dB Outdoor areas where people spend limited
amounts of time, such as school yards,
playgrounds, etc.
Indoor activity L,jn < 45 dB Indoor residential areas
interference and
annoyance ^eq(24) ^ 45 dB Other indoor areas with human activities
such as schools, etc.
L6q(24) represents the sound energy averaged over a 24-hour period while L,jn repre-
sents the Leq with a 10 dB nighttime weighting.
The hearing loss level identified here represents annual averages of the daily level over
a period of forty years. (These are energy averages, not to be confused with arithmetic
averages.)
EPA has determined that for purposes of hearing conservation alone, a level which is
protective of that segment of the population at or below the 96th percentile will protect
virtually the entire population. This level has been calculated to be an Leq of 70 dB over
a 24-hour day.
SOURCE: U.S. EPA (1974:3) See U.S. EPA (1974:29) for a more detailed description
of these levels.
contributes to hearing loss, it is impossible to apportion a specific part of
the observed hearing loss in the population to that source.
Students of industrial noise have established criteria for the prediction
of hearing loss as the result of continuing noise, but there remain several
sources of variability that make accurate assessments of the cause of
hearing loss difficult to determine after the fact:
a. individual differences in susceptibility to otological damage;
b. hearing loss differentials at different frequencies; and
c. the difficulty of separating industrial noise from other sources of
environmental noise in any individual's life history.
The problems are even more complicated for an analysis of the effects
of transportation noise or, more generally, environmental noise. The
-------
108 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
absence of any continuing survey of hearing impairment does not permit
a yearly estimate of the rate of otological damage. Even relatively
complete sources, such as the recent The Deaf Population of the United
States (Schein and Delk 1974), do not provide data that permit the
separation of cases of congenital and accident-caused hearing impair-
ment from those cases attributable to industrial or environmental causes.
There is another and more important difficulty in determining the
effects of noise on hearing. Current health and welfare levels for both
environmental noise (U.S. EPA 1974) and industrial noise (National
Institute for Occupational Safety and Health 1972) are founded on the
belief that it is the accumulation of noise stimulation that ultimately
impairs hearing. A person who undergoes hearing loss at age 50, even
when presbycusis (impairment of hearing due to advancing age) is
factored out—no simple matter in itself—is clearly manifesting the
consequences of 50 years of acoustic stimulation. While this makes it
exceedingly difficult to attribute a particular hearing loss to any one
episode or source, it implies that high levels of background noise, which
in our urbanized society come primarily from transportation, contribute
and are implicated in almost every case of general hearing loss. (A
detailed discussion of these issues can be found in the volume edited by
Henderson et al. 1976.)
OTHER HEALTH BENEFITS
Reduction of transportation noise may produce health benefits other
than a reduction in damage to hearing: it may affect mental health and
sleep disruption; it may reduce stress and cardiovascular involvement;
and it may even contribute to fetal health.
Mental Health Effects and Sleep Disruption
Intuitively, one might suppose that the intrusion of ambient noise levels
so high as to be continually irksome would, over the long run, produce
deficits in personality organization and functioning. However, there is
little evidence that noise "drives people crazy." Rather, a recurrent
finding is that humans adapt to noise to a remarkable degree (Davis 1948;
Davis et al. 1959; Glass et al. 1969, 1971; Reim et al. 1971). Most of the
studies linking noise to effective functioning center about sleep disrup-
tion, yet even there, adaptation seems quite usual. In many cases, it is the
shorter- rather than the longer-term exposure that is more disruptive.
The effects of noise on sleep are not well understood and no general
conclusion can be drawn. Whether noise rouses a sleeper seems to be
-------
Benefits of Noise Abatement 109
determined not just by the intensity of the noise, but by its spectral
distribution, the stage of sleep during which it occurs, and individual
characteristics of the sleeper. It is not known whether steady sound will
rouse sleepers more often than they usually waken, nor whether such
sound will prevent the onset of sleep. Bursts of sound, which may be more
likely to interfere with sleep, seem to be less effective rousers of sleep-
deprived people (Kryter 1970).
Of the different stages of sleep, one, named REM sleep for its
accompanying rapid eye movements, is thought to be important for
normal functioning; it occurs about two hours a night or 25 percent of the
total sleep time. Once deprived of REM sleep, there appears to be a
compensatory mechanism inducing people to spend a greater amount of
time in this state at another period (Kales 1969). Even if transportation
noise were not an important factor in overall sleep disruption, REM
disruption effects, if they could be established, might well be of special
importance.
General data on sleep disruption by noise in the population are not
available. Current studies (e.g., Lukas and Kryter 1970) indicate,
however, that disruption is an increasing function of age—older people
are more bothered than younger people. Given the gradual aging of the
American population, the problem of sleep disruption will affect an
increasing number and proportion of the population even if there is no
increase in ambient noise levels. In other words, a reduction in
transportation noise would be necessary to keep sleep disruption to its
current level and, in the future, a given reduction in noise may facilitate
the sleep of an increasing number of people.
Stress and Cardiovascular Involvement
In addition to its more general consequences for human health, noise has
effects upon the human cardiovascular system. Some of these appear to
be mechanical, others biochemical. Some investigators (Hattis et al. 1976)
have described several possible mechanisms through which noise can
affect the cardiovascular system. The general thesis is based on the
standard stress reaction model and general adaptation syndrome as
formulated by Selye (1956). This model asserts that all stressors produce
nonspecific as well as specific effects and that the nonspecific stress effect
is the same for all stressors and is cumulative. This suggests that even in
cases where noise is not sufficiently extreme to cause cardiovascular
problems by itself, it may add sufficiently to other nonspecific stressors
affecting an individual to produce such effects.
Empirical evidence about the relationships of noise to cardiovascular
-------
110 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
disease is scanty. The best available studies, those by Jansen (1959, 1969),
were conducted in an industrial setting and indicate that even when
control differences are taken into account, workers in noisy industries
have a significantly higher rate of cardiovascular disease than those in
quiet industries. However, any broad generalization of these conclusions
is unwarranted. A report of the NRC Committee on Hearing, Bioacous-
tics, and Biomechanics (CHABA) concluded:
So-called stress reactions in the human organism when continued for sufficiently
long periods can be physiologically harmful. However, it appears that the
psychological and physiological responses to noise (excluding changes in hearing)
are transitory, that they adapt out with continued exposure to the noise, and
therefore do not constitute harmful physiological stress (NRC 1971).
Pediatric and Fetus Effects
Although there is speculation about the effects of noise both during and
immediately after pregnancy, there is a dearth of information on the
effects of noise by itself or in combination with other stressors. A
CHABA working group on the effects of long-term exposure to noise on
human health has identified this specific gap in information and is likely
to recommend to the National Institute of Occupational Safety and
Health that research on this topic be given high priority.1
The most relevant of the available studies is an epidemiological survey
by Ando and Hattori (1973) that examines retrospectively the records of
women who carried to term. They studied records of over a thousand
births in Japan, comparing those of mothers who resided under noisy
airport flight paths with those who resided in quieter neighborhoods.
Their results must be regarded as suggestive because a lack of informa-
tion about procedures and measures makes it impossible to evaluate the
report fully. They find, however, that even with demographic variables
controlled, mothers experienced ill effects in noisy areas (with effects
starting at levels of 75 dB) at twice the rate of those in quiet areas. In
addition, the entire distribution of birth weights was somewhat lower for
the noisy areas: for example, a 50 percent increase in the proportion of
infants under 2500 grams at birth in the noisy areas. It is difficult to
project the results of this one study to produce a cross-cultural prediction
for the United States, but, if correct, it has implications for fetal and
neonatal care.
The number of infants with low birth weight is an extremely serious
National Research Council (In preparation) CHABA ad hoc work group on the Efiects of
Long-Term Exposure to Noise Upon Human Health. Report of the Committee on Hearing,
Bioacoustics and Biomechanics. Washington, D.C.: National Academy of Sciences.
-------
Benefits of Noise Abatement 111
matter. In the United States, which uses the same criterion of low birth
weight (less than 2500 grams), there are more than 250,000 such infants
born each year. This figure includes both premature infants and those
carried the full 37-week term. Of these infants, 45 percent die in the first
month of life. As a group they account for between 60 and 75 percent of
all first-year infant deaths in the United States. Those that do survive
show residual effects of their neonatal susceptibility to hypoglycemia,
acidosis, renal compensation, hyperbilirubinemia, response to infection
and many other diseases (National Institute of Child Health and Human
Development 1972, 1975). In later life, children whose birth weight was
low are still subject to a higher mortality risk and are more likely to have
physical defects or to be mentally retarded. It is clear that there would be
significant benefits from even a marginal reduction in the incidence of
low birth weight babies that might result from decreased noise.
WELFARE BENEFITS
ECONOMIC BENEFITS
Direct Productivity Increases
Abatement of industrial rather than environmental noise may produce
measurable savings through increases in productivity. There may also be
benefits from reduced transportation noise due to increased efficiency of
output.
Noise interferes with job performance in a number of ways: when noise
may mask a significant signal, when speed communication is required,
when a worker is overloaded with more than one task at a time, and
during vigilance tasks. If outside traffic noise were reduced, a secretary in
an office building next to a noisy street may be better at proofreading for
infrequent errors, may take dictation more accurately and may more
easily be able to act simultaneously as bookkeeper and receptionist
(Broadbent 1957, 1958; Glass and Singer 1972). There is some counter-
vailing evidence. There are indications that for simple, repetitive tasks,
some increase in noise level may serve as a general activator and increase
rather than decrease productivity (cf. Broadbent 1958 for examples).
Since this latter type of task is somewhat more specialized, it is probable
that, on balance, noise reduction would increase productivity.
Noise affects productivity not so much through direct reductions in
output as through higher error rates, greater variability of performance,
and an increased tendency of people to make quick decisions in
-------
112 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
ambiguous situations. These effects are documented and summarized by
Broadbent.2 Further, these effects may occur without those affected by or
annoyed by the noise being aware of the consequences3 (Singer 1976).
Some of the difficulties in trying to determine and to document
productivity effects of noise in any particular situation stem from the
probably small magnitudes of these effects. The evidence (Broadbent
1958, Kryter 1970, Miller 1974) is equivocal. But because of the relatively
small samples used in most laboratory studies and given the magnitude of
the probable errors, these experiments are not sufficiently powerful to
detect differences of the order of 1 or 2 percent between samples. Field
studies of industrial productivity have not been able to untangle the
specific effects of noise from other productivity factors, especiajly when
the noise effects are small.
However, even a productivity loss of 1 percent for workers affected by
noise—which is a value congruent with the effects reported in laboratory
and field studies cited above—can represent a sizable benefit from noise
reduction. However, the productivity increase resulting from the abate-
ment of transportation noise can only result in increased productivity by
those workers in industrial settings that are already quiet. Workers in a
metal manufacturing plant where the ambient noise levels are 85 dB
would not benefit from a reduction in adjacent freeway noise to 70 dB,
but the productivity of workers in urban office buildings or quiet
industries located near airport flight paths may be increased by a reduced
transportation noise.
Indirect Productivity Increases
Noise may affect productivity not only directly through interference with
activities, but also indirectly by influencing motivation and morale and
increasing absenteeism, personnel turnover, and retraining expenses.
Simply put, people may work better in a quieter workplace even though
the noise itself does not interfere directly with their work. This may occur
even when noise increases direct productivity—which may be the case for
simple, repetitive tasks. There is some evidence that the alienating effects
of this type of work are dissipated if the workers can form a friendly,
communicative social group (Schachter et al. 1961, Latane and Arrowood
1963), which would be more likely to occur with less noise. Noise
reduction may also be able to increase productivity indirectly by
minimizing speech interference and permitting easier work and social
2Broadbent, D.E. (In press) Human performance and noise. Chapter 17, Handbook of
Noise Control, edited by C.M. Harris, 2d ed. New York: McGraw-Hill.
3See note 2 above.
-------
Benefits of Noise Abatement 113
communication on the job. This point was made indirectly (e.g., Hattis et
al. 1976) in testimony at 1975 Hearings of the Occupational Safety and
Health Administration of the Department of Labor on a proposed
reduction in the level of maximum permissible industrial noise and was
stated directly in comments on the hearings (e.g., Woodcock 1975).
Even if the benefits of indirect productivity increases from the
abatement of noise are slight, the number of workers involved may be so
large that the aggregate potential savings would be considerable. As in
the case of direct productivity increases, these benefits would accrue only
for those workers whose workplace was relatively free of industrial noise
but affected by transportation noise from outside.
OTHER BENEFITS
Resource Savings
Resource savings occur when money that would have been spent for
noise abatement is saved because noise levels have been reduced by
alternative means. If noise is reduced at the source, receivers do not have
to insulate; if receivers insulate, sources do not have to reduce noise
emissions; if path barriers are erected, neither sources nor receivers must
expend resources. Since there are a number of alternative ways to reduce
noise, each with its associated costs, the choice among them is in part a
decision about who should bear the cost. Both pragmatic and historical
reasons suggest that the costs of control will more likely fall upon sources
of noise than upon receivers. If, as seems likely, the cost of source control
is less than that of the receiver insulation, there will be a net savings.
The issues related to resource savings also apply to savings in
construction costs. Concern for shielding the interiors of buildings from
exterior noise is negligible in the operational designs of buildings and in
current construction practices. Surprisingly, even structures such as
hospitals, for which one might expect noise abatement to be a significant
consideration, do not spend much money on it. A spokesman for the
Veterans Administration hospital construction section estimates that
only about 0.01 percent of the total building cost is spent explicitly for
noise treatment. This amounts to $6,000 out of the roughly $60,000,000
that is spent to construct a 500-bed hospital. In contrast, a study of the
soundproofing of houses in Los Angeles (Wyle Laboratories Research
staff 1970) reports that, for homes with a median value of $35,000, an
average of $4820 was required for a 25-dB reduction in noise. (The costs
of residential noise insulation are discussed in Chapter 8.) It should be
noted that if ambient noise levels were reduced not only by the quieting
-------
114 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
of transportation but by the imposition of noise standards for stationary
sources, large nonresidential buildings might be forced to incur addition-
al expenses to quiet their noise-producing heating, air conditioning, and
ventilating systems.
Systems Benefits
There are two kinds of systems benefits: those that result from a noise
abatement procedure and those that reduce noise as a consequence of
some other procedure, such as energy saving. Some noise abatement
procedures, though costly in themselves, produce offsetting savings. For
example, truck noise can be reduced by using cooling fan clutches that go
on only when engine temperatures reach a certain threshold. At high
speeds, when fans are least necessary, the fans are off, thus eliminating a
substantial amount of truck noise. While the installation of the fan
clutches is costly, there is an associated saving in fuel when the fan is
turned off. Such savings are referred to as systems benefits because they
are an integral part of the abatement process, arising with almost every
means of noise control. They include energy savings by vehicles traveling
at moderate rather than at high speeds, lowered expenses for heating and
air conditioning in better-insulated buildings, and reduced payments in
compensation awards for impaired hearing. It should be noted that many
systems benefits are unplanned or occur indirectly and that one can
rarely calculate directly the savings benefits of them.
Animal and Plant Production Increases
It has been conjectured that noise may have effects not only on human
beings but also on vegetative growth and on animal welfare. Some of
these effects, such as those on the well-being of wildlife or domestic pets,
are probably incalculable. Others, such as the increase in crop yield or
profit from husbandry resulting from noise reduction, are probably, in
principle, assessable in monetary amounts.
There has been very little research on the effects of noise on plants, and
research on the effect of noise on animal production is inconclusive. In
the Memphis State University's review paper on animals (1971), the only
effects even partially documented are those whose importance is difficult
to assess: for example, under noisy conditions hens tend to shift from
brooders to layers. Such an effect may be beneficial for egg producers and
detrimental for chicken producers, but the net consequences to society
are not obvious.
-------
Benefits of Noise Abatement 115
DIRECT BEHAVIORAL AND PSYCHOLOGICAL EFFECTS
Children's Cognition, Learning, and Language
As Mills' recent review (1975) of the effects of noise on children has
clearly argued, children are more likely to suffer from the effects of noise
than adults. One of the primary effects of high ambient noise levels is a
temporary disruption of speech and hearing. For adults, this means the
interruption of organized communication. For children, it means more.
Their speech is less redundant and its meaning is more likely to be lost.
More important, noise also disrupts the learning of language and the
acquisition of the ability to communicate. However, research in this area
is sparse. There are few results that show a relationship between high
ambient noise levels and reduction in language comprehension: Mills
(1975) reports the results of these studies and their shortcomings.
There are two studies that show the relationship of transportation noise
to the impairment of reading ability. Cohen et al. (1973) present evidence
that the noisier the home background of the child, at least at high levels of
noise, the less likely the child is to discriminate phonemes. This inability
to discriminate was related to reading level in the school, and children
from noisier homes performed more poorly on standardized reading tests.
Bronzaft and McCarthy (1975) studied a school situated next to an
elevated railroad. Students whose classrooms were adjacent to the train
tracks did significantly worse in reading than similar students whose
classrooms were on the other, quiet, side of the building. Since the effects
reported by Cohen et al. (1973) were ascribed to noisy homes and those of
Bronzaft and McCarthy (1975) were attributed to a noisy school, the
locations of both schools and homes are relevant.
Annoyance and Complaints
A number of studies have investigated the characteristics of noise
sources, the personalities of people who are annoyed or who complain,
and the mediating factors that influence which individuals are annoyed
and complain. Some of these studies have focused on aircraft noise
adjacent to airports, others on highway and automotive noise.
Overall, it is clear that higher noise levels produce somewhat more
annoyance and more complaints. The relationship between noise
intensity and annoyance, is, however, filtered by the social context in
which the noise occurs. It is not just the physical intensity of the noise
that will produce complaints (at least at levels generated by the common
transportation sources); people's interpretation of the source, reasons,
-------
116 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
and interfering qualities of the noise also affects the extent of annoyance
or the frequency of complaints.
Cederlof et al. (1967) investigated annoyance as a function of the
source of noise and found that even when various vehicles were equally
noisy, automobiles were not considered as offensive as trucks or buses.
Another study by Galloway and Jones (1973) found a similar dependence
on source: for example, they found that at equal noise levels sports car
noise is more offensive than sedan noise.
Mills and Robinson (1961), working with aircraft noise, investigated
what type of interference was most unpleasant. They found speech
interference most disturbing, interference with sleep and rest second.
(The third factor producing annoyance from aircraft noise was fear of
crashes. This is interesting because it is not the sound that is bothersome
but its signal value.) These findings are consistent with others in studies
by TRACOR (1971) and Galloway and Jones (1973) that conclude that
noise is most intrusive when it occurs in the evening and at the recipient's
home. Cederlof et al. (1967) reported that their respondents' annoyance
was a function of their beliefs about the considerateness of the sources.
Those who felt that pilots could avoid the noise but were inconsiderately
producing it were more annoyed.
In the TRACOR study (1971), people who complained about noise in
any of a variety of ways were surveyed. No particular personality pattern
was predisposed to complain about noise. Those who complained were
usually among the more affluent, better-educated members of their
community. Data from surveys suggest that those who complain are not
particularly hypersensitive to noise: that is, they are not bothered more
than residents who do not complain. Kryter (1970) reports that
complaints about noise from given sources diminish over time. It is
unclear whether this represents adaptation to the noise on the part of the
recipients or resignation to the belief that their complaints will produce
no action leading to abatement. Even if the effect reported by Kryter is
interpreted as adaptation, it applies to the direct effects and not to any
indirect effects that may occur (see Glass and Singer 1972 and below).
The TRACOR study (1971) attempted to establish a general model for
prediction of noise complaints. The authors wanted to be able to predict
the effects of different variables on annoyance (without respect to
particular activities disrupted). Their multiple classification analysis
resulted in the following list in order of importance in predicting
annoyance: fear of crashes in the neighborhood; susceptibility to noise;
distance from the airport; noise adaptability; city of residence; belief in
misfeasance on the part of those able to do something about the noise;
and the importance attributed to the airport and air transportation,
-------
Benefits of Noise Abatement 117
generally. Each of these variables in some form or another seems to affect
annoyance from noise. When these variables are combined with a
measure of noise intensity, CNR (Community Noise Rating, see Chapter
3), the prediction of annoyance was quite good: multiple R2 = 0.63. This
means that some 63 percent of the variation in annoyance may be
accounted for by these seven essentially "social" variables and CNR.
Different models that incorporate the same types of variables have also
been proposed. One has been constructed by the National Swedish
Institute for Building Research (1968), using a correlational framework,
and another, using a multiple regression-path analysis, has been
suggested by Leonard and Borsky (1973).
All these studies are beset by one major confounding effect. Those who
live in particularly noisy circumstances are most likely to have less
education and lower income and to include proportionally more non-
whites than the general population. They also tend to complain less—and
there is no evidence that their lower level of complaints reflect less
sensitivity to the noise. In one form or another, the primary explanation
offered for the relatively low volume of complaints from those most
severely bothered by noise relates to social control. Those who are more
educated and more affluent, it is argued, are more likely to feel that they
have the power to control or at least to influence their own destinies.
Thus, even though less severely affected by noise, they are more likely to
take action when disturbed because they are more likely to believe that
such actions will have some effect. To the extent that the system
producing noise is at all responsive to their complaints, this will be a self-
fulfilling belief. And the noise affecting them is as likely to be diverted to
the non-complaining, lower socioeconomic status groups as it is to be
abated.
Subjective Weil-Being
In addition to specific annoyance or complaints, ambient environmental
noise produces a reduction in what has been termed subjective well-
being, an aspect of what is called the quality of life. Besides disturbing
specific activities, noise also has aesthetic costs. The arguments against
allowing motorcycles or snowmobiles in wilderness areas are based as
much on aesthetic damage to the environment as on physical damage.
Effects of abatement on the quality of life are implicitly incorporated in
some of the benefits already discussed. For example, one of the
consequences of damage to the auditory system is a reduced sense of
communication with other people. It is not only speech that becomes
more difficult; use of the communications media such as radio, television,
-------
118 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
or telephone become progressively more difficult. Music becomes less
audible, and even participation in other activities such as sports and
games, to the extent that they involve speech and vocal communication,
becomes harder.
One step removed from the diminished quality of life for those with
unpaired hearing is the diminished quality of life of those with normal
hearing. Ambient environmental noise may move people indoors and
away from outdoor recreational activities, it may contribute to unwilling-
ness to use central cities, and it may bring about a noise escalation of its
own. In order to talk above higher noise levels, people must speak louder.
This in turn forces other sounds to grow louder, further increasing
background sound level and so escalating the cycle.
Despite the apparent agreement that subjective well-being is adversely
affected by sounds less intense than those which cause auditory damage,
methods for the measurement of these changes are not fully developed
and, consequently, the data collected are less than compelling. The main
approaches that have been used to study this problem have either tried to
use objective indicators of quality of life or have sought to assess
subjective well-being directly.
The report of the National Planning Association (Terleckyj 1975) uses
quantitative objective measures to assess quality of life (with reference to
noise). Among these are some economic indicators, such as the income at
the 20th percentile as a percent of the income at the 90th percentile. It
also uses social indicators, such as the number of hours per person per
year of discretionary time. At least in theory, it is possible to construct
similar noise-effect indices, such as the cumulative noise level at the 20th
percentile as a percent of the 90th percentile or the average number of
days per person per year above a particular Leq. However, even if
objective indices of quality of life can be defined, their subjective
interpretation still remains a problem. In other words, objective indica-
tors are not an equivalent substitute for an individual's satisfaction. Thus,
it is possible for people living in a central city to experience a rise in real
income, an increase in the average education of their family, to receive
any number of other social benefits, and yet feel less satisfied than they
did a decade earlier.
A second class of studies uses interviews of a sample of the population
to try to obtain a direct measure of subjective well-being. The surveys
attempt to assess directly how happy people are, how well they are
adjusted to their environment, or how much stress they are experiencing.
Examples of this approach are the studies by Bradburn and Noll (1969)
and those by Campbell et al. (1976). Campbell et al. use three dependent
measures: an overall happiness rating, a stress rating, and what they call
-------
Benefits of Noise Abatement 119
domain satisfactions. These reflect an individual's satisfaction with his or
her current status in a given number of specified areas or domains.
Although these procedures are well suited for overall measurement of
subjective well-being, the work is global in outlook. They embrace areas
so broad that noise never enters explicitly. The Bradburn studies (1969)
consist of interviews studying psychiatric adjustment, and as noted
above, few if any direct links have been found between noise and lack of
adjustment.
If none of the available studies permits a useful evaluation of the effects
of noise upon subjective well-being, what sort of study would? The study
by Campbell et al. (1976), the most complete attack on the subject that is
available, provides a model for what should be done. Presumably if their
domain satisfaction concept were broadened and particularly appropri-
ate samples were chosen, i.e., groups affected by noise as well as
appropriate control groups, then multiple-regression techniques might be
able to shed some light on this subject.
While further development of the survey techniques would contribute
to our ability to evaluate subjective well-being directly, such improve-
ments are necessary but not sufficient conditions. Even if these tech-
niques were perfected, two other issues would remain. First, there is the
distinction between prevalence and incidence: the incidence of a
characteristic is measured by the percentage of the current population
that will exhibit that characteristic at some time in their lives; the
prevalence is measured by the percentage of the population that displays
it now. Suppose, for example, that during their lifetime 50 percent of the
population will suffer from a severe toothache, but that at any given time
less than 1 percent will be afflicted. It is unlikely that a global assessment
of life quality would pick up enough toothache sufferers to investigate the
relationship between toothache and subjective well-being. On the other
hand, a study of a sample of people visiting dentists' offices would enable
an investigator to evaluate the disruptiveness of toothaches, but would be
likely to overstate the role of toothache in American society. The parallel
with noise is obvious.
A second issue relates to level of aspiration. Quality of life and
subjective well-being are probably not measurable on an absolute scale,
but rather are assessed by individuals in terms of their own expectations
and standards. A rising level of aspiration may make an unchanging or
even a more slowly improving quality of life seem to be deterioration;
comparison with the fortunes of others may alter people's assessment of
their own well-being. Consequently, in an evaluation of shifts in quality
of life or in annoyance over time, measurements of any shifts in
-------
120 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
aspirations and expectations may also be necessary. Campbell et al.
(1976) present a more complete discussion of this point.
These issues suggest that there is no readily available method, nor is
one likely to be designed soon, that would relate subjective well-being or
quality of life to changes in noise levels in the environment. Improved
quality of life, denned as an amalgam of enhanced subjective well-being
and reduced annoyance and complaints, seems to be one component of
the benefits to be derived from noise abatement. It is not easy to study
and does not lend itself to ready quantification or to expression in
pecuniary terms. Yet as has been pointed out effectively by the National
Research Council report (1975) on Decision Making for Regulating
Chemicals in the Environment, any benefit-cost analysis must deal with
these issues implicitly or explicitly if it is to avoid errors that may be
substantial and critical in their significance for public policy.
INDIRECT BEHAVIORAL AND PSYCHOLOGICAL EFFECTS
In addition to its direct effects upon health and behavior, noise may also
produce indirect effects—effects that may not be perceived by those
undergoing them or for which noise is not considered the cause. The
indirect effects can be classified as aftereffects, social effects, and learning
effects.
Aftereffects
People who work or perform a task under noisy conditions and do not
suffer direct impairment from it may nevertheless experience some loss in
their ability to do things after the noise has ceased relative to those who
have performed these tasks in quiet conditions. In about two dozen
experiments (Glass and Singer 1972), people of varying ages soon
adapted to the noise hi the first part of an experiment. They performed
the tests under noisy conditions as well as did people in no-noise control
groups. Yet when performing in a second part of the experiment, after the
intrusive noise had been eliminated, those previously exposed to noise did
worse than those not exposed. They found fewer errors in proofreading,
they did not persist as long in working on difficult or important problems,
and they were not able to process conflicting information as well. These
findings have direct relevance for the relationship between environmental
noise and productivity, for they imply that workers who inhabit noisy
homes will show aftereffects at work, irrespective of their adaptation to
noise at home or to the noisiness of the workplace.
-------
Benefits of Noise Abatement 121
Social Effects
Another class of indirect effects are those relating to social behavior.
Although noise may not interrupt task performance, it may have
consequences for various kinds of voluntary behavior. For example, two
laboratory studies had subjects administer electric shocks to someone
presumably engaged in a learning task (Geen and O'Neal 1969, Geen and
Powers 1971). They administered, at their discretion, a number of shocks
and had a choice of shock intensity. Those who administered the shocks
under noisy conditions gave a greater intensity of shock than those who
administered them under quieter conditions.
In a coordinated laboratory and field study (Mathews and Canon
1975), the effects of noise on altruism or voluntary helping behavior were
studied. In the laboratory situation, an experimental confederate dropped
an arm-load of books. The subject was less likely to help pick them up
when ambient noise levels were relatively high than when they were
relatively low. In a field replication, the confederate dropped an arm-load
of books when walking past a lawn mower. Subjects were considerably
more likely to pick up the books when the lawn mower was turned off
than when it was operating at a level of 84 dB(A). These effects are but
two of a class of normative rules likely to be disrupted by noise: types of
behavior influenced by modeling and imitation. One occurrence of noise-
influenced aggression may set the model of behavior for many others and
one instance in which helpfulness or altruism was inhibited may serve as
a standard for future acts. Thus, a small number of direct events may
affect large numbers of people.
Learning Effects
Whether noise affects learning directly is arguable, but noise is implicated
in incidental learning. People often have tasks that require them to learn
or process information about their environment. These can range from
the learning of people's names at a cocktail party to specific employment-
related materials. Though noise is unlikely to affect the direct learning, it
will reduce the peripheral information processed. Thus, in a laboratory
study, when subjects were presented with slides, each of which contained
a four-letter word in the center surrounded by three-letter words,
differences in noise level produced no differences in their ability to learn
the list of four-letter words. Those who learned under noisy conditions,
however, learned only the four-letter words; the control subjects learned
the three-letter ones as well (O'Malley and Poplawsky 1971). Noise
appears to produce concentration upon the primary learning task at the
-------
122 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
expense of the secondary, and in this case implicit, task. Since it is likely
that much of our everyday knowledge and information is acquired not
directly but indirectly, high ambient noise levels may require increases in
effort for people to reach given levels of knowledge and information.
It should also be noted that part of the failure of many investigators to
find general learning and performance effects of noise may be attribut-
able to their use of global or general measures. A finer-grained analysis
might reveal systematic noise effects. Thus in a study in which people
were required to proofread under noisy conditions (Weinstein 1974), their
general accuracy scores did not change. However, their ability to detect
spelling and mechanical errors increased while their ability to recognize
faulty grammar decreased. The net effect of noise was to focus attention
more effectively and carefully on less demanding problems and to leave
overall performance unchanged.
Most of the effects of noise on learning or task performance can be
subsumed under the general model put forth by Broadbent (1958,1971).4
Humans have a limit on their capacity to process information; noise
lowers this limit. If a task is well within this limit after a short adaptation
period, noise will have no direct effect. But if a task comes close to the
limit of an individual's information processing capacity, the addition of
noise may reduce this capacity below that required by the task.
SUMMARY
This discussion of the benefits of noise abatement provides an overview
of the changes that would occur and the areas in which benefits would
occur. Abatement of transportation noise would result in a reduction of
hearing loss, a reduction of non-auditory health effects, a decrease in
speech interference, and, maybe, a decrease in sleep disruption. Noise
abatement would also have a beneficial effect on worker productivity
through a variety of mechanisms, it would increase learning by children
living or studying in settings with high levels of noise, and, to some extent,
lessen disruptive social effects. The reduction of noise can also be
expected to reduce people's annoyance and increase their subjective well-
being. In short, it would improve physical and psychic well-being and
probably lead to an improvement in social relations: in a variety of ways,
it would contribute to the quality of life.
4See note 2 above.
-------
Benefits of Noise Abatement 123
REFERENCES
Ando, Y. and H. Hattori (1973) Statistical studies on the effects of intense noise during
human fetal life. Journal of Sound and Vibration 27:(1)101-110.
Bradburn, N.M. and C.E.Noll (1969) The Structure of Psychological Well-Being. Chicago:
Aldine Publishing Co.
Broadbent, D.E. (1957) Effects of noise on behavior. Pages 10:1-10:34, Handbook of Noise
Control, edited by C.M. Harris. New York: McGraw-Hill.
Broadbent, D.E. (1958) Perception and Communication. London, New York: Pergamon
Broadbent, D.E. (1971) Decision and Stress. London, New York: Academic Press.
Bronzaft, A.L. and D.P. McCarthy (1975) The effect of elevated train noise on reading
ability. Environment and Behavior 7(4):517-527.
Bums, W. (1973) Noise and Man. 2nd Ed. New York: J.B. Lippincott Co.
Campbell, A., P.E. Converse, and W.L. Rodgers (1976) The Quality of American Life:
Perceptions, Evaluations and Satisfactions. New York: Russell Sage Foundation.
Cederlof, R., E. Jonsson, and S. Sorensen (1967) On the influence of attitudes toward the
source on annoyance reactions to noise: a field experiment. Nordisk hygienisk tidskrift
48(2):46-55.
Cohen, S., D.C. Glass, and J.E. Singer (1973) Apartment noise, auditory discrimination and
reading ability in children. Journal of Experimental Social Psychology 9(5):407-422.
Davis, R.C. (1948) Responses to "meaningful" and "meaningless" sounds. Journal of
Experimental Psychology 38:744-756.
Davis, R.C., A.M. Buchwald, and R.W. Frankmann (1959) Autonomic and muscular
responses and their relation to simple stimuli. Psychological Monographs: General and
Applied 69(20): entire issue whole no. 405.
Galloway, W.J. and G. Jones (1973) Motor vehicle noise: identification and analysis of
situations contributing to annoyance. Pages 785-803, Proceedings of the International
Congress on Noise as Public Health Problem. Held at Dubrovnik (Yugoslavia), on May
13-18, 1973. Office of Noise Abatement and Control, EPA-550/9-73-008. Washington,
D.C.: U.S. Environmental Protection Agency; PB-241 060/3BE. Springfield, Va.:
National Technical Information Service.
Geen, R.G. and E.C. O'Neal (1969) Activation of cue-elicited aggression by general arousal.
Journal of Personality and Social Psychology 11(3):289-292.
Geen, R.G. and P.C. Powers (1971) Shock and noise as instigating stimuli in human
aggression. Psychological Reports 28(3):983-985.
Glass, D.C., J.E. Singer, and L.N. Friedman (1969) Psychic cost of adaptation to an
environmental stressor. Journal of Personality and Social Psychology. 12(3):200-210.
Glass, D.C., B. Reim, and J.E. Singer (1971) Behavioral consequences of adaptation to
controllable and uncontrollable noise. Journal of Experimental Social Psychology
7(2):244-257.
Glass, D.C. and J.E. Singer (1972) Urban Stress: Experiments on Noise and Social
Stressors. New York: Academic Press.
Hattis, D., N.A. Ashford, G.R. Heaton, and J.I. Katz (1976) Some Consideration in
Choosing an Occupational Noise Exposure Regulation. Prepared by the Center for
Policy Alternatives, Massachusetts Institute of Technology for the Office of Noise
Abatement and Control, EPA-550/9-76/007. Washington, D.C: U.S. Environmental
Protection Agency; PB-251 408/1GA. Springfield, Va.: National Technical Information
Service.
-------
124 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
Henderson, D., R.P. Hamernik, D.S. Dosanjh, and J.H. Mills, eds.( 1976) Effects of Noise
on Hearing. New York: Raven Press.
Jansen, G. (1959) Zur entstehung vegetativer funktionsstorungen durch larmeinwirkung.
(On the origination of vegative functional disturbances due to noise effect.) Archiv fur
Gewerbepathologie und Gewerbehygiene 17:238-261.
Jansen, G. (1969) Effects of noise on physiological state. Pages 89-98, Noise as a Public
Health Hazard, Proceedings. Conference held June 1968, Washington, D.C. Edited by
W.D. Ward and J.E. Fricke. ASHA Report No. 4. Washington, D.C.: American Speech
and Hearing Association.
Kales, A., ed. (1969) Sleep: Physiology and Pathology. Philadelphia: J.B. Lippincott Co.
Kryter, K. (1970) The Effects of Noise on Man. New York: Academic Press.
Latane, B. and A.J. Arrowood (1963) Emotional arousal and task performance. Journal of
Applied Psychology 47(5):324-327.
Leonard, S. and P.N. Borsky (1973) A causal model for relating noise exposure, psycho-
social variables and aircraft noise annoyance. Pages 691-705, Proceedings of the
International Congress on Noise as a Public Health Problem, May 1973, Dubrovnik,
Yugoslavia. Prepared by Office of Noise Abatement and Control, EPA-550/9-73-008.
Washington, D.C.: U.S. Environmental Protection Agency; PB-241 060/3GA.
Springfield, Va.: National Technical Information Service.
Lukas, J.S. and K.D. Kryter (1970) Awakening effects of simulated sonic booms and
subsonic aircraft noise. Pages 283-293, Physiological Effects of Noise, edited by B.L.
Welch and A.S. Welch. New York: Plenum Press.
Mathews, K.E. and L.K. Canon (1975) Environmental noise level as a determinant of
helping behavior. Journal of Personality and Social Psychology 32(4): 571-577.
Memphis State University (1971) Effect of Noise on Wildlife and Other Animals. EPA-
NTID 300.5. Washington, D.C.: U.S. Environmental Protection Agency; PB-206 720.
Springfield, Va.: National Technical Information Service.
Miller, J.D. (1974) Effects of noise on people. Journal of the Acoustical Society of America
56(3):729-764.
Mills, C.H.G. and D.W. Robinson (1961) The subjective rating of motor vehicle noise. The
Engineer 211:1070-1074.
Mills, J.H. (1975) Noise and children: a review of literature. Journal of the Acoustical
Society of America 58(4):767-779.
National Institute of Child Health and Human Development (1972) Perinatal Physiology. A
Report Prepared by the Perinatal Biology and Infant Mortality Program [National
Institutes of Health] for Presentation to the National Advisory Child Health and Human
Development Council. N.P.: National Institute of Child Health and Human Develop-
ment. (Photocopy)
National Institute of Child Health and Human Development (1975) Infant Morbidity &
Mortality: A Program for Prevention. A Report Prepared by the Perinatal Biology and
Infant Mortality Program [National Institutes of Health] for Presentation to the National
Advisory Child Health and Human Development Council. N.P.: National Institute of
Child Health and Human Development. (Photocopy)
National Institute for Occupational Safety and Health (1972) Criteria for a Recommended
Standard. Occupational Exposure to Noise. PB-213 463. Springfield, Va.: National
Technical Information Service.
National Research Council (1971) Non-Auditory Effects of Noise. National Academy of
Sciences-National Research Council. Committee on Hearing, Bioacoustics and Biome-
chanics. Report of Working Group 63. Washington, D.C.: National Academy of
Sciences.
-------
Benefits of Noise Abatement 125
National Research Council (1975) Decision Making for Regulating Chemicals in the
Environment. A Report Prepared by the Committee on Principles of Decision Making
on Regulating Chemicals in the Environment, Commission on Natural Resources.
Washington, D.C.: National Academy of Sciences.
National Swedish Institute for Building Research (1968) Traffic Noise in Residential Areas.
Study by the National Swedish Institute for Building Research and the National Swedish
Institute for Public Health, Report 36E. Stockholm: Statens Institut for Byggnadsforskn-
ing.
O'Malley, J.J. and A. Poplawsky (1971) Noise-induced arousal and breadth of attention.
Perceptual & Motor Skills 33:(3 Pt.l)887-890.
Reim, B., D.C. Glass, and I.E. Singer (1971) Behavioral consequences of exposure to
uncontrollable and unpredictable noise. Journal of Applied Social Psychology 1(1):44-
56.
Schachter, S., B. Willerman, L. Festinger, and R. Hyman (1961) Emotional disruption and
industrial productivity. Journal of Applied Psychology 45(4):201-213.
Schein, J.D. and M.T. Delk, Jr. (1974) The Deaf Population of the United States. Silver
Spring, Md.: National Association of the Deaf.
Selye, H. (1956) The Stress of Life. New York: McGraw-Hill.
Singer, J.E. (1976) Social and psychological impact of transportation noise. Pages 499-504,
Inter-Noise 76. Proceedings of the International Conference on Noise Control Engineer-
ing, held in Washington, D.C., April 5-7, 1976 by the International Institute of Noise
Control Engineering and the Institute of Noise Control Engineering concurrent with the
Acoustical Society of America. State College, Pa.: Institute of Noise Control Engineer-
ing.
Terleckyj, N.E. (1975) Improvements in the Quality of Life: Estimates of Possibilities in the
United States, 1974-1983. Washington, D.C: National Planning Association.
TRACOR, Inc. (1971) Community Reaction to Airport Noise, Vol. I. NASA CR-1761.
Washington, D.C.: National Aeronautics and Space Administration.
U.S. Environmental Protection Agency (1971) Community Noise. Office of Noise Abate-
ment and Control EPA-NTID 300.3. Washington, D.C.: U.S. Environmental Protection
Agency; PB-207 124. Springfield, Va.: National Technical Information Service.
U.S. Environmental Protection Agency (1974) Information on levels of environmental noise
requisite to protect public health and welfare with an adequate margin of safety. Office of
Noise Abatement and Control, EPA-550/9-74-004. Washington, D.C: U.S. Environ-
mental Protection Agency; PB-239 429/4BE. Springfield, Va.: National Technical
Information Service.
Weinstein, N.D. (1974) Effect of noise on intellectual performance. Journal of Applied
Psychology 59(5):548-554.
Woodcock, L. (1975) A. "This noise hurts." B. "What?" New York Times 7:17:29:1.
Wyle Laboratories Research Staff (1970) Home Soundproofing Pilot Project for the Los
Angeles Department of Airports, Final Report. Report No. WCR 70-1. El Segundo,
Calif.: Wyle Laboratories.
-------
7
Monetary Measures
of the Benefits
of Abatement:
Property-Value Analysis
INTRODUCTION
This chapter outlines the method most widely used to provide a monetary
evaluation of the benefits of non-industrial noise abatement. The method
is indirect, using as its basis observed differences in property values in
quiet and noisy neighborhoods (correcting for the effects of other
pertinent variables). It is necessary to use an indirect method because, as
indicated in the preceding chapter, the direct measures of the benefits of
noise abatement that are now available are largely qualitative. Of those
that are quantifiable, only some can be expressed directly in monetary
terms. Since a comparison of the benefits and the costs of any abatement
program requires both variables to be measured in the same units,
qualitative measures of benefits will not suffice for a cost-benefit analysis.
Therefore, analytic economic methods are used to construct a proxy, or
surrogate, measure of the benefits, which is calculated in monetary terms
and so is directly comparable with the cost estimates. This surrogate
measure is based on analyses of residential property values.
It should be emphasized that the purpose of the property-value method
is not to measure financial losses to property owners. Indeed, economists
do not even consider such losses in property values in themselves to
constitute a net loss—the financial loss to the seller is, after all, exactly
matched by the financial gain to the buyer. Rather, the property value
126
-------
Monetary Measures: Property-Value Analysis 127
approach is intended to estimate what monetary value the residents of
noisy areas place on the physical, psychic, and social damage that they
suffer.
The logic of the approach is straightforward. People who are offered a
house in which they may suffer hearing loss or speech interference or
other forms of damage from noise will be prepared to pay less for that
house than they will be prepared to pay for a quiet residence. Moreover,
the greater the noise damage they expect, the larger will be the resulting
discount in the amount they will offer. In fact, in an ideal market
arrangement, the size of the differential between the price of noisy and
quiet houses must be an exact measure of the valuation of noise damage
by buyers and sellers. For if the differential were less than this—that is, if
it were insufficient to compensate buyers for the noise damage—there
would be a flow of demand from noisy to quiet houses, forcing the
differential to increase, and the reverse would be true if the differential in
price more than made up for the damage caused by noisy houses.
In principle, then, the property-value method offers the prospect of a
monetary measure of the value of noise abatement to those who would
benefit from it. In practice, however, as will be emphasized later, the
market mechanism is far from perfect, and this and other considerations
require us to take the resulting estimates of the physical, psychic, and
social benefits of noise abatement with a considerable grain of salt. But,
at least for the moment, no better method has been designed to give an
overall quantitative measure of the monetary benefits of noise abate-
ment.1
The monetary estimates presented in this chapter are, perhaps, of some
value in themselves as a representative product of the current state-of-
the-art. But their primary purpose is to illustrate the issues raised by this
widely used method of benefit assessment rather than to provide yet
another set of figures as fuel to the controversy over the relative merits of
the available estimates.
1There has been considerable debate in the economics literature concerning the usefulness
of the property-value model for inferring the benefits of pollution abatement, particularly of
air pollution. (For a discussion of these studies, see NRC 1974 and Rubinfeld, D. [In press]
Market approaches to the benefits of air pollution abatement. Chapter 6, Approaches to
Controlling Air Pollution, edited by Ann F. Friedlaender. Cambridge, Mass.: M.I.T. Press.).
There are reasons to believe that this approach works better with regard to estimating the
effects of noise than with regard to other air pollutants because noise can be perceived by
the potential home buyer while many air pollutants cannot and because the level of noise at
a particular location can be estimated more accurately than the level of air pollution.
However, other criticisms of the approach remain, one of which is the difficulty in obtaining
accurate measures of real estate values. Many of the more technical criticisms are not
discussed in this report.
-------
128 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
The first section of this chapter discusses some of the analytical
approaches used to evaluate the benefits of noise abatement. The second
section describes a model of the noise problem that is applicable to a wide
variety of transportation modes and indicates the implications of this
model for the evaluation of benefits. The third section reviews some of the
evidence on the influence of airports and highways on property values.
The fourth section evaluates the benefits of reducing airplane, automo-
bile, and truck noise. These sections focus on the benefits to those who
are affected by noise and who are not themselves users of the
transportation system. The fifth section uses another approach to the
evaluation of benefits: a cost-effectiveness analysis to find efficient
combinations of interdependent noise abatement programs (QT given
levels of expenditures. This section then examines the values that decision
makers must place on noise abatement if those programs are to be
justified on benefit-cost grounds.
Finally, the last section of the chapter discusses some of the limitations
of the property value method of benefit measurement and seeks to
suggest the likely magnitude of some of the resulting errors.
ANALYTICAL PRINCIPLES
It is important to recognize that transportation noise is a by-product of a
service that has economic value. The amount of this undesirable by-
product is related to the levels and distribution of transportation activities
and to the quantities of resources that are allocated to the reduction of
the noise by-product.
An indicator of the benefits of reducing noise is the amount people are
willing to pay to be relieved from its effects. Unfortunately for both
analysts and policy makers, quiet, or the freedom from noise, cannot be
bought and sold by the decibel in the open market. If any site could be
exposed to transportation noise only after the noise maker had purchased
from the property owner an authorization to do so, it might be possible to
measure directly the value that people in our society place on a reduction
in their noise exposure. Such a market does not yet exist, although there
are some legal means by which third parties may be compensated, e.g., in
the purchase of noise easements by airport authorities. The absence of a
market that places a direct value on reduced noise leads economists to
use an indirect method to estimate this value. The indirect procedures
usually involve the search for some market in which noise exposures are
bought and sold implicitly as a tie-in with some other good. The most
common variant of this implicit market approach is the use of property
-------
Monetary Measures: Property-Value Analysis 129
values to find the change in market price associated with various noise
exposures.
For noise emanating from a well-defined source, the property-value
approach provides a reasonable approximation of the cost of noise. Since
noise decays at a smooth rate from its source, there are a variety of noise
intensities available from which an individual may choose a location for a
home or business.
Where the noise sources are so diffuse that they essentially become part
of the ambient noise level, however, the property-value approach is less
reliable. Because the ambient noise level is found throughout a given
area, an individual cannot choose between a location where it is present
and another where it is not. A market value cannot be placed on that
which is inescapable, for market values are always revealed by choices
among alternatives and the associated effects on prices. In the case of
ambient noise, the choices individuals make to modify the interior noise
environment—for example, by soundproofing their homes—can be used
to obtain a market valuation of quiet.
Ideally, an evaluation of the benefits of transportation noise abatement
would begin with the joint frequency distribution of noise from all
transportation sources at each location within the area to be analyzed.
Information about the interrelations among noise from all sources is
needed because the benefits of some marginal reduction in noise from
one source are conditional on the level of noise from other sources. The
benefits of an equal reduction in truck noise at two different locations will
differ, perhaps substantially, if one of the two locations is exposed to jet
airplane noise and the other is not. Similarly, the benefits of a program to
reduce airplane noise will depend on the character of any other noise
abatement programs that may be adopted simultaneously. The benefits of
various noise abatement programs are not independent, and the joint
noise frequency distribution is needed to measure the marginal contribu-
tion of each program and to establish priorities among programs as well.
Suppose, for example, one is choosing among three noise abatement
programs, call them Programs 1, 2, and 3. Evaluated separately,
Programs 1 and 2 may yield benefits exceeding their costs, while Program
3 does not. However, it is possible that adoption of Program 1 lowers the
benefits of 2 while raising the benefits of 3 to such an extent that a
combination of Programs 1 and 3 is preferable to any other combination.
If the programs are evaluated separately, a non-optimal choice will be
made.
Unfortunately, the required joint noise distributions are not available,
so this report evaluates separately the benefits of airport noise abatement
-------
130 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
and highway noise abatement. However, using work that has been done
on the joint noise distribution for Spokane, the evaluation of noise
abatement programs from a cost-effectiveness point of view will be
illustrated.
THE PROPERTY-VALUE MODEL
A site close to an airport or a highway will usually experience fairly
intensive noise, but it will also benefit from proximity to transportation.
Generally, proximity to transportation will raise the value of property
nearby relative to property further removed. This effect is referred to as a
pecuniary externality: the increase in values arising if a highway is routed
through location A would have been realized just as well had it been
routed through some other equally efficient location, and the rise in
property values along the highway are therefore matched by decreases
elsewhere. In other words, the increase in values does not correspond to
any real net social gain, but is simply a transfer of rents from one location
to another.
The noise emitted from an airport or highway is a technological
externality: noise uses up a real resource—quiet. When noise is
"dumped" on property, the productivity of that property is affected in
absolute and relative terms. Property affected by noise of high intensity is
less productive for virtually any use than comparable quiet property, and
its productivity as a housing site may be reduced even more than its
productivity as a site for commercial activity.
Quiet residential sites will be in greater demand than noisy ones, and
the resulting differences in residential property values should approxi-
mate the value that individuals place on residential quiet. Noisy
commercial property should also sell at a discount compared to quiet
sites, assuming equal access to labor and other inputs. Commercial
property will similarly be discounted if it is noisy, if all else is the same,
because workers will be less productive or customers will be less attracted
to the business. Theoretically, the discount will never exceed the least
costly way of completely eliminating the noise.
LOCATIONAL PREMIUMS AND NOISE DISCOUNTS: AIRPORTS
If the discount on noisy property is to be taken as an estimate of the cost
of noise, then the calculation must be carried out in a manner that
disentangles the locational premium from the noise discount. Consider an
airport with a surrounding commercial district. If the airport were
-------
Monetary Measures: Property-Value Analysis 131
absolutely quiet and did not emit pollutants or cause congested street
traffic, then the value of property near the airport would exceed the value
of property some distance away. Figure 7.la illustrates the typical
behavior of the proximity value and its relation to distance from the
desired location (the noiseless airport). The height of the p curve is the
premium on property near the airport relative to property farther
removed. The premium slopes downward at a rate roughly equal to the
additional cost of transportation.
Adding airport noise, there will now be a discount on property
reflecting the disutility of noise. The discount will be highest at locations
close to the airport (where noise is greatest) and will diminish at a rate
that reflects the disutility of noise and the rate at which noise attenuates
to the ambient level. The noise discount is shown by the d curve in Figure
7.1b, and can be denned as the reduction in property value associated
with a unit increase in the noise level index, all other things being equal.
In the empirical studies reviewed below, this discount is expressed in
dollars of depreciation on property per unit increase in noise or as the
percentage depreciation of an average property per unit increase in noise.
Putting the locational premium together with the noise discount yields
a net p-d curve, as shown in Figure 7.1c. The noise discount alters the
locational premium and transforms the curve of property values into
something like a crater. The precise effect an airport will have on property
values depends upon a number of factors, such as the ambient noise level
and the sound reflecting barriers or other influences affecting noise
attenuation, as well as the degree of commercial activity at or near the
airport and the efficiency of the transportation system. These features
differ from airport to airport, so the value curve will not have the same
shape near every airport.
In a city of high population density with moderate to high transporta-
tion costs, such as London, one might expect a pattern such as that shown
in Figure 7.2a. A city of low density with low transportation costs, such
as Los Angeles, might have a pattern like that in Figure 7.2b. Because of
the very high cost of transportation and the great consequent value of
proximity in the "London model," the airport increases property values
on balance everywhere within the region where it has any effects. In the
"Los Angeles model," the airport decreases property values over a large
region nearest to it, but produces small increases in the values of
properties lying some distance away. Neither of these diagrams necessari-
ly indicates the actual state of affairs in the two cities. Empirical studies of
their airports do suggest that these diagrams depict matters correctly, but
none of the studies carried out so far has completely disentangled the
-------
132 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
$
DISTANCE
DISTANCE
DISTANCE
FIGURE 7.1 Property value effects of a hypothetical airport.
-------
Monetary Measures: Property-Value Analysis 133
locational premium from the noise discount. Evidence relating to Love
Field in Dallas yields patterns similar to the "Los Angeles model."2
From the point of view of evaluation of the benefits of noise reduction,
it does not matter whether the effect of the airport on property values is
positive or negative on balance. Since the locational premium is a
pecuniary externality, it does not affect the real goods and services
available to society as a whole. A net gain in property values only
represents a windfall gain to those who happen to own property when the
airport is announced (which, if it had been captured by the airport
authority, could have helped to finance the airport's construction). The
benefits of noise abatement are approximated by the "bite" taken out of
the curve representing the value of proximity to the airport. This explains
the paradox in the conflicting claims sometimes made by airport
authorities and homeowners, with the authorities claiming the airport has
increased property values, and the homeowners claiming that their
property value has been reduced by the airport's noise. In this model,
they can both be right. From an efficiency point of view, however, it is
only the cost of the noise that is of concern, and the noise should be
reduced whenever the marginal value of the reduction exceeds the
marginal cost, irrespective of the overall effect of the airport on property
values.
THE EFFECTS OF NOISE REDUCTION
Suppose an airport has an initial value curve such as that shown in Figure
7.2. Now, let noise be diminished with no change in the level of
operations or employment at the airport. The value curve will rise near
the airport and decrease at locations distant from the airport due to the
increase in the supply of quiet sites relative to the supply of noisy sites. As
a consequence, the premium on quiet sites is diminished. Overall, the
benefits of the noise reduction include the area of property-value increase
in the graph minus the area of decrease. This difference will always be
positive because a saving in transportation costs must accompany the
noise reduction. With their exposure to noise the same as before,
individuals will now be able to live closer to the airport and thereby
conserve on the cost of access to the employment, commercial activity,
and transportation of the airport facilities. The cost of noise is the
transportation cost that individuals must bear in order to escape it,
2De Vany, A. (In press) An Economic Model of Airport Noise Pollution in an Urban
Environment. Carbondale, 111.: South Illinois University Press.
-------
134 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
a. "London Model"
P-d
DISTANCE
DISTANCE
b. "Los Angeles Model"
FIGURE 7.2 Two patterns of property value effects of airports.
-------
Monetary Measures: Property-Value Analysis 135
including both the money cost, the value of the time lost in the process,
and any associated disutility.3
The cost of an increase in noise can be estimated by asking what
additional transportation cost will be borne if individuals so exposed are
relocated to new locations having noise exposure equivalent to their
original site. In an orderly market, property values will reflect this
transportation cost. Consequently, the cost of noise can be measured
either as the differential in property values or as the increased transporta-
tion costs individuals are willing to bear to reduce their noise exposure.
(Homeowners can do things other than move to alter their noise
exposure; they can soundproof, air-condition and close windows, and
otherwise alter their living styles.) The discount on a house with a noise
exposure forecast (NEF) of 40, for example, relative to one with an NEF
of 30, can never exceed the least costly means of achieving a living
environment in the 40 contour that is equivalent in terms of utility to an
environment in the 30 contour.
HIGHWAYS
This model applies to highways as well as airports (see Gamble et al.
1974). Access to highways has a positive value and highway traffic also
emits noise and other pollutants. It is expected that a highway will
transform the curve of property values in a manner similar to that of an
airport. In this case, however, noise is radiated over a smaller distance
and attenuates at a more rapid rate because noise from a low altitude
source usually encounters a profusion of reflection agents and barriers
such as vegetation, buildings, etc.; therefore, the noise discount is likely
to be confined to a fairly narrow band around the highway. The
locational premium will slope away at a rate representing the value of
access to the highway.
3This can also be expressed in terms of a moving rule for the household (Walters 1975:37).
Let N be the differential noise evaluation for two properties that are identical except for
noise exposure. Then
NS+D+R (move)
NS+D+R (stay put)
where S is the difference in consumer surplus associated with the two residences, D is the
capital loss due to differences in prices, and R is search and removal (transportation) costs.
The values of N, S, R, and D are the present values of expected future outlays or valuations
at some particular rate of time preference.
-------
136 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
EMPIRICAL EVIDENCE ON THE NOISE DISCOUNT
The various noise indices that have been designed to measure airplane or
road traffic noise do not have as their objective an explanation of the
influence of noise on property values or locational choices of individuals.
There is evidence, however, that noise, as measured by the standard
indices, does adversely affect property values, all other things being
equal. The various noise indices do contribute to an understanding of
urban property values and are, therefore, measuring a phenomenon to
which individuals react in their economic behavior.
AIRPLANE NOISE
Two very able reviews of the evidence from studies through 1974 are
available (Walters 1975, Nelson 19754). While there are several unsettled
technical issues, the available evidence suggests that airplane noise
reduces property values, and the amount and quality of the evidence is
reasonably impressive.
Table 7.1 summarizes the studies of airplane noise and property values
through 1974. The studies cited place the percent reduction in average
property value per unit NEF in the range of 0.4 to 2.0 percent.5 The cities
included in the samples of those studies are diverse in climate, population
density, and mean housing values, and the functional forms employed by
the authors also differ, so it is not surprising to find some differences in
the estimates of property values. (Indeed, the discussion of the preceding
section suggests that because of differences in the characteristics of
different cities this can be expected to be the rule rather than exception.)
One point to be noted in evaluating these studies is the range of NEF
values considered. The Emerson study (1969) deals with a range from
about 30 to 55 NEF and is the only study that includes values above 45. It
therefore measures response to noise well above the range considered by
the other studies; however, it does not include any data in the 20- to 30-
NEF range, which is the range within which airplane noise is generally
considered to become noticeable. (The ambient noise level is about 20
4A slightly revised and expanded analysis of the data in the Nelson Study appears in An
Analysis of Jet Aircraft Noise and Residential Property Values. Institute for Research on
Human Resources. University Park, Pa.: Pennsylvania State University. (Unpublished)
The 2.0 percent noise discount is from Paik's study (1972), which employs 1960 data. All
other studies in Tables 7.1 and 7.2 use data from the period 1967-1971. A comparison of
1960 results with the later period suggests a decline in the noise discount over time. This may
be due to an adjustment toward a new long-run equilibrium or it may reflect soundproofing
or air-conditioning of homes and the introduction of new, quieter, wide-body jets. For a
study of the noise discount over time, see Crowley (1973).
-------
Monetary Measures: Property-Value Analysis 137
NEF.) In the other studies, NEF areas above 45 are not considered so
that the most heavily affected areas are not included.
More recent work has been done by Nelson (1975, 1976), De Vany,6
and Mieszkowski and Saper.7 Their estimates are compared in Table 7.2.
If the reductions in value per unit of NEF are adjusted to a mean housing
value of $35,000, which is the sum reported in the Mieszkowski and Saper
study,8 then the discount becomes $350 and $204 per NEF for the Nelson
and De Vany studies, putting them in fairly close agreement with the
Mieszkowski and Saper discount of $210. The De Vany study9 indicates
that the overall effect of the airport on land values is positive, even though
there is substantial noise damage.
HIGHWAY NOISE
The model suggests that empirical studies of the effects of highways on
property values should find a narrow belt of net noise damage around
each highway, surrounded by a region in which property values are
increased by the accessibility afforded by the highway. Although large-
scale, multivariate statistical studies for highway noise are limited in
number, the available evidence from three studies is of reasonable quality
and consistency. Unfortunately, each study employed a different index of
traffic noise, which somewhat complicates comparison of the damage
estimates.
Table 7.3 summarizes the studies that relate property values to highway
noise. While the percentage of damage per unit of noise in dB(A) seems
rather high for the Bogota (New Jersey) sample, the other areas exhibit a
fairly narrow range of damages from 0.20 to 0.60 percent. A more exact
comparison of marginal damages can be made by using the traffic noise
index (TNI), which is the noise level exceeded 10 percent of the time
minus the noise level exceeded 90 percent of the time (see Chapter 3). A
1-unit change in the TNI is equal to about a 1.1 unit change in the noise
pollution level (NPL) index and about a 1.25-unit change in Leq (Illinois
Institute for Environmental Quality 1976). Converting to TNI units,
marginal damages are $147, $168, and $102, respectively.
Unfortunately, none of these studies attempted to disentangle com-
pletely the locational premium from the noise discount. Using informa-
tion on the value of access compiled by the Washington, D.C. Council of
6See note 2 above.
'Mieszkowski, P. and A.M. Saper. An estimate of the effects of airport noise in property
values. Journal of Urban Economics. (Forthcoming)
8See note 7 above.
9See note 2 above.
-------
TABLE 7.1 Summary of Jet Airplane Noise Pollution Studies1
Study
Emerson (1969)
Paik (1972)
Dygert (1973)
Price (1974)
Functional Form
for Noise
Log
Log
Semi-Log
Semi-Log
Linear
Noise
Coefficient
-0.003
-0.018 to
-0.025
-0.005 to
-0.007
-1.267
R2
0.79
0.78
0.60
0.50
Marginal
Damage Estimate2
-$123 /NEF
-$560/NEF
-$140/NEF
-$100/NEF
Range of
Noise Values
100-125 CNR
(30-55 NEF)
20-40 NEF
25-45 NEF
25^5 NEF
Percent Reduction
in Average Property
Value per Unit NEF
0.4
2.0
0.5
0.4
•ived from Nelson (1975:8-10). See also Walters (1975:103).
2In 1970 dollars and relative to a $28,000 property.
-------
TABLE 7.2 Summary of Recent Jet Airplane Noise Pollution Studies1
Study
Nelson (1975 and
1976)
Mieszkowski and
Saper(1975)
De Vany (1976)
Distance to Airport
Within 1 mile
1 to 2 miles
2 to 3 miles
Functional Form
for Noise
Semi-Log
Linear
Semi-Log
Log
Noise
Coefficient
-0.010
N.A.
-0.065
-0.050
-0.123
R2
0.86
0.90
0.71
0.88
0.79
Marginal
Damage Estimate
-S280/NEF
-$210/NEF
-$ 33/NEF
-$ 52/NEF
-S164/NEF
Range of
Noise Values
20^5 NEF
25-35 NEF
20-55 NEF
20-50 NEF
20-45 NEF
Percent Reduction
in Average Property
Value per Unit NEF
1.0
0.60
0.22
0.22
0.58
Data derived from Nelson (1975:8-10). See also Walters (1975:103).
2In 1970-71 dollars. The average property values for the three studies are about $28,000, $35,000, and $22,000 (all areas), respectively.
-------
140 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 7.3 Summary of Highway Noise Pollution Studies
Study
Nelson (1975)
Vaughan and
Huckins (1975)
Gamble et al.
(1974)
Area
Suburban Wash., D.C.
Chicago
North Springfield, Va.
Bogota, N.J.
Rosedale, Md.
Towson, Md.
All areas
Noise
Measure
TNI
Leq
NPL
NPL
NPL
NPL
NPL
Marginal Damage
Estimate Per
Unit Noise
-$147
-$135
-$ 69
-$646
-$ 60
-$141
-$ 82
Percent Reduction
in Average Property
Value per Unit Noise
0.40
0.60
0.20
2.22
0.24
0.42
0.26
1 In 1970-71 dollars, average property values are about $32,000, $22,500, and $31,000 (all areas),
respectively.
Governments, Gamble et al. (1974) were able to put together a partial
picture of the effect of a highway for the North Springfield area. Using a
constant noise damage value of $69 per dB(A) per residence and a
constant accessibility value of $2,955 per residence, one obtains a value
curve like that shown in Figure 7.3. At a distance of roughly 1000 feet
from the highway, noise from the highway falls to the ambient level,
about 55 dB(A), and there is no further loss in value. The assumption that
the accessibility value is constant over distances up to 1000 feet is
probably realistic, although for much greater distances one would expect
the accessibility value to fall with increasing cost of access to the
highway.
BENEFITS OF NOISE ABATEMENT
AIRPLANES
When the property-value model is used to evaluate the benefits of noise
abatement, a series of assumptions and extrapolations must be made.10
Jet airplanes typically serve many airports, so the benefits of making the
airplanes quieter are distributed among many places, yet rigorous
property value studies have been conducted for about 10 airports at most.
It is necessary, therefore, to extrapolate a damage value per NEF from
one study or an average of several studies. Furthermore, the available
data on noise exposure are far from complete. What is available is an
10For some discussion of the the technical aspects of these issues, see Freeman (1974), Oron
et al. (1974), and Polinsky and Shavell (1976).
-------
Monetary Measures: Property-Value Analysis
141
P-d
DISTANCE
SOURCE: Modified from Gamble et al. (1974)
FIGURE 7.3 Property value effects of a highway.
estimate of the number of people who reside within the NEF-40 and
NEF-30 contours (but see U.S. EPA 1974c). As a consequence, the
analyst is forced to assume that those persons who live outside the 30
contour obtain essentially no benefits from noise abatement, in spite of
evidence that noise causes some annoyance within the 25 contour.
This report has previously argued that it is erroneous to take a project-
by-project approach when, in fact, the benefits of airplane noise
-------
142 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 7.4 Estimated Number of People Residing in NEF 30+ and NEF
40+ and Associated Land Area in 1972
Noise Level Population Acres1
NEF 30+ 6,200,000 965,000
NEF 40+ 630,000 114,000
Land area includes residential, industrial, commercial, and farmlands, as well as high-
ways and surface transportation facilities. Not all this land area is incompatible with the
imposed noise levels.
SOURCE: Safeer (1975)
reduction depend on other projects that affect environmental noise. Since
it is always preferable to give priority to the least costly way of achieving
a given reduction in noise, it would be best to evaluate the benefits of
airplane noise abatement programs only after we know how many people
would be saved from exposure to NEF-30 by the adoption of, for
example, all more cost-effective highway noise control programs.
Similarly, it would be best to evaluate highway noise abatement programs
only in terms of the benefits to those people who would remain in the
NEF 30 contour after all more cost-effective aircraft noise abatement
programs have been adopted. However, since the data necessary for such
analyses are not available, aircraft and highway noise are considered
separately in this discussion.
Table 7.4 provides estimates of the size of the population or land area
affected by noise. The estimates of noise damage per residence from one
or several of the studies reviewed above can be translated into benefits of
noise reduction per residence, per person, or per acre of land, but only if
there are specific alternative noise abatement programs to evaluate. For
any given level of abatement, the program or combination of programs
that is cost-effective—i.e., that achieves the target level of abatement at
least cost to society—is to be selected. The benefits of these programs,
starting with the most cost-effective program and adding incremental
programs so long as marginal benefits exceed marginal costs, are then
evaluated.
Safeer (1975) has provided a relative ordering of the major options for
airplane noise abatement in terms of the numbers of people and land area
removed from the NEF 30+ and NEF 40+ areas. Five major
alternatives were analyzed:
1. retrofitting of all JT3D- or JT8D-powered aircraft with new nacelles
containing sound absorption material (SAM);
-------
Monetary Measures: Property-Value Analysis
143
2. retrofitting of all JT8D-powered aircraft with refanned engines and
new nacelles (REFAN);
3. modifying approach procedures (two-segment);11
4. modifying takeoff procedures (thrust cutback); and
5. acquiring land within the NEF-40 contour.
The results of Safeer's analysis are shown in Figures 7.4 and 7.5. Each
option number is listed to the right of the figure under a column heading
indicating the date at which the program in question is assumed to be in
full operation. It should be noted that some of the procedures examined,
e.g., two-segment landings, are not fully applicable to all aircraft. Safeer
then gives the population and land area removed from the NEF-30 + and
NEF-40 + contours by the various alternatives together with the benefits
associated with the programs. He employs a benefit estimate using Paik's
study (1972) based on 1960 census data.
5r
E 4
i
C/3
8
DC
(D
O
cc
0.
ONDIT
ASE
ON
ASE t 2 SEG
ASE +• (
JB
SE +2 SEG *C/B
M 30
M30
M 30
M30
M80
M80
M80
M80
M 30/
2 SEG
C/B
2 SEG * c/B
2 SEG
C/B
2 SEG + C/B
30
M 30/80 + 2 SEG
M 30/
30 + C/B
M 30/80 * t SEG • C/B
M 30
RFN 80
M 30-RFN 80 + 2 SEG
M 30
RFN 80 *C/B
AM 30-RFN 80 t 2 SfcG ^ C/B
1978
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
1981
17
18
19
20
06-^_L
20 40
100
IMPACTED LAND AREA REMOVED FROM NEF 30+ i
SOURCE: Safeer (1975)
FIGURE 7.4 Cost vs. effectiveness, NEF 30+.
"The FAA has very recently decided not to prescribe a two-segment approach, primarily
for reasons of safety. In its stead, the FAA will require noise abatement by means of landing
flap setting procedures (Federal Aviation Administration and the U.S. Department of
Transportation 1976).
-------
144
BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
Each dot in the figure shows the cost and noise reduction correspond-
ing to the identified program or combination of programs. The aim is to
select the option that is least costly, given the land area removed from the
NEF-30+ or NEF-40+ contour. These cost-efficient options are given
by the envelope curve that goes through the lowest dots in the diagram.
For example, the curve in Figure 7.4 indicates that options 2,3,4,16, and
20 are the most cost-effective means by which to remove land incremen-
tally from the NEF-30+ contour.
Nelson's study (1975) contains a more explicit calculation of benefits
based on the empirical studies of airplane noise and property values for
1967-1971 data. Four major alternative abatement strategies were
examined.
1. No Change. Even with no new control programs, major reductions
in noise levels will occur as the result of introduction of new, quieter jets
(B747, DC-10, L-1011, and others) under the standards of the FAA's
Federal Aviation Regulation Part 36 (FAR 36) and the phasing out of the
airplanes now in use.
2. Two-Segment Approach. A 6°/3° two-segment landing approach
_ 5
o
•a
C/J
O
O
1
O
o
oc
D-
CONDITION
BASE
BASE + 2SEG
BASE + C/B
BASE + 2 SEG * C/B
SAM 30
SAM 30 + 2 SEG
SAM 30 + C/B
SAM 30 + 2 SEG + C/B
SAM 80
SAM 80 + 2 SEG
SAM 80 + C/B
SAM 80 + 2 SEG + C/B
SAM 30/80
SAM 30/80 + 2 SEG
SAM 30/80 + C/B
SAM 30/80 + t SEG ' C/B
SAM 30- RFN 80
SAM30-RFNBO + 2SEG
SAM 30-RFN80 * C/B
SAM 30- RFN 80 + 2 SEG + C/B
1978
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
1981
17
IB
19
20
0
20
40
60
80
100
IMPACTED AREA REMOVED FROM NEF 40+ i
SOURCE: Safeer (1975)
FIGURE 7.5 Cost vs. effectiveness, NEF 40+.
-------
Monetary Measures: Property-Value Analysis 145
for all airplanes will reduce noise levels, especially outside the NEF-40
contour. It is assumed that two-segment instrument landing systems can
be installed and tested and approach procedures instituted during 1976
and 1977.
3. SAM 8D/3D. This program requires all old JT8D- and JT3D-
powered airplanes to be fitted with acoustically treated nacelles begin-
ning in 1975. It is assumed that all civilian airplanes can be fitted with
quiet nacelles by the end of 1978.
4. REFAN 8D/SAM 3D. This requires all old JT8D-powered air-
planes to be fitted with refanned engines beginning in 1978 and all JT3D
engines to be fitted with acoustically treated nacelles beginning in 1976. It
is assumed that these modifications can be completed by the end of 1981
and 1978, respectively.
Nelson evaluates alternatives 2, 3, and 4 in terms of the resulting
incremental noise reduction as compared with the no change option. This
comparison correctly anticipates that, within the period considered by
the study (1975-1997), some of the noisier airplanes will be retired from
the fleet and that new airplanes coming into the fleet will be somewhat
less noisy as a result of FAR 36. Unfortunately, it is extremely difficult to
anticipate the exact rate at which older airplanes will be retired. Nelson's
analysis, which uses a Department of Transportation forecast, is
probably overly optimistic with regard to the retirement of older
airplanes. As a consequence, the benefits of abatement measures are
probably undervalued.
In order to calculate aggregate benefits, it is necessary to forecast: (1)
the reduction in NEF levels over time resulting from each noise
abatement alternative; (2) the percentage of persons within the NEF-
30+ contour who experience reduced noise levels; and (3) the dollar
value of benefits per person or per residence per NEF.
Table 7.5 shows the effect of each noise abatement alternative on noise
exposure through the year 1987. The information in this table indicates
the reduction in NEF values from each abatement program and the
percentage of persons remaining inside the NEF-30+ contour. Aggre-
gate benefits are calculated under several alternative assumptions about
what is the appropriate population: for example, whether or not those
people removed from the NEF-30+ contour continually share in the
benefits of a specific abatement program. In light of evidence that the
most stringent abatement program would only reduce the NEF-30
contour to about NEF 25, it seems appropriate simply to use the
population in 1975 inside the NEF-30 + contour or about 5.2 million
people (84 percent of 6.2 million).
-------
146 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 7.5 Impact of Airplane Noise Abatement on Noise Exposure,
1975-1987
Year3
Abatement Program1'2 1975 1978 1981 1987
No Change
ANEF
Efficiency
Two-Segment Landing
ANEF
Efficiency
SAM8D/3D
ANEF
Efficiency
REFAN8D/SAM3D
ANEF
Efficiency
-2.1
84.0%
0.0
84.0%
-0.4
84.0%
0.0
84.0%
-2.1
72.0%
-0.5
61.0%
-1.7
56.0%
-1.7
56.0%
-2.5
67.0%
-0.6
54.0%
-1.6
57.0%
-5.3
27.0%
-2.4
70.0%
-0.6
55.0%
,.-1.3
58.0%
-4.9
31.0%
'The A NEF data were supplied by John E. Wesler, Office of Noise Abatement, U.S.
Department of Transportation.
The efficiency factor is the percentage of the 1972 population remaining inside NEF
30+ due to each abatement program after accounting for introduction of new airplanes.
Population efficiency data are from unpublished supporting data from Bartel, Sutherland
and Simpson (1974), summarized on page 3-33 of their report. These data were adjusted
to account for the population efficiency of each alternative program without two-
segment landing.
3NEF measurements depend on the number of airplane flights. Therefore, A NEF factors
decrease and efficiency factors increase after 1978 or 1981 due to increases in air travel.
The efficiency factors for two-segment landing are affected by the fact that this option
would apply to all airplanes (if feasible).
SOURCE: Derived from Nelson (1976).
As his measure of benefits, Nelson uses an estimate of $140 per
residence per NEF (in 1970 dollars). To evaluate this figure, note that the
average property value in 1970 for metropolitan suburbs was about
$21,000. Thus, a noise discount of $140 per NEF is about 0.7 percent of
the average residential property value in 1970. A review of Tables 7.1 and
7.2 suggests that for studies using 1967-1971 data, the noise depreciation
value is in the range of 0.4 to 1.0 percent. Although it would be preferable
to obtain aggregate benefits by separate calculations for each airport,
reflecting the diversity that no doubt exists, $140 per NEF is at the mid-
point of the range of the damage figures available from empirical studies.
Table 7.6 shows the estimates of the discounted present value of the
four abatement programs. These estimates assume a real interest rate of 8
percent, the retirement of all old jet airplanes by 1997, and a 2-3 percent
growth in benefits to reflect income growth and associated increases in
-------
Monetary Measures: Property-Value Analysis 147
TABLE 7.6 Total Discounted Benefits of Jet Airplane
Noise Abatement, 1975-1997 (millions of 1974 dollars)
At an 8% Interest Rate
Abatement Program Until the Year 1997
No Change $ 998.3
Two-Segment Landing 214.2
SAM8D/3D 426.5
REFAN8D/SAM3D 1,109.2
SOURCE: Nelson (1976).
willingness to pay for noise abatement. Estimated benefits, in 1974
dollars, range from $214 million for two-segment landings to $1,109
million for the REFAN program.
Nelson's noise discounts can be compared to the price paid for flyover
easements for two airports. The average easement cost in Columbus
(Ohio) was $2414 for 30 easements, and the average cost in Denver
(Colorado) was $1000 for 32 easements (National Bureau of Standards
1971). If the residences on which the easements were purchased were in
the NEF-45 contour, this would suggest a discount per NEF of about
$112 (average easement of $1684 for a reduction from NEF 45 to NEF
30, which is not far out of line with the results of the studies cited above).
MOTOR VEHICLES
The estimation of benefits from reduction of noise from motor vehicles is
far less advanced than it is for aircraft noise abatement. None of the
studies examined has shown that central city urban property values are
affected significantly by motor vehicle noise per se. This is partly to be
expected because noise in highly urbanized areas comes from so many
sources. As a result, a feasible reduction in traffic noise will usually not
decrease noise exposure sufficiently to make a significant difference to the
home owner. Moreover, the relative uniformity of the noise level in an
urban area means that it is difficult to ascribe statistically any significant
portion of the difference in property values between one neighborhood
and another to differences in noise levels; usually, the location decision
that most affects an individual's noise exposure is in the choice between
an urban or suburban location. Consequently, only a comparison of
urban and suburban property values, adjusted for accessibility, would
seem to offer a way of discovering what value is placed on relief from high
ambient noise levels. As an alternative, one might investigate how
apartment rents vary with noise exposure or how much individuals are
-------
148 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
willing to pay to modify their interior noise level relative to the ambient
level. The studies by Nelson (1975) and Vaughan and Huckins (1975)
suggest that an ambient noise level of about 50 dB(A) is the approximate
threshold below which a change in the noise level has no impact on
property values. But it is not possible to determine from the available
evidence if an increase in the ambient level from, say, 50 dB(A) to 60
dB(A) has any effect on property values, assuming that the frequency and
intensity of intermittent sounds remain unchanged.
On the other hand, reasonably strong results are reported in studies of
the effect of highway noise on suburban and urban residential property
values, most notably in the area of freeways. In such areas, there is a well-
defined source of noise, and noise exposure can be varied substantially by
the choice of proximity to the freeway. In addition, EPA (1974a:6-15) has
provided estimates of populations exposed to various noise levels that
also take account of proximity to urban streets. These data provide a
basis for an extrapolation of the benefits of highway noise abatement,
although we feel these estimates are tentative at best and should be
revised as more data become available. (See also Vaughan and Huckins
[1975] and Illinois Institute for Environmental Quality [1976] for benefit-
cost comparisons of selected abatement programs.)
Nelson's report (1975), with his later corrections, provides a basis for
estimation of the benefits of abatement of noise from medium- and
heavy-duty trucks in excess of 10,000 pounds gross vehicle weight rating
(GVWR). Four abatement programs are considered (see U.S. EPA
1974a, 1974b).
1. Current operating rules for interstate motor carriers and new cars.
The Interstate Motor Carrier Noise Emission Standards (U.S. EPA
1974b) require that all motor vehicles above 10,000 pounds GVWR
operated by motor carriers engaged in interstate commerce meet the
following standards as of October 1975:
a. no more than 86 dB(A) at 50 feet in speed zones at or under 35
mph under all conditions, and
b. no more than 90 dB(A) at 50 feet in speed zones over 35 mph
under all conditions.
2. Model 1. An illustrative regulatory program under which new
trucks of over 10,000 pounds GVWR will be required not to exceed the
following noise levels after October of the year indicated:
a. 1976 83dB(A)
b. 1980 80dB(A)
c. 1982 75dB(A)
-------
Monetary Measures: Property-Value Analysis 149
TABLE 7.7 Reduction in Day-Night Sound Level (Ldn) in dBA Relative
to 1974 Values
Abatement Program
Freeways
Operating rules and new cars
Model 1
Model 2
Model 3
Urban Streets
Operating rules and new cars
Model 1
Model 2
Model 3
1976 1980
-2.4 -2.4
-1.2
-2.0
-1.2
-0.7 -1.2
-0.3
-0.6
-0.3
1982
-2.4
-2.6
-3.8
-2.6
-1.4
-0.7
-1.1
-0.7
1990
-2.4
-6.0
-6.2
-6.0
-2.0
-2.9
-3.0
-2.9
1992
-2.4
-6.2
-6.2
-6.2
-2.0
-3.3
-3.5
-3.3
SOURCE: U.S. EPA (1974a:6-20).
3. Model 2. A program whose restrictions are the same as in Model 1
but whose effective dates are different:
a. 1976 83dB(A)
b. 1977 80dB(A)
c. 1980 75dB(A)
4. Model 3. A program establishing separate standards for gas
engine and diesel engine powered trucks with the following effective
dates:
a. 1976
b. 1977
c. 1980
d. 1982
Gas
80dB(A)
80dB(A)
75 dB(A)
75 dB(A)
Diesel
83 dB(A)
83 dB(A)
80dB(A)
75 dB(A)
Table 7.7 shows the estimated reduction in the day-night sound level
(Ldn) associated with each program. The reductions indicated are
incremental so that, for example, the current operating rules together
with Model 1 regulations would yield a total reduction of 3.6 dB(A) along
freeways in 1980. Table 7.8 presents estimates of the population expected
to be exposed to noise levels in excess of 55 Ljn under the four programs
for the period 1974-1992.
For his estimate of benefits, Nelson used a (corrected) discount of 0.4
percent per dB(A) per residence or about $147 per dB(A) per residence.
Since this value is based on property values in suburban Washington,
D.C., it may somewhat overestimate the costs of noise in the United
States as a whole. The total benefit calculations assume that: (1) the 1974
-------
150 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 7.8 Populations Exposed to Day-Night Sound Level (Ldn) Greater
than 55 Under Alternative Programs (millions of people)
Abatement Program 1974 1976 1980 1982 1990 1992
Operating Rules and New Cars
Freeways
Urban Streets
Model 1
Freeways
Urban Streets
Model 2
Freeways
Urban Streets
Model 3
Freeways
Urban Stieets
2.7
34.6
2.7
34.6
2.7
34.6
2.7
34.6
2.1
31.5
2.1
31.5
2.1
31.5
2.1
31.5
2.1
29.4
1.8
28.0
1.7
27.0
1.8
28.0
2.1
28.4
1.6
25.6
1.4
23.2
1.6
25.6
2.1
26.0
1.1
15.9
1.0
14.9
1.1
15.9
2.1
26.0
1.0
14.9
1.0
13.8
1.0
14.9
SOURCE: U.S. EPA (1974a:6-21).
TABLE 7.9 Total Discounted Benefits of Heavy-
Medium Duty Truck Noise Abatement, 1976-2010
(billions of 1970 dollars)
At a 10% Interest Rate
Abatement Program Until the Year 2010
Operating Rules-New Cars $2.53
Model 1 2.53
Model 2 2.99
Model 3 2.53
SOURCE: Nelson (1975:10-1 7 as corrected).
populations continually receive benefits from reductions in noise through
the year 2010; (2) benefits increase at a rate of 5 percent per annum,
which is the predicted growth rate for new truck sales; and (3) the
appropriate real interest rate is 10 percent. The resulting estimates of the
total discounted benefits from each of the four programs are shown hi
Table 7.9 in 1970 dollars. For example, Model 1 would yield discounted
benefits of about $2.5 billion hi 1970 dollars.
PARAMETRIC EVALUATION OF INTERDEPENDENT
NOISE ABATEMENT PROGRAMS
As discussed earlier, it is incorrect to evaluate noise abatement programs
independently when they are in fact interdependent; that is, when the
-------
Monetary Measures: Property-Value Analysis 151
marginal benefits of each program depend upon the magnitudes of the
standards selected for other programs. Examples of these interdependen-
cies are abundant. For instance, the benefits of truck noise abatement will
differ according to the programs adopted for automobiles and airplanes if
portions of the population are exposed simultaneously to each of these
noise sources.
Two problems arise when interdependent programs are evaluated
separately. First, the marginal benefits of an individual program may be
over- or under-valued if they are estimated alone. For example, the
marginal benefits of airplane noise abatement may be increased relative
to those of other programs if a strong program of automobile traffic noise
abatement is adopted simply because airplane noise is no longer masked
effectively by urban traffic. At the same time, however, the marginal
benefits of low-speed truck noise abatement would probably diminish.
Consequently, the optimal combination of noise abatement strategies
cannot be determined unless their interdependence is recognized.
While these interdependencies are often recognized in evaluations of
programs designed to cope with noise emanating from one transportation
mode (see, for example, the Safeer study [1975] discussed earlier), this
important issue is generally ignored when several modes are under
consideration. Thus, very little work has been done to integrate all modes
of transportation in the calculation of cost-effective or cost-benefit
programs for the abatement of transportation noise. As a consequence,
the total benefits of noise abatement have almost certainly been
miscalculated, although we do not know, in general, whether they have
been overestimated or underestimated. More important, the combination
of programs recommended by several independent analyses is almost
certain to be inefficient. In principle, each program should be carried to
the point at which its marginal benefits in noise reduction received for
each dollar of expenditure is equal to that of all other programs for all
transportation modes. This is necessary to ensure that the maximum
noise reduction is secured for any given level of expenditure.
AN ILLUSTRATIVE STUDY
A better approach is illustrated in a pioneering study of noise in Spokane
by Wyle Labs.12 The Wyle study attempted to characterize the joint
12On March 17, 1977, Wyle and the Motor Vehicle Manufacturers Association circulated a
letter stating: "both Wyle and MVMA believe that the Spokane study is not sufficiently
accurate or definite to be used as a data reference . . . [but they do] believe that this unique
approach in studying community noise and counter-measure cost-effectiveness is appropri-
ate and its development as a policymaking tool should be continued."
-------
752 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
distribution of noise from all transportation sources and then defined
cost-effective combinations of noise abatement programs considering all
transportation modes for various given levels of expenditures. Using the
results of this study, it can be seen what implicit (negative) money values
one would have to attribute to noise in order to justify the funding of the
cost-effective noise programs, that is, how to determine the minimum
benefit levels at which such abatement programs become worth the cost
of carrying them out. Whether these break-even benefit values exceed the
corresponding figures derived from the property-value approach is also
considered.
The community of Spokane is described in the Wyle study in terms of
cells of population, with the people in each cell all living within an
(approximately) homogeneous noise environment. For one particular
time period, a noise level figure, defined by Leq, is calculated at a central
point of each cell, taking into account noise from all sources and
propagation losses. The effectiveness of a given noise reduction is then
defined as the reduction in the percentage of people in the cell reacting
adversely. This percentage of people reacting adversely is called a noise
impact index (Nil).
The Wyle study selected three expenditure levels—$5, $10, and $30
million—and made a set of assumptions about the measures that can be
used to abate noise from each source and about the cost of these
measures. It then determined the particular combination of such
measures that would minimize NIL As would be expected, the relative
expenditures for different measures and for different noise sources change
as the total amount to be spent is varied. For example, using the Wyle
data as illustrative, if $5 million is to be spent on medium-cost measures,
the optimal allocation assigns 72 percent of the total to reducing
automobile noise and nothing to barriers, home insulation, and reloca-
tion; with $10 million, the optimal allocation assigns 44 percent to
reducing automobile noise and nothing to barriers, home insulation, and
relocation; and with $30 million, the optimal allocation assigns 23
percent to reducing automobile noise and 57 percent to barriers, home
insulation, and relocation.
These comparisons are based on strong assumptions. Unfortunately,
the procedures, analysis, and data in the Wyle report do not permit an
actual calculation for Spokane but only an example of the method. The
point of this exercise is to demonstrate the general methods that can be
used to permit a benefit-cost analysis to take account of the interdepen-
dencies among transportation noise abatement programs.
-------
Monetary Measures: Property-Value Analysis 153
QUALIFICATIONS FOR THE PROPERTY-VALUE APPROACH
The primary objective in the adoption of property-value analysis as a
means of estimating the benefits of noise abatement is to obtain figures, in
dollars, directly comparable to those of the costs of abatement. By
themselves, these benefit estimates are of interest as indicators of the
magnitude of noise problems, but, more importantly, they are of value as
input to other analytic techniques. (One of these, a cost-effectiveness
analysis, was illustrated earlier in this chapter.) Their most common use is
in cost-benefit analyses. These analyses are widely used in the formula-
tion and evaluation of policy proposals. Chapter 9 contains two
illustrative cost-benefit exercises, one for jet aircraft, the other for
medium- and heavy-duty trucks. They serve as specific examples of the
ways in which the property-value analysis and the consequent benefit
estimates arrived at in this chapter can be used to provide information for
and help in the evaluation of noise abatement proposals.
There are aspects of the property-value model that need further
discussion. The Committee recognizes that the use of property values as
the sole index of benefits is likely to lead to evaluations that are far from
perfect. Yet, at least for the moment, no satisfactory measure of benefits
calculated independently of market values is available to cost-benefit
analysts.
The property-value approach, though it may work well or badly, is an
attempt to put quantitative pecuniary magnitudes on the damages
produced by noise on physical health, psychic well-being, and social
behavior. A neighborhood in which noise damages hearing, causes lack
of sleep, and leads to social disruption will be an undesirable neighbor-
hood to live in and we would expect that to some degree this will be
reflected in rents and property values. It must be emphasized that we are
interested in the relation between noise and property values not because
of any financial loss to property owners (which is just a transfer of wealth
from one social group to another); we are interested in property values
only to the extent that they reflect health, psychic, social and any other
forms of real noise damage. However, there are differences among
Committee members about the magnitudes of the likely errors of this
approach in carrying out that task and even, in some cases, about the
likely direction of those errors.
SOME QUALIFICATIONS REQUIRED FOR THE BENEFIT ESTIMATES
There are several problems involved in inferring estimates of the benefits
of noise abatement from the estimates of noise-induced discounts in
-------
154 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
property values. These problems vary in seriousness and in the extent to
which they suggest inaccuracies in the estimates. This section lists some
of the major concerns but will not try to resolve them; for many of these
issues, the arguments are moot, and for others, the discussion is too
technical for a general-purpose report. The nature of these difficulties can
be gathered from the following illustrative list of issues.
1. Some of the land affected by transportation noise is used for non-
commercial and non-residential purposes such as schools, parks, or
hospitals. Other land is used for streets and sewers. Some part of the
damage of noise to users of these properties will already be reflected in
depressed values of nearby homes—a worsening of schools does reduce
the price of homes nearby. But not all such noise damage will be reflected
in this way, and so some estimate of the residual damage to schools,
hospitals, and other such properties should be incorporated into the
property-value estimates of the benefits of noise abatement.
2. There are tax incentives to home ownership that induce house
purchasers to spend more for housing than they otherwise would spend
with their incomes. To the extent that they therefore pay a higher price
for quiet than they otherwise would, the benefits estimates have to be
adjusted downward.
3. If the effects of noise or of an abatement program lead people to
move to obtain quieter dwellings, the cost of those moving activities must
be deducted from the property-value estimates of benefits.
4. Often, the property-value calculations are used to estimate benefits
expected at some future time. In these cases, a discount rate is used to
translate figures for different dates into comparable units. However, the
calculated values of the benefits can be affected very substantially by the
number chosen for the discount rate, and there is no general agreement
on the way this rate should be chosen.
5. The adoption of an abatement program confers benefits on those
who own the property at the time, just as the original imposition of noise
imposed costs on those who owned the property at that time. Similarly,
the effects of noise or abatement may have different consequences for
owners than for tenants. These considerations raise questions of
distributive equity, which are ignored in the property-value method of
benefit estimation even though they may be considered vital for policy
decisions.
6. The property-value analysis estimates differentials in quiet and
noisy residences. If noise is so pervasive as to affect all properties, the
analysis may not be applicable. If the quietest location is noisy in
-------
Monetary Measures: Property-Value Analysis 155
absolute terms, no quiet residences will be available for comparison and
it may therefore not be possible to estimate willingness to pay for quiet.
7. Prices may reflect noise damage inadequately if buyers have very
imperfect information about the magnitude and the effects of noise. If
buyers are, for example, unaware that noise can produce deafness (and
there are incentives for the sellers not to disseminate such information),
buyers' willingness to pay for quiet may be different than if they are fully
informed. There are undoubtedly some effects of noise about which little
scientific evidence is now available, but which may be documented in the
future. Current willingness to pay, and, consequently, current property
value differentials, clearly cannot reflect noise damage that no one really
knows about.
8. The use of property values to infer people's willingness to pay for
quiet rests upon the assumption that there are no external constraints on
people's choices. If there is discrimination on the part of mortgage
lenders or realtors against particular racial, ethnic, or age groups, or if
certain neighborhoods are red-lined, i.e., are disqualified for mortgage
loans, the market mechanism will not operate freely, and the property-
value estimates will be in error.
9. If rents or prices are determined or heavily influenced by an outside
mechanism, such as a rent control law, the differences between quiet and
noisy property will not reflect willingness to pay and the benefits
estimates obtained from property-value analyses will be in error.
10. Statistical procedures must be used to disentangle the effects of
noise on property values from the effects of all other variables influencing
property values. This separation of influences becomes very difficult
when the factors affecting property values—e.g., noise, proximity to
airports, quality of schools, size of homes—are closely correlated, that is,
when a change in one factor is usually accompanied by a similar change
in some or all of the other factors. In cases where there are high degrees of
colinearity, that is, where the movements in the variables are closely
parallel, the calculated property-value discounts will be less reliable and
the associated benefit estimates less stable and useful.
11. The inference that property price differentials are reflections of
people's willingness to pay for quiet rests on assumptions on the nature of
human choices. There are substantial differences of opinion on the extent
to which fiscal decisions correspond to individuals' preferences. If there is
large variation between the amount of quiet an individual is willing to
purchase and the individual's preference for quiet, the estimates of
benefits based on property values will not reflect those preferences.
-------
156 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
VARIATIONS IN ESTIMATES
Tables 7.1 and 7.2 contain the estimates for marginal damage and
percentage reduction in average property values obtained from a series of
studies of jet aircraft noise. Table 7.3 contains similar estimates for
highway noise. There is a large amount of variation in these estimates
from study to study. It varies as a function of the economic, demographic,
and social characteristics of the geographic area investigated, as well as
with the data and techniques used by the investigator. For example, the
study of aircraft noise by Paik (1972) uses data collected in 1960, when jet
aircraft were recently introduced; these aircraft are different from the
aircraft under consideration in the other six studies. There are similar
problems in the studies of highway noise. The Gamble et al. study (1974)
of Bogota (New Jersey) was conducted in an area with background noise
of approximately 70 dB, which is considerably greater than that of the
other areas studied, thereby influencing the location of the origin in the
regression equation. (Background levels of NEF 25 for the aircraft studies
and 50 dB(A) for the highway studies [the 2 figures are not quite
equivalent] were assumed to be levels below which there is no noise
effect—i.e., they are the origin of the regression line.)
It should be noted that the range of variation from study to study is
large relative to the magnitude of the correction factors already
discussed. For the seven studies of aircraft noise shown in Tables 7.1 and
7.2, the mean of the marginal damage estimates is -$214/NEF with a
standard deviation of S167/NEF. (With the Paik study eliminated, the
mean marginal damage estimate is -$156/NEF with a standard deviation
of $75/NEF.) For the studies of highway noise shown in Table 7.3, the
mean deviation estimate is -S183/NEF with a standard deviation of
S207/NEF. (With the Bogota study eliminated, the mean deviation
estimate is -$106/NEF with a standard deviation of S40/NEF.)
Similarly, the percentage reduction in average property values reported
in Tables 7.1 and 7.2 has a mean of 0.75 percent with a standard deviation
of 0.60 percent. (With the Paik study omitted, the mean is 0.54 percent
and the standard deviation falls to 0.24 percent.) For highway noise, as
reported in Table 7.3, the mean is 0.62 percent with a standard deviation
of 0.72 percent. (With the Bogota study omitted, the mean is 0.35 percent
with a standard deviation of 0.15 percent.)
This range of variation has been taken into account in Chapter 9, in
which the benefit estimates have been used in cost-benefit analyses. In
that chapter, the analyses used benefit estimates ranging from the largest
value to about the mid-point of the calculated figures.
-------
Monetary Measures: Property-Value Analysis 157
THE MAGNITUDES OF THE REQUIRED ADJUSTMENTS
The illustrative reservations listed above can obviously be of considerable
significance, and they can make a considerable difference for the
estimated values obtained from observation of real estate prices. Just for
its suggestive value, we undertook some illustrative calculations in one or
two cases where plausible guesses seemed possible. For example, a rough
calculation based on the size of the relevant areas and the degree to which
they are likely to be affected by noise suggests that the figure for the
benefits of noise abatement obtained from a property-value calculation
should be adjusted upward by between 2 and 15 percent to allow for
benefits to schools, hospitals, and other properties that would not already
be reflected in neighborhood property prices.
On the other hand, a similar hybrid between guesswork and analysis
suggests that the tax advantage accorded to home ownership calls for a
downward adjustment in the benefit figures of the real-estate calculation
of between 8 and 10 percent.
These two figures are clearly not intended to be accepted literally, nor
is their objective to suggest that the required upward and downward
adjustments will approximately cancel out. Yet it is worth observing that
(1) the required adjustments do not all go in the same direction; (2) they
are not insignificant in size; (3) at this point, at best, we can offer only the
roughest sort of evaluations of their magnitudes; and (4) for some of the
adjustments we cannot even offer a reasonable conjecture about the
amount involved.
IMPERFECT INFORMATION AND THE
PROPERTY-VALUE ESTIMATE OF BENEFITS
The qualifications that have just been discussed include some that have
rather technical aspects, of interest primarily to specialists. However, to
illustrate the sorts of issues involved, we next examine in somewhat
greater detail one of the qualifications—that relating to imperfect
information on the part of purchasers of property.
One reason the difference in market values of quiet and noisy
properties may not be the same as the true cost of the noise is that
property buyers may simply not know at the time they make their
purchases how noisy the property really is or not realize how serious the
damaging effects of the noise will be. If at the time of purchase they think
a noisy house is less noisy than it really is, or if they underestimate the
resulting discomfort and damage to themselves and their families, they
are likely to pay a higher price for the property than they would have
-------
158 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
otherwise. As a result, the market prices of noisy houses will be closer to
the market prices of quiet houses and the property-value method will
underestimate the true noise damage. On the other hand, the opposite
will be true if home buyers overestimate noise damage—thinking it has
more serious physiological consequences than it really does or believing
that it will constantly disturb their sleep even though they may soon grow
used to it.
Instinctively, one tends to believe that the first of these possibilities is
more likely—that imperfect information will most frequently lead home
buyers to underestimate the noisiness of their new homes and, therefore,
that the property-value method will on this account be biased toward
underestimation of the true cost of noise. After all, there is no motivation
for sellers to exaggerate the noisiness of homes, and they do have much to
gain by concealment of noisiness. For this reason, a number of members
of this Committee are inclined to believe that imperfection of buyers'
information requires an upward adjustment in the abatement benefit
figures derived from property-value data.
However, it must be recognized that there is no firm evidence on this
matter and the arguments on the other side are strong. It is at least
possible that people imagine the degree of disturbance noise will cause
them to be greater than it is. Some observers assert that in a number of
cases this seems to have been true, with real estate values plunging
temporarily in areas that were merely suspected to be under consider-
ation as airport sites.
Those who question the view that buyers are systematically misin-
formed about noise point out that the discounts in the values of noisy
property are reasonably consistent from city to city. For example, the
effect of airplane noise, measured in dollars of lost property value per
NEF, is roughly the same for such diverse cities as Boston, Minneapolis-
St. Paul, and San Francisco. Furthermore, there is also some evidence to
suggest that there is no relationship between noise and length of
occupancy for owner-occupied housing (De Vany 1974) once other
factors are taken into account. This suggests that few recent buyers are
putting their homes back on the market, having discovered that their new
property is noisier than they had believed at the time of purchase.
There are other arguments that can be adduced on both sides, but they
would merely confirm our finding that the issue is far from settled.
CONCLUSION
A benefit-cost analysis provides no more than a reasonably well-defined
starting point from which to begin an examination of proposed public
-------
Monetary Measures: Property-Value Analysis 159
programs, such as those designed to reduce noise and its effects. It is
essential to proceed beyond mere benefit-cost calculations, to examine
issues such as income-distribution effects, political and technical feasibili-
ty, legality, and overall social consequences. Cost-benefit techniques have
only a limited capacity to incorporate information about social values,
political effectiveness, or moral judgments—considerations that influence
public decisions. It follows that an economic analysis, particularly one
relying on surrogate measures such as property values to evaluate health,
psychic, and social consequences, leaves a variety of judgments that must
be made by the decision maker as an adjunct to the economic
calculations. Decisions about noise abatement programs are also
decisions about style and quality of life, about the social benefits of health
and welfare, about government intervention in personal decisions, and
about the relative value of short- and long-term effects. The cost-benefit
analysis properly constitutes the beginning of the decision process, not its
end.
REFERENCES
Crowley, R.W. (1973) A case study of the effects of an airport on land values. Journal of
Transport Economics and Policy 7:144-152.
De Vany, A. (1974) The Measuremental Cost of Airport Noise. Environmental Quality
Program. Note 17. College Station, Tex: Texas A&M University.
Emerson, F.C. (1969) The Determinants of Residential Value with Special Reference to the
Effects of Aircraft Nuisance and Other Environmental Features. PhD dissertation.
Minneapolis: University of Minnesota.
Federal Aviation Administration and the U.S. Department of Transportation (1976)
General operating and flight rules: Noise abatement landing flap amendment and
decision not to prescribe two-segment approach requirements submitted by Environmen-
tal Protection Agency. 41 FR 52388.
Freeman, A.M., III (1974) On estimating air pollution control benefits from land value
studies. Journal of Environmental Economics and Management l(l):74-83.
Gamble, H.B., C.J. Langley, R.D. Pashek, O.H. Sauerlender, R.D. Twark, and R.H.
Downing (1974) The Influence of Highway Environmental Effects on Residential
Property Values. Institute for Research on Land and Water Resources, Research
Publication No. 78. University Park, Pa.: Pennsylvania State University.
Illinois Institute for Environmental Quality (1976) Economic Impact Study of the Proposed
Motor Vehicle (In-Use) Noise Regulations. Report of the Task Force on Noise,
Document No. 76/10. Chicago: Illinois Institute for Environmental Quality.
National Bureau of Standards (1971) The Economic Impact of Noise. EPA-NTID 300.14.
Washington, D.C.: U.S. Environmental Protection Agency; PB-206 726. Springfield,
Va.: National Technical Information Service.
National Research Council (1974) Air Quality and Automobile Emission Control. Volume
4: The Costs and Benefits of Automobile Emission Control. A Report by the
Coordinating Committee on Air Quality Studies, National Academy of Sciences,
-------
160 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
National Academy of Engineering for the Committee on Public Works. Committee
Serial No. 93-24,93rd Congress, 2nd Session.
Nelson, J.P. (1975) The Effects of Mobile-Source Air and Noise Pollution on Residential
Property Values. DOT-TST-75-76. Washington, D.C.: U.S. Department of Transporta-
tion; PB-241 570/1GA. Springfield, Va.: National Technical Information Service.
Nelson, J.P. (1976) Aircraft Noise, Residential Property Values and Public Policy. Institute
for Research on Human Resources. University Park, Pa.: Pennsylvania State University.
Oron, Y., D. Pines, and E. Sheshinski (1974) The effect of nuisances associated with urban
taffic on suburbanization and land values. Journal of Urban Economics l(4):382-394.
Paik, I.K. (1972) Measurement of Environmental Externality in Particular Reference to
Noise. PhD dissertation. Washington, D.C.: Georgetown University.
Polinsky, A.M. and S. Shavell (1976) Amenities and property values in a model of an urban
area. Journal of Public Economics 5(1,2): 119-129.
Safeer, H.B. (1975) Analysis of the costs, effectiveness, and benefits of aircraft noise
reduction programs. Paper No. 750595, Warrendale, Pa.: Society of Automotive
Engineers.
U.S. Environmental Protection Agency (1974a) Background Document for Proposed
Medium and Heavy Truck Noise Regulations. Surface Transportation Branch, EPA-
550/9-74-018. Washington, D.C.: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency (1974b) Background Document for Interstate
Motor Carrier Noise Emission Regulations. Office of Noise Abatement and Control,
EPA-550/9-74-017. Washington, D.C.: U.S. Environmental Protection Agency; PB-242
554/4BE. Springfield, Va.: National Technical Information Service.
U.S. Environmental Protection Agency (1974c) Civil Subsonic Turbojet Engine-Powered
Airplanes (Retrofit and Fleet Noise Levels). EPA Project Report. Office of Noise
Abatement. Washington, D.C.: U.S. Environmental Protection Agency.
Vaughan, RJ. and L. Huckins (1975) The Economics of Expressway Noise Pollution
Abatement. Report No. P-5475. Santa Monica, Calif: Rand Corporation.
Walters, A.A. (1975) Noise and Prices. London: Oxford University Press.
-------
8
Costs of
Noise
Abatement
INTRODUCTION
There is a lack of information on the general costs of noise abatement
except for a few types of vehicles: commercial aircraft (see FAA 1976),
heavy and medium trucks (see U.S. EPA 1976), and locomotives (see U.S.
EPA 1975) are the major exceptions. However, even for these vehicles
there is a wide difference of opinion on the magnitude of some of the
costs, the appropriate method of accounting for the costs, the meaning of
the benefits derived, and the appropriate relationships among future costs
and benefits.
Two basic types of cost-abatement programs merit particular empha-
sis: retrofitting (the installation of noise-reduction equipment to existing
vehicles) and regulation of new vehicles. Retrofitting programs lend
themselves to reasonably reliable cost estimates once the noise-reduction
equipment has been designed and tested, as in the FAA programs. Cost
estimates for regulatory programs are more difficult and suffer from
considerable uncertainty, primarily because it is difficult to estimate the
cost imposed by adding a noise performance requirement to the other
performance requirements associated with the design of a new vehicle.
Consequently, the approach often is based on the cost of adding noise-
reduction equipment to a new vehicle of existing design, an approach that
probably overestimates the real future cost by a significant margin.
Overall, the costs of noise abatement for transportation vehicles are
significant. For surface vehicles, such as automobiles and trucks, this is
161
-------
162 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
largely attributable to the large number of vehicles involved; for example,
with 10 million automobiles purchased every year, a $10 increase per new
automobile yields a total cost of $100 million per year. For aircraft, the
unit cost of abatement is very high; for example, a per-vehicle cost of
$500,000 for 2000 aircraft yields a total cost of $1 billion.
The remainder of this chapter suggests the magnitudes of the costs
associated with some proposed or possible noise abatement programs,
with detailed examples for commercial aircraft. A variety of estimates of
the costs of noise abatement are listed for commercial aircraft and for
other vehicles that contribute to urban noise. The cost estimates are
highly controversial and are presented in this report only to provide
examples of the general magnitude of the costs that can be anticipated
and that can be used in examples of cost-benefit analysis.
AIRCRAFT
EXPECTED TRENDS IN AIRCRAFT NOISE
When Federal Aviation Regulation Part 36 (FAR 36) was promulgated in
1969, it was expected that as new airplanes complying with the noise
standards of FAR 36 entered the fleet (replacing those certificated before
1969), the noise near airports would diminish. Those expectations were
encouraged by the certification of the DC-10 and L-1011 aircraft (whose
noise levels are 13 to 18 decibels lower than those of the B-707 and DC-8
aircraft, which they were expected to replace), and between 1970 and
1973 total noise near airports was reduced. However, at the end of 1973
several events occurred that not only slowed the reduction in noise, but
also reversed the trend. The oil embargo, accompanied by the general
economic recession, led to a decrease in air travel, which in turn resulted
in an increase in air transportation costs, excess capacity, declining
industry profits, and, by 1975, net losses for the air carrier industry. In
response, some airlines grounded their quieter wide-body airplanes in
favor of the smaller, but noisier, narrow-body airplanes; other airlines
sold their wide-body airplanes; and, in general, orders for newer, quieter
airplanes were either deferred indefinitely or cancelled outright. Thus, by
mid-1976,7 years after the passage of FAR 36, only 22 percent of the U.S.
air carrier fleet met the noise standards. It is now estimated that unless
there is a drastic reversal of industry economic trends or specific federal
action, some 48 percent of the air carrier fleet in 1990 will still not meet
the noise standards of FAR 36.
Total noise exposure is a function of the absolute noise levels of the
individual airplanes and the number of operations at any airport. In
-------
Costs of Noise Abatement
163
LL
O
to
FAR-36 Airplanes
Introduced
1st Generation of
Newer, Quieter
Airplanes Introduced
Replacement
Complete
2nd Generation of
Newer, Quieter
Airplanes Introduced
2nd Replacement
Complete
TIME
FIGURE 8.1 Cumulative noise exposure.
order to keep the cumulative noise level constant, a 3-dB reduction in
aircraft noise levels is required for every doubling of operations. As a
result, even if all of the new airplanes acquired in the future meet the
standards of FAR 36, a short-term gradual reduction in cumulative
exposure would eventually be reversed as the increase in operations will,
once again, increase total exposure. Since airplanes are kept in service for
10 to 15 years (or longer if warranted) depending upon economic
conditions, complete turnover of the fleet can take as long as 30 years.
Over the long run, therefore, one may expect a series of oscillations in
cumulative exposure, with total exposure decreasing as newer, quieter
airplanes replace older, noisier ones, and then, once replacement is
complete, cumulative exposure increasing as the number of airplanes and
operations increase until the next generation of airplanes is introduced
and the cycle starts again. The increase in cumulative exposure will occur
in two ways: a modest increase in exposure at current major airports, and
the exposure of new populations near new and expanding airports. This is
illustrated in Figure 8.1. Some of the predicted increase in cumulative
noise exposure will come not only from an increase in operations, but
also from the introduction of commercial jet aircraft to airports now
served by propellor-driven aircraft.
-------
164 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
MEASURES FOR REDUCTION OF AIRCRAFT NOISE
The immediate problem is how to reduce noise levels at a rate faster than
shown in Figure 8.1. There are 4 major ways in which the noise emitted
by airplanes can be reduced:
1. Retrofitting older, noisier airplanes with new engine nacelles
containing sound absorbing material (SAM);
2. Replacing the engines of older airplanes with quieter engines that
are fuel-efficient or modifying old engines as proposed in the REFAN
program for JT8D engines (discussed in Chapter 7);
3. Accelerating replacement of older airplanes with quieter, more
fuel-efficient airplanes using new technology; and
4. Modifying airplane operating procedures, including
a. reduced thrust on takeoff and
b. use of reduced flap/reduced thrust approaches.
In addition to these four ways to modify the airplanes themselves or the
way they are operated, cumulative noise exposure near an airport can be
reduced by changing airport operations or conditions near airports:
1. reducing the total number of operations;
2. limiting the number of operations of "noisy" airplanes;
3. reducing the number of night-time operations;
4. routing airplanes over nonresidential areas;
5. modifying land use regulations to permit only activities compat-
ible with noisy airports; and
6. insulating homes to reduce interior noise levels.
COSTS OF AIRCRAFT NOISE ABATEMENT
The cost of reducing aircraft noise obviously depends on the per-unit
cost, the number of airplanes affected, and the time period over which the
program is conducted. In the case of aircraft, one of the important
influences on the costs of noise abatement is the rate of retirement of
older, noisier aircraft.
While the issues and programs to reduce noise can be concisely
presented, the costs of each program vary with the assumptions and time
period for which each action is proposed. Table 8.1 lists the number of
aircraft of various types that did not meet the FAR 36 standards in 1976
and an estimate of the number of aircraft that will still not be in
compliance with the standards in 1982. Table 8.2 reports the estimated
-------
Costs of Noise Abatement 165
TABLE 8.1 Estimated Air Carrier Fleet Not Meeting FAR 36 Noise Stand-
ards, 1976 and 1982
Type 1976 1982
B-707, DC-8, B-720 487 391
B-727 572 459
B-737, DC-9 480 448
B-747 45 45
Total 1,584 1,343
SOURCE: FAA(1976:B-1)
costs of bringing these current aircraft into compliance with FAR 36.
(These unit cost data are incorporated into the cost-benefit analysis in
Chapter 9.)
From Table 8.2 we can observe that, on a unit-cost basis, REFAN
modifications are substantially more costly than SAM retrofitting alone.
For example, SAM retrofitting of a B-737—the most numerous type of
aircraft not meeting FAR 36 standards—would require a capital outlay
of about $0.27 million per aircraft compared with $1.92 million for a
refanned engine in addition to the use of sound absorption materials.
This implies that the incremental cost per aircraft per REFAN installa-
tion is $1.65 million ($1.92-$0.27). The total cost for the less expensive
SAM retrofitting of the 1982 fleet of non-compliant aircraft listed in
Table 8.1 is about $704 million (FAA cost estimates expressed in 1975
dollars).
MOTOR VEHICLES
MOTOR VEHICLE NOISE EMISSIONS
There are six types of surface vehicles that are important noise emitters:
heavy trucks, medium trucks, light trucks, automobiles, motorcycles, and
buses. Before the cost of the reduction of motor vehicle noise can be
determined, the amount of noise emitted by each type of vehicle must be
known. The magnitude of the sound produced by vehicles in each of
these categories can be described in terms of noise energy emitted as they
pass by 50 feet from a fixed monitoring point. Tables 8.3 through 8.8
describe these values for, respectively, heavy trucks, medium trucks, light
trucks and automobiles, motorcycles, intercity transit buses, and school
buses. In those tables, the regulatory levels for motor vehicles represent
the maximum permissible noise levels under any mode of operation. A
-------
TABLE 8.2 Estimated Costs per Aircraft to Comply with FAR 36 (Millions of 1975 Dollars)
Aircraft
Type1
SAM
B-707
DC-8
B-727
B-737
DC-9
B-747
REF AN (Includes SAM)
B-727
B-737
DC-9
Capital Costs
FAA
Est.2
$1.2007
1.2007
0.225
0.270
0.270
0.250
-
—
-
DOT
Est.3
$1.200
1.020
0.225
0.264
0.231
-
$2.250
1.920
1.270
REF AN
Lost Time
Cost4
$0.094
0.102
0.
0.
0.
-
$0.078
0.020
0.028
Percent IncrC3.sc
in Direct
Operating Costs5
0.5
0.6
0.1
0.2
0.1
-
2.35
2.58
2.52
Percent Increase
in Fuel
Consumption
0.2
0.2
0.
0.
0.
-
2.5
2.5
0.5
SAM = Sound absorption material applied to engine nacelles.
REFAN = Refanned engines in JT8D-powered aircraft.
2FAA(1976:D-39).
3Bartel et al. (1974:2-112), converted to 1975 dollars.
4Bartelet al. (1974:2-116), converted to 1975 dollars.
SBarteletal. (1974:2-117).
6Bartel et al. (1974:2-125).
7$1.200 million per aircraft if 270 aircraft modified; 2.6 million per aircraft if 100 aircraft are modified.
-------
Costs of Noise Abatement 167
vehicle must be designed, therefore, so that during some of its operation it
will emit less noise than the maximum permissible in order not to exceed
the standard when in its noisiest mode of operation. The test conditions
used to collect the data for Tables 8.3 through 8.8, as described in those
tables, result in baseline noise averages—current operations—and
expected values under future regulatory levels of various degrees of
severity.
Table 8.3 presents the noise data for heavy trucks. Noise emissions
vary as a function of the operating conditions of the vehicle; the levels
presented represent only a sample of the noise estimates that could be
made. For example, more noise would be emitted if the trucks were
traveling at higher speeds. Table 8.4 presents similar data for medium
trucks.
Noise emission levels for automobiles and light trucks operating at 25
and 35 mph are described jointly in Table 8.5, along with the assumptions
and conditions that underlie the estimates. Similar information for
motorcycles is presented in Table 8.6, for intercity buses in Table 8.7, and
for school buses in Table 8.8.
COSTS OF MOTOR VEHICLE NOISE ABATEMENT
The costs of abating motor vehicle noise vary with the severity of the
regulatory standard desired: the more stringent the regulatory standard,
the higher (usually) the cost of producing and operating the affected
vehicles. The cost estimates in this section consider several alternative
regulatory levels. These figures and their methods of calculation are
described in Tables 8.3 through 8.8.
Heavy and Medium Trucks
Table 8.9 lists the annual production for 1976 and for 1984 of four types
of truck—heavy and medium, gas and diesel.
The production and operating costs for the four types of trucks are
shown in Table 8.10 for each of four regulatory levels of noise emission.
For example, to attain a reduction of the noise emissions of heavy diesel
trucks to an 83-dB(A) level would cost $185-$431 per vehicle. The lower
of the two sets of estimates ("most likely capital costs") in Table 8.10 take
into account the probable use of quieter engines and any decreases in
manufacturing costs that may result from increased future production.
Annual operating (maintenance and fuel) costs for heavy diesel trucks
would increase approximately $48 per vehicle while the savings in
operating costs would total $429.
-------
TABLE 8.3 Estimated Energy Average Maximum Passby Noise Levels for Heavy Trucks at 50 ft at Urban Speeds1
Operating Mode Mixed2
Regulatory
Level
None
83dB(A)
80dB(A)
75dB(A)
25-mph Cruise
dB(A)
80.9
77.4
74.7
70.9
35-mph Cruise
dB(A)
81.9
78.2
76.2
73.8
Acceleration
dB(A)
86.6
79.2
76.4
72.3
25-mph Cruise
dB(A)
82.8
78.0
75.8
71.2
35-mph Cruise
dB(A)
83.3
78.4
76.2
73.5
For unregulated heavy trucks, the levels given in Table 8.3 for 25-mph cruise are based on survey data (Sharp 1974). The estimates for 35-mph
cruise are based on the 25-mph data with an appropriate correction for tire noise, which is approximately 66 dB at 25 mph and 72 dB at 35
mph (Hornett and Williamson 1975). For cruising heavy trucks subject to noise emission regulations, the engine-related noise is assumed to be
approximately 6.5 dB(A) below the regulatory level (National Bureau of Standards 1970); 2.5 dBA as a design tolerance for compliance with a
not-to-exceed regulatory level, 3.0 dBA for differences in test and cruise modes of operation, and 1.0 dBA to compensate for differences in test
and roadside sites. For accelerating heavy trucks, this procedure is repeated, except a 1.0-dBA difference is used for test and acceleration modes
of operation. Regulations are assumed to be based on test procedures of the Society of Automotive Engineers (SAE), where passby noise levels
are measured at 50 ft under wideopen throttle conditions.
20% acceleration and 80% cruise.
-------
TABLE 8.4 Estimated Energy Average Maximum Passby Noise Levels for Medium Trucks at 50 ft at Urban Speeds1
Operating Mode Mixed2
Regulatory
Level
None
83dB(A)
80dB(A)
75db(A)
25-mph Cruise
dB(A)
74.3
74.3
73.4
70.9
35-mph Cruise
dB(A)
76.4
76.4
75.5
73.8
Acceleration
dB(A)
78.6
77.5
76.4
72.3
25-mph Cruise
dB(A)
79.6
75.2
74.2
71.2
35-mph Cruise
dB(A)
76.9
76.6
75.7
73.5
!The data were estimated using the same set of assumptions described in Table 8.3 for heavy trucks. In cases where the level estimated from the
regulatory level is higher than the level for existing medium trucks, the regulations are assumed to have no impact on the median passby level
and the median level for existing medium trucks entered in the table. Regulations are assumed to be based on SAE test procedures, where pass-
by noise levels are measured at 50 ft under wide-open throttle conditions.
220% acceleration and 80% cruise.
-------
TABLE 8.5 Estimated Energy Average Maximum Noise Levels for Automobiles and Light Trucks at 50 ft1
Operating Mode Mixed2
Regulatory
Level
None
70dB(A)3
67dB(A)
65dB(A)
25-mph Cruise
dB(A)
65.6
64.7
60.2
59.1
35-mph Cruise
dB(A)
67.0
66.1
63.5
62.5
Acceleration
dB(A)
68.6
67.5
65.5
63.5
25-mph Cruise
dB(A)
69.0
67.4
63.5
61.7
35-mph Cruise
dB(A)
69.6
67.9
64.6
63.5
Existing automobiles and light trucks, accelerating and cruising at 35 mph, emit median levels of 68.6 and 67 dBA, respectively. For 25-mph
cruising automobiles and light trucks, the 35-mph tire noise level is corrected to a speed of 25 mph, using 40 log V, where V is the vehicle speed.
Since the noise levels measured according to the SAE J986a test procedure do not correlate well with the levels observed under typical operat-
ing modes, an energy-average multimodal test is assumed for regulations on noise emissions from automobiles and light trucks. The levels for
the 35-mph cruise, 1/4-g acceleration, idle and wide-open throttle operating modes are selected such that the weighted energy-average is 2.5
dBA below each regulatory level in the table. The 25-mph cruise is computed by correcting the 35-mph tire noise level to a speed of 25 mph.
Light trucks and automobiles measured under wide-open throttle conditions, such as specified in the SAE I986a test procedure, correlate
poorly with passby levels measured under typical operating conditions (see Rentz, P. E. and L. D. Pope, Description and Control of Motor Ve-
hicle Noise Sources. Vol. 2, Establishment of Standards for Highway Noise Levels, Final Report. Prepared by Bolt, Beranek and Newman, Inc.
BBN Report 2739 for the Transportation Research Board, NCHRP 3-7/3, 1974). Therefore, the assumed regulations on light trucks and auto-
mobiles are based on a multimodal test in which a weighted energy-average of passby levels measured under different operating conditions is
taken.
-------
TABLE 8.6 Estimated Energy Average Maximum Passby Noise Levels for Motorcycles at 50 ft1
Operating Mode Mixed
2
Regulatory
Level
None
83dB(A)
80dB(A)
75dB(A)
25-mph Cruise
dB(A)
73.2
72.1
70.8
66.4
35-mph Cruise
dB(A)
73.2
72.1
70.8
66.4
Acceleration
dB(A)
82.9
79.0
77.0
71.0
25-mph Cruise
dB(A)
78.9
75.7
73.9
68.5
35-mph Cruise
dB(A)
78.9
75.7
73.9
68.5
'the estimated energy average levels for existing motorcycles operating in cruise and acceleration modes are 73.2 dBA and 82.9 dBA, respec-
tively (private communication from S. Edwards, EPAIONAC, 23 June 1976). The same level is used for 25-mph and 35-mph cruise, since tire
noise, the most speed-dependent noise component, is expected to be negligible. For regulated motorcycles, the energy average level under ac-
celerating conditions is approximately 4 dBA below the regulatory level to allow for design tolerance to comply with a not-to-exceed regulatory
level, differences in test and typical acceleration operational modes and compensate for differences in test and roadside sites. Regulations are
assumed to be based on SAE test procedures, where passby noise levels are measured at 50 ft under wide-open throttle conditions.
233% acceleration and 67% cruise.
-------
TABLE 8.7 Estimated Energy Average Maximum Passby Noise Levels for Intercity Transit Buses at 50 ft1
Regulatory
Level
None
83dB(A)
80dB(A)
75dB(A)
Operating Mode
25-mph Cruise2
dB(A)
76.4
75.5
74.7
70.9
35-mph Cruise
dB(A)
N/A
N/A
N/A
N/A
Acceleration
dB(A)
81.5
79.2
76.4
72.3
Mixed3
25-mph Cruise
dB(A)
N/A
N/A
N/A
N/A
35-mph Cruise
dB(A)
79.6
77.7
75.6
71.6
The energy average maximum noise levels for existing intercity buses under accelerating conditions is 81.5 (Warnix 1974). The level for 25-
mph cruise is estimated from 75 to 78 dBA 40-mph cruise noise levels by correcting the 40-mph tire noise to a speed of 25 mph (Motor Vehicle
Manufacturers Association of the United States 1927-1976). The levels for regulated intercity buses are the same as the levels for regulated
heavy trucks. Regulations are assumed to be based on SAE test procedures, where passby noise levels are measured at 50 ft under wide-open
throttle conditions.
Cruise includes deceleration.
50% acceleration and 50% cruise or deceleration.
-------
TABLE 8.8 Estimated Energy Average Maximum Passby Noise Levels for School Buses at 50 ft1
Operating Mode Mixed4
Regulatory
Level
None
83dB(A)
80dB(A)
75dB(A)
25-mph Cruise2
dB(A)
74.3
74.3
73.4
70.9
,35-mph Cruise3
dB(A)
76.4
76.4
75.5
73.8
Acceleration
dB(A)
81.9
79.2
76.4
72.3
25-mph Cruise
dB(A)
79.6
77.4
75.2
71.6
35-mph Cruise
dB(A)
80.0
78.0
76.0
73.1
'the energy average maximum noise levels for existing school buses under accelerating conditions is 81.9 dB(A) (Warnix 1974). The level
for 25-mph cruise is estimated from 75 to 78 dB(A) 40-mph cruise noise levels by correcting the 40-mph tire noise to a speed of 25 mph
(Motor Vehicle Manufacturers Association of the United States 1927-1976). The levels for regulated school buses are the same as the levels for
regulated medium trucks. Regulations are assumed to be based on SAE test procedures, where passby noise levels are measured at 50 ft under
wide-open throttle conditions.
2Cruise includes deceleration.
3 Cruise includes deceleration.
450% acceleration and 50% cruise or deceleration.
-------
174 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 8.9 Annual Production by Type of Truck, 1976 and 1984
Thousands Produced
Truck Type 1976 1984
Medium Gas 204 229
Heavy Gas 3 3
Medium Diesel 40 39
Heavy Diesel 165 248
SOURCE: U.S. EPA (1976:B-2).
The table indicates the rapidity with which costs mount as noise
abatement becomes increasingly stringent. For example, a reduction in
medium gas truck sound levels from 83 to 80 dB(A) is most likely to
increase capital costs by $96 (from $11 to $107). But a reduction from 78
to 75 dB(A) is most likely to increase the cost by $251 (from $195 to
$446).
When the production data presented in Table 8.9 are combined with
the data on cost per vehicle in Table 8.10, the total annual costs, as
displayed in Table 8.11, can be evaluated. The costs shown are sensitive
to the assumptions concerning capital costs as well as to the regulatory
level.
Light Trucks, Automobiles, Motorcycles, and Buses
The costs of noise abatement for other motor vehicles can also be
estimated. Tables 8.12, 8.13, and 8.14 present data on the annual
production of these vehicles, the per-unit production and operating costs,
and the total annual costs for noise abatement. Table 8.12 presents these
data for a regulatory level of 83 dB(A) for motorcycles and buses and 70
dB(A) for automobiles and light trucks; Table 8.13 for a regulatory level
of 80 dB(A) for motorcycles and buses and 67 dB(A) for automobiles and
light trucks; and Table 8.14 for a regulatory level of 75 dB(A) for buses
and motorcycles and 65 dB(A) for automobiles and light trucks.
All Vehicles
The total annual national cost of compliance with each regulatory
standard is shown in Table 8.15. Once again, the costs of compliance
increase sharply with regulatory stringency. For the most stringent
level—(70 dB(A) for automobiles and light trucks and 83 dB(A) for other
vehicles)—there is even a possibility of a small net savings, but each
-------
TABLE 8.10 Estimated Costs per Truck to Comply with Noise Emission Standards (1975 Dollars per Truck)
Truck Type/
Standard
Medium Gas
83 dB(A)
80 dB(A)
78 dB(A)
75 dB(A)
Heavy Gas
83 dB(A)
80 dB(A)
78 dB(A)
75 dB(A)
Medium Diesel
83 dB(A)
80 dB(A)
78 dB(A)
Heavy Diesel
83 dB(A)
80 dB(A)
78 dB(A)
75 dB(A)
Worst Case
Capital
Cost1
$ 42
218
399
805
$ 151
309
460
866
$ 516
1,029
1,283
$ 431
713
1,042
1,651
Most Likely
Capital
Cost2
$ 11
107
195
446
$120
218
334
586
$ 69
207
405
$185
328
385
770
Avg. Annual
Maintenance
Cost1
$ 11
23
108
117
$ 23
45
131
162
$ 59
96
298
$ 42
104
167
280
Avg. Annual
Maintenance
Cost
(Saving)1
$ 0
0
0
0
$ 0
0
0
0
$(66)
(66)
(66)
$(66)
(66)
(66)
(66)
Avg. Annual
Fuel
Cost1
$ 0
1
1
4
$ 1
2
2
7
$ 3
9
15
$ 6
15
18
62
Avg. Annual
Fuel
Cost
(Saving)1
$ (53)
(95)
(126)
(126)
$(308)
(308)
(308)
(308)
$ (91)
(191)
(217)
$(363)
(363)
(363)
(363)
1U.S. EPA (1976:D-2, D-3 and 6-25), converted to 1975 dollars. Assumes fan-off compliance testing for capital costs.
2U.S. EPA (1976:6-13, Table 6-6) converted to 1975 dollars.
-------
776 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 8.11 Costs of Producing and Operating Quieter Trucks (in 1975
million $ and 1976 production quantities)
Highest Most Likely
Truck Type/Standard Capital Cost (Saving) Capital Cost (Saving)
Medium Gas
83 dB(A)
80 dB(A)
78 dB(A)
75 dB(A)
Heavy Gas
83 dB(A)
80 dB(A)
78 dB(A)
75 dB(A)
Medium Diesel
83 dB(A)
80 dB(A)
78 dB(A)
Heavy Diesel
83 dB(A)
80 dB(A)
78 dB(A)
75 dB(A)
0.0
30.0
77.9
163.2
(0.4)
0.1
0.9
2.2
16.8
35.1
52.5
8.3
66.5
131.7
258.1
(6.3)
7.8
36.3
90.0
(0.5)
(0.1)
0.5
1.3
(1.0)
2.2
17.4
(32.3)
3.0
23.3
112.7
increment of abatement becomes progressively more costly, reaching a
likely capital cost of $700 million when auto and light truck noise is
reduced to 65 dB(A) and that of other vehicles to 75 dB(A).
PATH ABATEMENT AND INSULATION
In addition to treatment of the source, noise can be abated by erecting
barriers that interrupt the noise path or by insulating the receiver. This
section briefly considers path and receiver treatment for noise produced
by motor vehicles and receiver treatment for noise produced by aircraft.
COSTS OF PATH ABATEMENT
Vegetation
It is maintained by some that planting vegetation to a depth of 100 feet
will reduce motor vehicle noise by 5-8 dB(A) (Reethof 1973; see also
Beaton and Bourget 1973). There is considerable difference among
experts as to the acoustical effectiveness of plantings, although the
-------
TABLE 8.12 Estimated Annual Costs of Noise Reduction for Compliance with Regulatory Level of 83 dB(A) for Buses and
Motorcycles and 70 dB(A) for Automobiles and Light Trucks
Vehicle
Type
Intercity Bus1
School Bus1
Motorcycle 0-100 cc2
Motorcycle 100-200 cc2
Motorcycle > 200 cc2
Automobiles
Light Trucks3
Total
Population
23,000
310,000
-
-
-
-
-
Annual
Production
2,500
33,500
172,000
282,000
543,000
10,949,000
1,999,000
Production
Cost/Vehicle
(1975 $)
237
23
2
10
21
1
3
Annual
Operating
Costs/Vehicle
(1975 $)
39
9
-
—
-
—
-
Total Costs
(1975 million $)
0.7
1.1
0.3
2.8
11.4
10.9
6.0
'The bus population estimates are taken from Warnix (1974). The annual production figures are estimated from a total annual bus production
of 36,000 and the population percentage of each bus type. Because of the similarity of the noise treatments of buses and trucks, cost estimates
comparable to those for heavy trucks are applied to intercity buses and the estimates for medium trucks applied to school buses.
2The estimates of motorcycle annual production are based on a total production of 1,000,000 (private communication from S. Edwards,
EPAIONAC, 23 June 1976) and a percentage of breakdown of 17.1 percent for 0-100 cc, 28.2 percent for 100-200 cc, and 54.3 percent for
greater than 200 cc motorcycles (see Singh, J. and R. A. Renner, The Impact of Noise Abatement Standards upon the Motorcycle Industry.
A Study by International Research and Technology Corporation for the Environmental Protection Agency, 1974, unpublished). Production
costs are based on production cost estimates presented by Singh and Renner (see Singh, J. and R. A. Renner). No data are available on changes
in operating costs for noise-tr'eated motorcycles.
3The production estimates for automobiles and light trucks were obtained from Motor Vehicles Manufacturers Association of the United States
(1919-1975). The production costs for light trucks regulated at the first level (70 dBA) and the second level (67 dBA) are sales-weighted
averages of estimates given by Remington and Burroughs (see Remington, P. J. and C. B. Burroughs, Noise Control Technology for Light Trucks,
BBN Interim Report No. 3252, 28 February 1976, unpublished). The estimate for the third regulatory level (65 dBA) is derived from the esti-
mates for the first two levels. The cost estimates for automobiles were computed by multiplying the cost estimates for light trucks by the ratio
of costs for light trucks and the costs for automobiles given by General Motors (Vehicular Noise Control Environmental Activities Staff 1973).
Data are not available on changes in operating costs for noise-treated automobiles and light trucks.
-------
TABLE 8.13 Estimated Annual Costs of Noise Reduction for Compliance with Regulatory Level of 80 dB(A) for Buses and ^i
Motorcycles and 67 dB(A) for Automobiles and Light Trucks)
Vehicle
Type
Intercity Bus
School Bus1
Motorcycle 0-1 00 cc2
Motorcycle 100-200 cc2
Motorcycle > 200 cc2
Automobiles
Light Trucks3
Total
Population
23,000
310,000
_
_
_
_
-
Annual
Production
2,500
33,500
172,000
282,000
543,000
10,949,000
1,999,000
Production
Cost/Vehicle
(1975 $)
393
120
4
22
39
15
25
Annual
Operating
Costs/Vehicle
(1975 $)
97
20
—
—
_
_
-
Total Costs
(1975 million $)
1.2
4.7
0.7
6.2
21.2
164.2
50.0
The bus population estimates are taken from Warnix (1974). The annual production figures are estimated from a total annual bus production
of 36,000 and the population percentage of each bus type. Because of the similarity of the noise treatments of buses and trucks, cost estimates
comparable to those for heavy trucks are applied to intercity buses and the estimates for medium trucks applied to school buses.
The estimates of motorcycle annual production are based on a total production of 1,000,000 (private communication from S. Edwards,
EPAIONAC, 23 June 1976) and a percentage of breakdown of 17.1 percent for 0-100 cc, 28.2 percent for 100-200 cc, and 54.3 percent for
greater than 200 cc motorcycles (see Singh, J. and R. A. Renner, The Impact of Noise Abatement Standards upon the Motorcycle Industry.
A Study by International Research and Technology Corporation for the Environmental Protection Agency, 1974, unpublished). Production
costs are based on production cost estimates presented by Singh and Renner (see Singh, J. and R. A. Renner). No data are available on changes
in operating costs for noise-treated motorcycles.
3The production estimates for automobiles and light trucks were obtained from Motor Vehicles Manufacturers Association of the United States
(1919-1975). The production costs for light trucks regulated at the first level (70 dBA) and the second level (67 dBA) are sales-weighted
averages of estimates given by Remington and Burroughs (see Remington, P. J. and C. B. Burroughs, Noise Control Technology for Light Trucks,
BBN Interim Report No. 3252, 28 February 1976, unpublished). The estimate for the third regulatory level (65 dBA) is derived from the esti-
mates for the first two levels. The cost estimates for automobiles were computed by multiplying the cost estimates for light trucks by the ratio
of costs for light trucks and the costs for automobiles given by General Motors (Vehicular Noise Control Environmental Activities Staff 1973).
Data are not available on changes in operating costs for noise-treated automobiles and light trucks.
-------
TABLE 8.14 Estimated Annual Costs of Noise Reduction for Compliance with Regulatory Level of 75 dB(A) for Buses and
Motorcycles and 65 dB(A) for Automobiles and Light Trucks)
Vehicle
Type
Intercity Bus1
School Bus1
Motorcycle 0-100 cc2
Motorcycle 100-200 cc2
Motorcycle > 200 cc2
Automobiles3
Light Trucks3
Total
Population
23,000
310,000
—
-
-
—
-
Annual
Production
2,500
33,500
172,000
282,000
543,000
10,949,000
1,999,000
Production
Cost/Vehicle
(1975 $)
909
443
8
30
60
30
50
Annual
Operating
Costs/Vehicle
(1975 $)
276
101
—
—
—
_
-
Total Costs
(1975 million $)
3.0
16.9
1.4
8.5
32.6
328.5
100.0
1The bus population estimates are taken from Warnix (1974). The annual production figures are estimated from a total annual bus production
of 36,000 and the population percentage of each bus type. Because of the similarity of the noise treatments of buses and trucks, cost estimates
comparable to those for heavy trucks are applied to intercity buses and the estimates for medium trucks applied to school buses.
2The estimates of motorcycle annual production are based on a total production of 1,000,000 (private communication from S. Edwards,
EPAIONAC, 23 June 1976) and a percentage of breakdown of 17.1 percent for 0-100 cc, 28.2 percent for 100-200 cc, and 54.3 percent for
greater than 200 cc motorcycles (see Singh, J. and R. A. Renner, The Impact of Noise Abatement Standards upon the Motorcycle Industry.
A Study by International Research and Technology Corporation for the Environmental Protection Agency, 1974, unpublished). Production
costs are based on production cost estimates presented by Singh and Renner (see Singh, J. and R. A. Renner). No data are available on changes
in operating costs for noise-treated motorcycles.
3 The production estimates for automobiles and light trucks were obtained from Motor Vehicles Manufacturers Association of the United States
(1919-1975). The production costs for light trucks regulated at the first level (70 dBA) and the second level (67 dBA) are sales-weighted
averages of estimates given by Remington and Burroughs (see Remington, P. J. and C. B. Burroughs, Noise Control Technology for Light Trucks,
BBN Interim Report No. 3252, 28 February 1976, unpublished). The estimate for the third regulatory level (65 dBA) is derived from the esti-
mates for the first two levels. The cost estimates for automobiles were computed by multiplying the cost estimates for light trucks by the ratio
of costs for light trucks and the costs for automobiles given by General Motors (Vehicular Noise Control Environmental Activities Staff 1973).
Data are not available on changes in operating costs for noise-treated automobiles and light trucks.
-------
180 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 8.15 Cost of Compliance with Regulatory Noise Levels for 6 type
of Vehicles: Automobiles, Light Trucks, Medium Trucks, Heavy Trucks,
Motorcycles, and Buses
Heavy and Medium Truck
Capital Expenditure Assumption
Regulatory Level
Automobiles and Light Trucks: 70 dB(A)
Other vehicles: 83 dB(A)
Automobiles and Light Trucks: 67 dB(A)
Other vehicles: 80 dB(A)
Automobiles and Light Trucks: 65 dB(A)
Other vehicles: 75 dB(A)2
Highest
(1975 million $)
57.9
379.9
966.9
Most Likely
(1975 million
-6.9
261.1'
712.3
$)
1976 production estimates.
2Medium diesel trucks at a level of 78 dB(A).
aesthetic value is undenied and may have an effect on people's attitude
towards the noise emitter. However, the cost is not low. One estimate of
the cost of planting a mixture of shrubs and trees is $7500 per 100 square
feet or about $49,000 for a typical city block (Vaughan and Huckins
1975:46), exclusive of the costs of the land. In addition to its high cost,
this method does not lend itself to widespread use because of space
limitations in areas adjacent to highways.
Solid Barriers
Any solid barrier can serve as an effective noise attenuating device if it is
tall enough to intercept the noise path. An earthberm that would reduce
noise levels by 10 dB(A) costs between $17,000 and $29,000 per city
block, depending on whether fill must be hauled to the site (Vaughan and
Huckins 1975:48). A concrete wall that can reduce noise levels by 12-15
dB(A) costs $55-$75 per foot, or $36,000-$50,000 per city block.
Aesthetics aside, use of an earthberm or concrete wall to reduce noise
levels will depend on the initial noise levels, the site characteristics and
alternatives, and the density and value of nearby residences.
COSTS OF INSULATION
The noise emitted by the commercial aircraft fleet can be abated by
insulating receivers. Table 8.16 presents the costs of soundproofing all
-------
Costs of Noise Abatement 181
TABLE 8.16 Estimated Cost of Insulation-Soundproofing all Residences1
in NEF 30 Area by 1980
Noise Level Reduction Total Cost (Billions of 1975 Dollars)
3-7 dB 1.9
8-12 dB 3.8
13-16 dB 7*2
Sound proofing costs for residential dwellings vary with the type of construction, size
of dwelling, materials used, and level of noise reduction to be attained. In this estimate,
the first three variables were averaged.
SOURCE: Wyle Laboratories (1970).
residences currently within the NEF-30 contour for three levels of noise
reduction. A program of insulation of residences, however, does not
alleviate noise problems out-of-doors or inside of non-residential
buildings.
CONCLUSION
This chapter has presented some estimates of the costs of noise
abatement, primarily for treating commercial jet aircraft and motor
vehicles. The costs have been estimated in monetary terms, suitable for
use in other analytic techniques, such as the cost-benefit analyses
illustrated in the next chapter. While many of the cost figures in this
chapter must be treated as approximations, at best, there is far more
agreement about the methods that should be used in estimating them
than there is about the methods that should be used to calculate benefits.
The main conclusion that emerges from this chapter is that the cost of
any significant noise abatement certainly will not be small.
REFERENCES
Bartel, C., L.C. Sutherland, and L. Simpson (1974) Airport Noise Reduction Forecast:
Volume 1. Summary Report for 23 Airports. DOT-TST-75-3. Washington, D.C.: U.S.
Department of Transportation PB-239 387/4GA. Springfield, Va.: National Technical
Information Service.
Beaton, J.L. and Bourget, L. (1973) Traffic noise near highways: testing and evaluation.
Highway Research Record 448:32-45.
Federal Aviation Administration (1969) Certification procedures for products and parts:
Part 36—Noise standards: Aircraft type certification. 34 FR18355.
-------
182 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
Federal Aviation Administration (1976) FAR Part 36 Compliance Regulation; Final
Environmental Impact Statement Pursuant to Section 102(2XQ, P.L. 91-190. Washing-
ton, D.C.: U.S. Department of Transportation.
Hornett, H. and I.M. Williamson (1975) Evaluation of Stationary and Moving Motorcycle
Noise Test Methods for Use in Proposed Regulations. A3-13E-469. Prepared by
McDonnel Douglas Astronautics Company—West. Washington, D.C.: Motorcycle
Industry Council.
Motor Vehicle Manufacturers Association of the United States (1919-1975) Automobile
Facts and Figures. Annual. Detroit: Motor Vehicle Manufacturers Association of the
United States.
Motor Vehicle Manufacturers Association of the United States (1927-1976) Motor Truck
Facts. Annual. Detroit: Motor Vehicle Manufacturers Association of the United States.
(Now combined with Automobile Facts and Figures in Motor Vehicle Facts and Figures.
[1976-D.
National Bureau of Standards (1970) Truck Noise-I. Peak A-Weighted Sound Level Due to
Truck Tires. Building Research Division. Washington, D.C.: National Bureau of
Standards; PB-204 188. Springfield, Va.: National Technical Information Service.
Reethof, O. (1973) Effect of plantings on radiation of highway noise. Journal of the Air
Pollution Control Association. 23(3): 185-189.
Sharp, B.H. (1974) A Survey of Truck Noise Levels and the Effect of Regulations. Wyle
Research Report WR 74-8. Study performed under EPA Contract No. 68-01-1860 for the
Office of Noise Abatement and Control, U.S. Environmental Protection Agency. El
Segunda, Calif.: Wyle Research.
U.S. Department of Transportation (1976) Aviation Noise Abatement Policy November 18,
1976. Office of the Secretary, Federal Aviation Administration. Washington, D.C.: U.S.
Department of Transportation.
U.S. Environmental Protection Agency (1975) Background Document for Railroad Noise
Emission Standards. Office of Noise Abatement and Control, EPA-550/9-76/005.
Washington, D.C.: U.S. Environmental Protection Agency; PB-251 713/4BE.
Springfield, Va.: National Technical Information Service.
U.S. Environmental Protection Agency (1976) Background Document for Medium and
Heavy Truck Noise Emission Regulations. Office of Noise Abatement and Control,
EPA-550/9-76-008. Washington, D.C.: U.S. Environmental Protection Agency.
Vaughan, RJ. and L. Huckins (1975) The Economics of Expressway Noise Pollution
Abatement. Report No. P-5475. Santa Monica, Calif.: Rand Corporation.
Vehicular Noise Control Environmental Activities Staff (1973) Proceedings: Conference on
Motor Vehicle Noise. April 3-4, 1973. G.M. Desert Proving Ground, Mesa, Arizona.
Warren: Mich.: General Motors Corporation.
Warnix, J. (1974) Cost Effectiveness Study of Major Sources of Noise. Volume IV—Busses.
Wyle Research Report WR 73-10. Study performed under EPA Contract No. 68-01-1537
for the Office of Noise Abatement and Control, U.S. Environmental Protection Agency.
El Segunda, Calif.: Wyle Research.
Wyle Laboratories (1970) Final Report-Home Soundproofing Pilot Project for the Los
Angeles Department of Airports. Wyle Research Report WCR-70-1, DA 831. Los
Angeles, Calif.: Los Angeles Department of Airports.
-------
9
Cost-Benefit
Analysis:
Some Illustrations
INTRODUCTION
Cost-benefit analysis is a technique that assesses the probable gains from
a proposed policy or action and weighs them against probable losses. It
requires that both costs and benefits be measured in comparable terms;
the usual standard of measurement is monetary. Quite often benefits are
not directly expressible in terms of dollars, but have to be estimated
indirectly through property values, compensation payments, court
awards for damages, and the like.
The estimates for the benefits used for the cost-benefit analyses in this
chapter, and by most economic studies of transportation noise, were
obtained by multiple regression analyses of differences in property
values. Thus, there are three particular techniques involved in these cost-
benefit analyses: (1) the use of property values to estimate benefits; (2)
the use of multiple-regression methods to calculate property-value
differences; and (3) the cost-benefit analysis. However, use of the cost-
benefit technique in general does not depend on property values or on
multiple regression; it does not even depend on monetary estimates. For
example, an elected official may choose to evaluate potential costs and
benefits of a proposed policy in terms of votes for and against reelection.
The particular cost-benefit analyses in this chapter, therefore, are
illustrative both with respect to the specific examples of transportation
noise and with respect to the specific type and method of estimating
benefits.
183
-------
184 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
There is a strong analytical basis for the conclusion that some noise
abatement will yield benefits that exceed costs. That is, society must
achieve net gains on balance from some amount of noise abatement. This
is because noise emission, a by-product of economic activities, is
damaging to society, but its emitters do not pay the social costs resulting
from the damaging activities. As a result, one can be certain that noise
emitters will find it in their financial interest to spend less for noise
abatement than the amount required for maximization of the well-being
of society. Thus, at least some increase in expenditures for noise
abatement will be beneficial to society.1
While one can conclude that some noise abatement will be beneficial,
one still has to determine just what programs will in fact yield benefits
greater than their costs and which of the available alternatives would be
most effective. This is the task that cost-benefit analysis is intended to
carry out. Unfortunately, as will be illustrated in this chapter, data
imperfection, problems of method, and other problems often prevent this
method from yielding categorical conclusions.
As indicated in the previous chapters, the range of available noise
abatement techniques and programs is very broad. Since the purpose of
this chapter is to illustrate the application of benefit-cost analysis to
transportation noise abatement, no attempt is made to provide a ranking
of all possible programs for each transportation mode. Indeed, the state-
of-the-art only permits the analysis in this chapter to cover five jet aircraft
noise abatement projects which are not identified explicitly and four
noise abatement programs for medium and heavy trucks. It also gives
selected references to a few other benefit-cost studies, including analyses
of projects not considered here.
The somewhat limited scope of the analysis in this chapter reflects in
part the difficulty of benefit and cost measurements and the comparative-
ly few studies that have so far been carried out. The analysis described
here does not represent any original research by the Committee. The
discussion does, however, attempt to provide evaluative comments on
past studies where they seem appropriate.
1One may ask, however, whether it is really necessary for the state to intervene and require
noise emitters to abate noise or to pay for their emissions. If there were only a few noise
makers and a few sufferers, one might expect that they would be able to come to a voluntary
agreement or contract about the appropriate amount of noise to be generated and the
payments to be made for such amelioration of the environment. (This is the essence of the
Coase Theorem; see Coase 1960.) Unfortunately, however, voluntary arrangements of this
kind may not be possible in the abatement of noise. There are a very large number of
sufferers—and in some cases (automobiles) a large number of emitters—and it is virtually
impossible to bring these large numbers of people together in a voluntary agreement. The
state does have a role to play.
-------
Cost-Benefit Analysis: Some Illustrations 185
The first step in a benefit-cost analysis is to select an indicator of
achievement or success by which alternative projects can be ranked.
Where the total budget for noise abatement projects is fixed, total
expenditures should be distributed among projects so that no increase in
total benefits can be achieved by an incremental reallocation from any
one project to another. This requires a ranking based on the absolute
differences between benefits and costs of the different projects after
discounting for those costs and benefits that are expected in the more
distant future. This criterion is referred to as the maximum net present
value.
As an alternative criterion, projects are sometimes ranked by their
benefit-cost ratio, defined as the present value of project benefits divided
by the present value of project costs. It can be shown that the benefit-cost
ratio will normally lead to a different and inferior set of choices, but this
has not prevented its widespread use. The analysis in this chapter
employs both of these criteria. Finally, no attempt has been made to
extend the analysis to take actual budget constraints or other pertinent
complications into account.
TYPES OF BENEFITS AND COSTS
It is hardly enough to say that all "relevant" or "pertinent" benefits and
costs should be included in a benefit-cost analysis. The important
problem is to decide which benefits and costs are relevant, whe.ther or not
they can be measured, and how they should be valued. Some aspects of
this problem have been discussed in Chapters 6, 7, and 8, so that this
section will only summarize briefly some of the issues involved. There is
also the important issue of the distribution of the costs and benefits
among different groups of the population. (Some parts of this issue are
considered in Chapter 1 in this report.) Distribution issues are sometimes
handled by the assignment of explicit weights to different groups of
recipients, with benefits usually assigned higher weights if they go to
poorer recipients. The analysis here only gives unweighted values for
benefits and costs. (For an example of weighting in cost-benefit analysis
of a noise abatement issue, see Nwaneri's study of the site for the Third
London Airport [Nwaneri 1970]; see also Pearce and Wise [1972].)
REAL VS. PECUNIARY BENEFITS AND COSTS
Tables 9.1 and 9.2 list illustrative categories of benefits and costs that may
be expected from aircraft and truck noise abatement projects, respective-
ly. The tables only provide a few examples in each category. The
-------
186 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 9.1 Illustrative Benefits and Costs of Aircraft Noise Abatement1
Types Benefits Costs
Real
Direct tangible Reduction in hearing loss Costs of inputs
intangible Reduction in annoyance Government intervention
in local affairs
Indirect tangible Improved worker productivity Costs of regulation
intangible Reduction in antisocial Disutility of household
behavior moving
Pecuniary Relative improvement in the Relative reduction in the
economic position of the economic position of
aviation industry commercial airlines
1The benefits and costs in this table merely illustrate the types of benefits and costs that
can occur; it makes no attempt to be comprehensive.
TABLE 9.2 Illustrative Benefits and Costs of Truck Noise Abatement1
Types Benefits Costs
Real
Direct tangible Improved learning Cost of inputs
intangible Reduction in sleep loss More complaints
Indirect tangible Fuel savings Costs of regulation
intangible Reduction in antisocial Less masking of other
behavior noises
Pecuniary Relative improvement in Relative reduction in the
the economic position economic position of
of equipment manufacturers the trucking industry
The benefits and costs in this table merely illustrate the types of benefits and costs that
can occur; it makes no attempt to be comprehensive.
categories are designed to suggest some of the associated problems of
relevance, ease of measurement, and valuation.
The first distinction suggested by Tables 9.1 and 9.2 is that between
real and pecuniary costs and benefits. Real costs represent the use of
physical resources required for the abatement of noise—the metal and
fuel needed to produce abatement devices or insulation. Real benefits
represent the increase in quiet to the consumers who no longer suffer as
much physiological or psychological harm or annoyance as before.
Pecuniary benefits and costs, on the other hand, are those resulting from
price changes caused by a noise abatement project. For example, it may
tend to raise the wages of skilled labor employed in retrofitting aircraft
-------
Cost-Benefit Analysis: Some Illustrations 187
and to decrease the wages of, say, workers employed in the construction
of insulated porches and double-glazed windows. The gains that accrue
to some parts of society are offset by losses to other parts and generally
represent neither a net gain or loss to society as a whole. Thus,
employment or sales changes in the aerospace or airlines industries
should not be included in an analysis of the real benefits and costs of
aircraft noise abatement.2
DIRECT VS. INDIRECT BENEFITS AND COSTS
A second distinction suggested by Tables 9.1 and 9.2 is that between
direct (primary) and indirect (secondary) benefits or costs, a distinction
that is to some extent arbitrary. For example, the use of cooling fan
clutches for the abatement of truck noise will yield an indirect benefit in
fuel saving. It may also impose an indirect cost if truck operators are
induced to avoid regulations by routing trucks over back roads to escape
inspection. While indirect real costs and benefits are relevant, they are
difficult to measure exhaustively since they are likely to be spread widely
in the economy and to take unexpected forms.
TANGIBLE AND INTANGIBLE BENEFITS AND COSTS
The final distinction indicated in Tables 9.1 and 9.2 is that between
tangible and intangible benefits and costs: tangible benefits and costs are
those whose monetary value is observable directly; those whose monetary
value cannot be observed directly are intangible. Noise abatement
benefits are, by and large, intangible; consequently, they must be
quantified by indirect procedures, which are more likely to give rise to
serious measurement questions. The use of noise easements (see Baxter
and Altree 1972) does approximate a market for quiet on a local level,
although various special problems inhibit the use of this mechanism as a
general basis for the valuation of the benefits of noise abatement.
At present, the only method that has been used systematically to
measure the intangible benefits of noise abatement is based on observa-
tion of the association between residential property values (or apartment
rents) and levels of environmental noise. This method, which assumes
that the difference in the prices of properties in quiet and noisy
neighborhoods reflects to some degree the valuation of quiet by the
2This reasoning may break down if there is widespread unemployment, so that every lost job
represents a major loss to society as well as to the person directly affected. But a particular
abatement project may not be the best way of using the unemployed. Unless their use in
other projects is considered, the analysis may be seriously deficient.
-------
188 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
general public, has been described in detail and discussed critically in
Chapter 7. Because no satisfactory substitute is, at least so far, available,
it will be the basis for all of the quantification of noise abatement benefits
in the next two sections of this chapter, with the exception of indirect fuel
savings for trucks.
ILLUSTRATIVE BENEFIT-COST ANALYSES OF
AIRCRAFT NOISE ABATEMENT
This section illustrates the procedure that can be used to calculate the
costs and benefits of commercial jet aircraft noise abatement projects.
Whenever possible, an effort should be made to examine the sensitivity of
the analysis to variations in some of the critical data whose values may be
subject to question, and this, too, is illustrated in this section. A benefit-
cost analysis accompanied by a sensitivity analysis not only shows a
project's ranking, but also gives an indication of the extent to which the
ranking will change as a result of changes in such variables as the state of
the economy (e.g., the rate of economic growth, interest rates) or in the
preferences of the policy maker (e.g., the implicit weight assigned to the
employment or technological effect of a retrofitting program). This
section examines two analyses of the benefits and costs of some aircraft
noise abatement projects. (For additional examples, the reader is referred
to Walters [1975], Council on Wage and Price Stability [1975a, 1975b],
and FAA[ 1976].)
ILLUSTRATIVE ANALYSIS I
The first example of a benefit-cost analysis for jet aircraft noise
abatement is based on the work of Nelson (1976), which is a revision of
earlier work (Nelson 1975). It examines a policy of no change and three
alternative programs, called A, B, and C.
The case involving no change in current noise abatement policy is
defined as the baseline program; it is intended primarily as a standard of
comparison for the other programs considered. (All noise abatement
programs considered here reflect noise reductions resulting from thrust
cutbacks on take-off.) Even with no change in policy, noise levels will
decrease as the new, wide-body jets (B-747, DC-10, L-1011, and others)
are introduced into the fleet and older, noisier jets are retired from
commercial use. The baseline benefit estimate, reflecting the noise
reductions expected from fleet mix changes, thus depends on a forecast of
the rate of retirement of existing jets and levels of aviation activity in
general.
-------
Cost-Benefit Analysis: Some Illustrations 189
Because the purpose of this chapter is to describe and comment on the
techniques of benefit-cost analysis rather than to endorse any evaluation
of some particular abatement procedure, we do not identify the three
alternatives that will be described. All of them involved some
modification in the operating procedures or the equipment used in older
aircraft now in operation. Because of limitations of the data and because
in each case some important considerations, such as the effect upon
safety, are not taken into account, the figures as given could not be
legitimately used as the basis for policy conclusions about the choice of
noise abatement technique.
The general method used to calculate costs and benefits for each of the
programs was described in Chapters 7 and 8. Since for each program the
relevant costs are the additions in costs resulting from the introduction of
the program, no cost estimates are needed for the baseline program. The
types of cost included in the calculations were: (a) investment and
installation costs required to carry out any modifications in equipment
required by the three programs; (b) any lost flight time incurred during
installation of any new equipment; and (c) any resulting direct and
indirect increases in operating cost. Direct and indirect operating cost
increases are taken to be incurred continually until the aircraft in
question is retired from the fleet.
Benefit estimates were based on: (a) the estimated incremental
reductions in noise exposure forecast (NEF) levels ascribable to each
program; (b) an estimated incremental benefit per NEF of $175 per
household in 1975, as indicated by real-estate values; and (c) an assumed
increase in annual benefits per household of 3 percent per year from 1975
to 1987 and 2 percent per year from 1988 to 2001. The growth rate of
incremental benefits is based on anticipated increases in average real
incomes and increases in the willingness to pay for quiet.
Nelson examined the sensitivity of the benefit estimates to changes in
the total population exposed to NEF 30+ and in the discount rate used
to calculate present values for the period 1977-2001. For the year 1972,
the size of the U.S. population exposed to NEF 30+ has been estimated
to be 6.2 million (Safeer 1975:1). However, even with no policy changes,
that number would have fallen to an estimated 6 million by the year 1976
as a result of the introduction of wide-body jets (FAA 1976:D-24). This
figure is somewhat smaller than that found in several reports by EPA, and
obviously excludes populations in the range of NEF 20-30.
Nelson's work also provides a sensitivity analysis of the discount rate
used to translate future benefits and costs into current dollars. As
alternatives, values of 4 percent and 8 percent were used for the real
discount rate. (Real discount rates of 4 and 8 percent are equivalent to
-------
190 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 9.3 Total Discounted Costs and Benefits of Jet Aircraft Noise
Abatement, 1977-2001, at $175 Per NEF in 1975 (Millions of 1975 Dollars)1
Abatement
Program
Baseline (No-Change)
Program A
Program B
Program C
Costs
At a 4% Real
$ 77.6
726.4
2,769.5
Benefits
Benefits
less
Costs
Benefits
Costs
Discount Rate Until the Year 2001
$1,708.5
370.6
643.5
1,696.5
$ 293.0
(82.9)
(1,073.0)
4.78
0.89
0.61
At an 8% Real Discount Rate Until the Year 2001
Baseline (No-Change)
Program A
Program B
Program C
_
$ 70.3
602.9
2,068.9
$1,097.1
237.5
467.6
1,161.7
_
$ 167.2
(135.3)
(907.2)
_
3.38
0.78
0.56
Discounted marginal benefits are $175 per NEF per household in 1975 based on an
average $2 5,000 property value and a noise discount of 0.7 percent per NEF per property.
SOURCE: Based on Nelson (1976)
market interest rates of about 8 and 12 percent, respectively, when the
rate of inflation is about 4 percent per year.) Current policies of the U.S.
Office of Management and Budget call for a 10-percent market interest
rate for evaluation of public projects (U.S. OMB 1969).
Table 9.3 shows estimated total discounted benefits and costs for the
period 1977-2001 in constant 1975 dollars. For program A, estimated
benefits always exceed costs while for programs B and C costs always
exceed benefits in all our illustrative calculations. The sensitivity analysis
reveals that the absolute differences between benefits and costs are
indeed sensitive to changes in the choice of the real discount rate.
The sensitivity analysis can be extended by varying the estimate of the
incremental benefit of noise abatement. Table 9.3 uses an estimate of
$175 per household per NEF, based on an average property value of
$25,000 in 1975 and a noise discount of 0.7 percent per NEF. The
evidence (see Chapter 7) from past empirical studies suggests a noise
discount for property values in the range of 0.4-1.0 percent per NEF,
although at least two studies have yielded discounts in excess of 1 percent
for areas near some airports (e.g., Paik 1972, Dygert 1973). The sensitivity
of the analysis to variations in this parameter can be examined by
multiplying the benefit estimates in Table 9.3 by the appropriate ratio of
-------
Cost-Benefit Analysis: Some Illustrations 191
TABLE 9.4 Total Discounted Costs and Benefits of Jet Aircraft Noise
Abatement, 1977-2001, at $250 Per NEF in 1975 (Millions of 1975 Dollars)1
Benefits
Abatement less Benefits
Program Costs Benefits Costs Costs
At a 4% Real Discount Rate Until the Year 2001
Baseline (No-Change)
Program A
Program B
Program C
$ 77.6
726.4
2,769.5
$2,440.6
530.0
919.2
2,423.5
_
$452.4
192.8
(346.0)
_
6.83
1.27
0.88
At an 8% Real Discount Rate Until the Year 2001
Baseline (No-Change)
Program A
Program B
Program C
$ 70.3
602.9
2,068.9
$1,567.2
339.6
668.0
1,659.5
_
269.3
65.1
(409.4)
_
4.83
1.11
0.80
'Discounted marginal benefits are $250 per NEF per household in 1975 based on an
average $25,000 property value and a noise discount of 1.0 percent per NEF per property.
SOURCE: Based on Nelson (1976)
noise discounts, e.g., 1.0/0.7= 1.43 or 0.4/0.7=0.57. Table 9.4 shows the
estimates of total discounted benefits and costs when the discounted
benefits of abatement are $250 ($25,000 X 0.01) per NEF per household in
1975. It shows, for example, that for program B, benefits would now
exceed costs. In other words, any ranking of noise abatement projects is
quite sensitive to estimates of the discounted benefits. This parameter is
crucial to the policy decisions to be reached with respect to aircraft noise
abatement.
ILLUSTRATIVE ANALYSIS II
The benefit-cost calculations in Tables 9.3 and 9.4 use a baseline forecast
for aviation activity that is somewhat out of date for the near future. As
the rate of aggregate economic growth declined in 1973-75, so did the rate
at which commercial aircraft were retired from the fleet. Assuming that
this pattern will continue in the future, it means that noisier, narrow-body
jets will be continued in use somewhat longer than is assumed in Tables
9.3 and 9.4.
The effect of a reduction in aircraft replacement is to increase, perhaps
significantly, the benefits from the abatement programs relative to the
-------
192 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 9.5 Assumed Incremental Reductions in the Noise Exposure
Forecast (A NEF) under Program B 1977-1995, Under Slower Retirement
of Existing Aircraft
A NEF
1977 1980 1985 1990 1995
Abatement Program B 0.0 -0.2 -0.9 -0.8 -0.1
SOURCE: FAA(1976:D-33).
revised baseline program. If more narrow-body jets are in use in each
year, the incremental reduction in noise levels achievable by the
abatement program will be greater and will extend over a longer period of
time. At the same time, direct and indirect operating costs will increase as
a result of physical obsolescence. These cost increases, however, will not
result in any further reductions in noise levels and, indeed, one would
expect the acoustical efficiency of the abatement programs to decrease
somewhat over time, so that the outlays required for a given reduction in
noise will increase. Consequently, it is not legitimate to conduct a
sensitivity analysis in this case by just varying the benefit estimates while
leaving the cost estimates unchanged.
We can examine the sensitivity of the earlier analysis to such changes
in fleet mix, accompanied, for the sake of illustration, by the assumption
that the number of aircraft to which abatement programs apply is
simultaneously reduced. For this calculation several assumptions are
made:
1. Program B affects 6 million people or about 2 million households
over the period 1976-2001.
2. The reductions in the noise exposure forecast (A NEF) are based on
the most recent FAA data and are shown in Table 9.5 for program B. The
assumed incremental reductions in NEF values for program B are shown
in Table 9.5, using the most recent FAA data.
3. The marginal capitalized benefit of a unit reduction in NEF is $175
per household in 1975 dollars. Annual benefits per household increase at
3 percent per year from 1975 to 1987 and 2 percent per year from 1988 to
2001.
4. Direct and indirect operating costs (in constant 1975 dollars)
increase after 1990 as a result of the continued operation of the fleet and
cost increases due to physical obsolescence. These increases occur despite
the more modest program contemplated here: it is assumed that 1217 jets
will be encompassed by the program between 1979 and 1986.
-------
Cost-Benefit Analysis-Some Illustrations 193
TABLE 9.6 Example of Discounted Costs and Benefits Under Slower Re-
tirement of Existing Aircraft, 1977-2001 (Millions of 1975 Dollars)1
Abatement
Program
Costs Benefits
Benefits
less
Costs
Benefits
Costs
At a 4% Real Discount Rate Until the Year 2001
Baseline
Program B
$ 34.9
$609.1 $319.9
$(289.2)
0.53
At an 8% Real Discount Rate Until the Year 2001
Baseline - $ 37.8 - -
Program B $492.3 221.6 $(270.7) 0.45
Discounted marginal benefits are $175 per NEF per household in 1975 based on an
average $25,000 property and a noise discount of 0.7 percent per NEF per property.
SOURCE: Based on Nelson (1976)
The net effect of these assumptions is to reduce both the benefits and
costs of the abatement programs, although it should be pointed out the
assumptions must be offered with somewhat stronger reservations than
those applicable to Tables 9.3 and 9.4.
Using alternative real discount rates of 4 and 8 percent, the present
values of costs and benefits were determined and shown in Table 9.6. This
illustrates the basic point to be derived from this analysis: a more modest
abatement program and a slowdown in the rate of attrition of existing
aircraft will change both the incremental benefits and costs of such a
program. As a result, this parameter (rate of attrition) turns out to be less
crucial to the ranking of projects than the noise discount or interest rate.
Table 9.6 shows that the calculated benefits of programs B and C still do
not exceed the calculated costs despite the assumed continuation of
slowdown in retirements. Indeed, this would be the case even at the
higher marginal benefit of $250 per NEF per household.
BENEFIT-COST ANALYSES OF TRUCK NOISE ABATEMENT
This section examines two studies of the benefits and costs of the
abatement of noise from medium and heavy trucks. (For analyses of
other selected projects, the reader is referred to Merewitz [1975], Vaughan
-------
194 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
and Huckins [1975], and the Illinois Institute for Environmental Quality
[1976] and the DOT 1975 projections3.) Before examining the study by
Nelson (1975) and an extension of a study by the Council on Wage and
Price Stability (1975c, 1975d) two preliminary issues are discussed briefly.
An analysis of the benefits and costs of truck noise abatement involves
greater uncertainty than a similar analysis of aircraft noise. In part, this is
a result of the limited information available on the benefits of truck noise
abatement, but it also reflects the great variety of trucking equipment and
the complexity of the urban/suburban noise milieu. It is not clear, for
example, that a reduction in truck noise will result in a corresponding
reduction in annoyance, especially if other major sources of noise in
residential areas are left unchanged. For this reason, the benefit analysis
employs a wide range of possible values for marginal abatement benefits.
Unlike the case of aircraft/airport noise, it is not entirely clear that the
benefits of abating urban/suburban truck noise can be analyzed by
considering only those properties immediately affected—the property
values near major highways and streets. Because so many areas are
affected substantially, significant changes in urban/suburban truck noise
levels would undoubtedly produce general changes in residential proper-
ty values. The analyses that follow ignore these marketwide changes in
property value as well as other analogous complications.
As a consequence of these two factors, the benefit-cost estimates in this
section are subject to fairly large and undetermined probable errors,
although this does not mean that tentative project ranking cannot be
derived from the studies. The uncertainty in the calculation does,
however, imply that any project ranking is subject to less confidence than
might be the case for aircraft noise abatement. The benefit-cost
calculations that follow are provided to indicate the current state-of-the-
art and certainly do not constitute the last word on this subject.
ILLUSTRATIVE ANALYSIS I
The first example of a benefit-cost analysis of medium and heavy truck
noise is provided by Nelson (1975). This study used population and noise
level projections from a preliminary report by EPA (1974a) to calculate
truck noise abatement benefits from residential property values. Cost
estimates were also obtained from the EPA report.
EPA has identified medium and heavy trucks with a gross vehicle
weight rating (GVWR) in excess of 10,000 pounds as a major source of
3U.S. Department of Transportation, Office of Noise Abatement (1975) Comparative
Benefits and Costs Projected for Proposed New Medium- and Heavy-Duty Truck Noise
Emission Standards. Washington, D.C.: U.S. Deprtment of Transportation. (Unpublished)
-------
Cost-Benefit Analysis: Some Illustrations 195
noise (see 39 Federal Register 38338 1974). Four alternatives, including
three noise abatement programs for these vehicles, were considered in
this benefit-cost analysis.
Baseline. The baseline assumes use of current operating rules. The
Interstate Motor Carrier Noise Emission Standards (U.S. EPA 1974b)
requires that all motor vehicles above 10,000 pounds GVWR operated by
motor carriers engaged in interstate commerce meet the following
standards as of October 1975:
a. No more than 86 dB(A) at 50 feet in speed zones at or under
35 mph under all conditions, and
b. No more than 90 dB(A) at 50 feet in speed zones over 35 mph
under all conditions.
Abatement Program 1. This program requires new trucks of over 10,000
pounds GVWR not to exceed the following noise levels after October of
the year indicated:
a. 1976 83dB(A)
b. 1980 80dB(A)
c. 1982 75dB(A)
Abatement Program 2. This program has the same noise levels as the
previous one, but with different effective dates:
a. 1976 83dB(A)
b. 1977 80dB(A)
c. 1980 75dB(A)
Abatement Program 3. This program sets separate standards for gas-
engine and diesel-engine powered trucks with the following effective
dates:
a. 1976
b. 1977
c. 1980
d. 1980
e. 1982
Gas
80dB(A)
80dB(A)
75 dB(A)
75 dB(A)
75 dB(A)
Diesel
83 dB(A)
83 dB(A)
83 dB(A)
80dB(A)
75 dB(A)
Incremental cost estimates for each noise level were obtained from the
-------
196 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
EPA report (1974a:7-29) for each noise level standard—83, 80, and 75
dB(A). The cost totals include depreciation, interest, and operating and
maintenance expenses for the first full year during which the^ limits on
noise levels become effective. Total costs were assumed to increase at 5
percent per year for 1976-1985 (Nelson 1975:10-15), reflecting the growth
of new truck sales. Thereafter, the estimate of the incremental annual
costs ascribable to the 75 dB(A) standard was fixed at $185 million per
year.
Benefits from the abatement of truck noise were calculated in a manner
analogous to the procedures used for aircraft noise benefits. For 1976-
1992, the EPA report (1974a:6-20-21) indicated the reduction in the Day-
Night Noise Levels (Lda) in dB(A) (relative to 1974) and the equivalent
number of people exposed to Ldn 55 + for each abatement program. The
analysis assumes that a l-dB(A) reduction in noise levels would result in a
marginal capitalized benefit of $64 per household. Total benefits were
assumed to grow at 5 percent per year for 1977-1983, but after 1983
benefits were extrapolated linearly, so that maximum total benefits were
assumed to be attained in 1992.
Using these procedures, annual costs and benefits were calculated for
1977-2000 and then discounted to 1976 using real discount rates of 5 and
10 percent. (These rates correspond to market interest rates of approxi-
mately 9 and 14 percent, respectively.) Tables 9.7 and 9.8 show the basic
results where the low-benefit estimate is based on a gradual decrease in
the relevant population as some individuals are no longer exposed to Ldn
55 + . The high-benefit estimate is based on an equivalent residential
population in 1974 of 37.3 million (U.S. EPA 1974a:6-20). Tables 9.7 and
9.8 show that Programs 1 and 2 are the top-ranked projects using a 5-
percent real rate of interest. However, any ranking will be sensitive to the
population affected by truck noise abatement since the low-benefit
estimate always yields a negative net present value.
The sensitivity of the analysis may again be examined for changes in
the estimated marginal benefits of noise abatement. The analysis in
Chapter 7 suggests a noise discount for traffic noise that varies between
0.2 and 0.6 percent per dB(A) for an average residential property. In
1975, the average residential property in the United States had a value of
about $25,000. This implies that marginal capitalized damages from
traffic noise range from about $50 to about $150 per dB(A), with a mean
of about $100 per dB(A). Thus, Table 9.7 is based on a value near the
lower limit of this range.
Table 9.8 shows the benefit-cost estimates when marginal capitalized
benefits are assumed to be $100 per dB(A) per household. This table
shows that higher benefit values will affect the ranking of projects, so that
-------
Cost-Benefit Analysis: Some Illustrations 197
TABLE 9.7 Total Discounted Costs and Benefits of Truck Noise Abatement,
1976-2000, at $64 per dB(A) in 1975 (Billions of Dollars)1
High
Benefits High
Abatement High , Low less Benefits
Program Costs Benefits2 Benefits2 Costs Costs
At a 5% Real Discount Rate Until the Year 2000
Baseline
Program 1
Program 2
Program 3
$1.80
2.10
1.80
$1.60
1.90
2.10
1.90
$1.30
0.90
1.00
0.90
$0.10
(0.00)
0.10
1.06
1.00
1.06
At a 10% Real Discount Rate Until the Year 2000
Baseline
Program 1
Program 2
Program 3
_
$1.00
1.20
1.00
$1.00
0.90
1.10
0.90
$0.80
0.50
0.50
0.50
_
($0.10)
(0.10)
(0.10)
_
0.90
0.92
0.90
Discounted marginal benefits are $64 per dB(A) per household.
The range of benefit estimates reflects alternative assumptions about the total equiva-
lent population that is exposed to high levels of environmental noise, i.e., Ldn > 55
dB(A).
SOURCE: Based on Nelson (1975)
Program 2 then receives the highest ranking according to the criterion of
maximum net present value. Program 2 uses an earlier time schedule,
imposing the 75 dB(A) standard in October 1980 rather than October
1982. This advanced time schedule is reflected in the present values of
both costs and benefits. The greater the residential damages due to noise,
the more there is to be lost by postponement of the 75 dB(A) standard
until 1982.
ILLUSTRATIVE ANALYSIS II
The analysis in this section parallels that contained in two reports
prepared by Robert L. Greene for the Council on Wage and Price
Stability (1975c, 1975d). These reports provide explicit and more up-to-
date information on truck noise abatement costs. In addition, Greene
estimated indirect abatement benefits arising from truck fuel savings
ascribable to the use of a demand-actuated fan clutch, reduced back
pressure in the exhaust system, less restrictive air intakes, and lower
horsepower ratings. Most of the fuel savings are due to the fan clutch.
-------
198 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 9.8 Total Discounted Costs and Benefits of Truck Noise Abatement,
1976-2000, at $100 per dB(A) in 1974 (Billions of 1975 Dollars)1
Abatement
Program
Costs
High
Benefits2
Low
Benefits2
High
Benefits
minus
Costs
High
Benefits
Costs
At a 5% Real Discount Rate Until the Year 2000
Baseline
Program 1
Program 2
Program 3
$1.80
2.10
1.80
$2.50
2.97
3.28
2.97
At a 10% Real Discount
Baseline
Program 1
Program 2
Program 3
$1.00
1.20
1.00
$1.56
1.41
1.72
1.41
$2.03
1.41
1.56
1.41
Rate Until the Year
$1.25
0.78
0.78
0.78
_
$1.17
1.18
1.17
2000
_
$0.41
0.52
0.41
—
1.65
1.56
1.65
_
1.41
1.43
1.41
1 Discounted marginal benefits are $100 per dB(A) in 1974 based on an average $25,000
property and a noise discount of 0.4 percent per dB(A) per property.
2The range of benefit estimates reflects alternative assumptions about the total equiva-
lent population that is exposed to high levels of environmental noise, i.e., L^n > 55.
SOURCE: Based on Nelson (1975)
Greene examined the variation in benefits and costs as the severity of
noise level standards is varied. Total benefits can, of course, be expected
to increase as noise levels are reduced, but beyond some point the rate of
increase in total benefits from additional quiet can be expected to decline.
Total costs of increased abatement, on the other hand, can be expected to
increase continually since it becomes more and more difficult to achieve
an additional increment of quiet with current technology.
This incremental information permits an analysis beyond the usual
cost-benefit results that, in effect, grade any proposed project on a pass-
fail basis. Instead, with incremental data, one can determine the degree of
abatement that yields maximal net social benefits. This analysis, based on
the most recent data to be found in the EPA report, Background
Document for Medium and Heavy Truck Noise Emission Regulations (U.S.
EPA 1976), examines four alternative abatement projects.
Project 1. Under this project, new trucks over 10,000 pounds GVWR
-------
Cost-Benefit Analysis: Some Illustrations 199
would be required not to exceed 83 dB(A) in 1978, with no further
reductions required after that date.
Project 2. This project requires a standard of 83 dB(A) in 1978, 80
dB(A) in 1982, and no further reductions thereafter.
Project 3. This project requires a standard of 83 dB(A) in 1978, 78
dB(A) in 1984, and no further reductions thereafter.
Project 4. This project requires a standard of 83 dB(A) in 1978, 80
dB(A) in 1982, and 75 dB(A) in 1984 and thereafter. This project is
essentially the same as Program 1 in Tables 9.7 and 9.8 ("Illustrative
Analysis I," above) except that the tune schedule is retarded by about 15
months.
The EPA data can be used to examine the incremental or marginal
costs and benefits of an increasingly severe noise emission standard. This
important issue is largely ignored in Tables 9.7 and 9.8, where only the
marginal costs and benefits of an advance in the abatement time schedule
are considered.
The benefits considered in this analysis include changes in residential
property values that result from reduced noise levels and fuel savings that
will result from abatement equipment or hardware. Property value
benefits reflect: (a) a marginal capitalized damage cost of $50-$ 150 per
dB(A) per household, based on an average $25,000 property value and a
noise discount of 0.2-0.6 percent per dB(A); (b) a growth rate for real
benefits of 3 percent per year for 1975-1987 and 2 percent per year for
1988-2000; and (c) equivalent populations of 31.4 million in urban areas
and 2.6 million in suburban areas for 1978 and beyond. Annual fuel
savings per truck are based on EPA data (1976:6-23) and assume average
fuel prices per gallon of $0.60 for gasoline and $0.45 for diesel fuel in
1975. Total indirect benefits were determined from projected data on new
truck sales and annual mileages, accumulated over four classes of trucks.
Costs include the capital costs incurred in equipping the quieter trucks
and the increased maintenance and operating costs of the additional
equipment. Total capital costs are based on a so-called worst case and on
average values for four classes of trucks (U.S. EPA 1976:6-3, 6-7)
adjusted to 1975 dollars, and projected data on new truck sales adjusted
for higher prices. Maintenance and operating expenses are obtained
directly from Appendix E of the EPA report.
Discounted costs and benefits are summarized in Tables 9.9 and 9.10,
respectively. Various net-benefit estimates were then calculated and these
estimates are shown in Table 9.11. The top-ranked project in all cases is
Project 1, which imposes an 83 dB(A) standard in 1978. Note that net
benefits fall when the 80, 78, or 75 dB(A) standards are imposed. This
-------
200 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 9.9 Total Discounted Costs of Truck Noise Abatement, 1977-2000
(Billions of 1975 Dollars)
Abatement
Program
Capital Costs
(1)
Low1
(2)
High2
(3)
Op. & Mnt.
Costs3
Totals
(4)
Low
(5)
High
At a 4% Real Discount Rate Until the Year 2000
Baseline
1 - 83 dBA
2-83/80
3 - 83/78
4 - 83/80/75
$2.2
3.8
5.2
8.3
$2.4
4.4
5.7
8.8
$1.5
3.2
5.4
8.4
$ 3.7
7.0
10.6
16.7
$ 3.9
7.6
11.1
17.2
At an 8% Real Discount Rate Until the Year 2000
Baseline
1 - 83 dBA
2-83/80
3-83/78
4 - 83/80/75
$1.4
2.4
3.2
5.0
$1.5
2.8
3.5
5.3
$0.9
1.9
3.0
4.7
$2.3
4.3
6.2
9.7
$ 2.4
4.7
6.5
10.0
1 Based on unit cost data in U.S. EPA (1976:6-14).
2Based on unit cost data in U.S. EPA (1976:6-3).
3Derived from Appendix E of U.S. EPA (1976).
suggests that marginal costs exceed marginal benefits for these projects,
although in almost all cases the benefit-cost ratio would exceed 1. For
example, Table 9.9 shows that the 80 dB(A) standard has a marginal cost
of $3.3 billion ($7.0 billion minus $3.7 billion) when the low estimate of
total costs is employed, while Table 9.10 shows a marginal benefit of only
$1.5 billion ($17.3 billion minus $15.8 billion) of 1975 dollars. Further-
more, these results hold over the range of property value effects
considered, $50-$ 150 per dB(A) per household. Thus, unlike the analysis
for aircraft noise abatement, the results here are not particularly sensitive
to measures of intangible, direct benefits based on the property value
method. The increase in property values is, by itself, insufficient to offset
the increase in costs arising from the more stringent standards, given the
small increase in fuel savings at standards below 83 dB(A).
Manufacturers, however, have indicated that it may be possible to
meet the 83 dB(A) standard without the installation of a fan clutch,
depending on testing procedures and the timetable for noise reductions.
In this event, fuel savings at 83 dB(A) might be minimal and a more
stringent noise standard would be required to attain the significant
indirect benefits associated with this hardware. It is possible that most of
-------
Cost-Benefit Analysis: Some Illustrations 201
TABLE 9.10 Total Discounted Direct and Indirect Benefits of Truck Noise
Abatement, 1977-2000 (Billions of 1975 Dollars)
Consumer Benefits
(H\
Abatement
Program
Op. & Mnt. (7)
Savings1 Low2
(8)
High3
Totals
(9)
Low
(10)
High
At a 4% Real Discount Rate Until the Year 2000
Baseline
1 - 83 dBA
2-83/80
3-83/78
4 - 83/80/75
$15.0
16.0
16.4
16.4
$0.8
0.8
1.3
1.4
1.6
$2.5
2.5
3.9
4.1
4.9
$ 0.8
15.8
17.3
17.8
18.0
$ 2.5
17.5
19.9
20.5
21.3
At an 8% Real Discount Rate Until the Year 2000
Baseline
1 - 83 dBA
2 - 83/80
3 - 83/78
4 - 83/80/75
_
$ 9.7
9.8
10.0
10.0
$0.5
0.5
0.8
0.8
1.0
$1.7
1.5
2.3
2.4
2.9
$ 0.5
10.2
10.6
10.8
11.0
$ 1.7
11.2
12.1
12.4
12.9
'Derived from Appendix E of U.S. EPA (1976).
Marginal capitalized noise damages are $50 per dB(A) per household in 1975.
Marginal capitalized noise damages are $ 150 per dB(A) per household in 1975.
the fuel savings benefits would be lost if the 80 dB(A) standard were not
imposed. Additional analysis will have to be conducted before this
important issue can be resolved.
OTHER BENEFIT-COST ANALYSES
The results of the preceding illustrative benefit-cost analyses are not
intended to constitute definitive guides to policy. The steps of cost-benefit
analysis are not a cut-and-dried matter over which there is universal
agreement; much depends on the judgment of the analyst. There is a
scant, but growing, literature of benefit-cost analyses of transportation
noise abatement; nearly all of it deals with aircraft noise. Three studies
that examine the issues covered here are those by the Council on Wage
and Price Stability (COWPS) (1975a), Safeer (1975), and Federal
Aviation Administration (1976). We offer some comments on the relation
of the results of those studies to those reported by Nelson (1975).
The COWPS study compared the benefits and costs of a SAM8D/3D
retrofitting program that involves the installation of sound absorption
materials (SAM) in the engine nacelles of most existing commercial jets
-------
202 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
TABLE 9.11 Net Benefits of Truck Noise Abatement, 1977-2000 (Billions
of 1975 Dollars)
Net Benefit Equals:1
Abatement
Program (6)-(4) (9)-(4) (10)-(4) (10)-(5)
At a 4% Real Discount Rate Until the Year 2000
Baseline
1 - 83 dBA
2-83/80
3-83/78
4-83/80/75
$11.3
9.0
5.8
(0.3)
$ 0.8
12.1
10.3
7.2
1.3
$ 2.5
13.8
12.9
9.9
4.6
$ 2.5
13.6
12.3
9.4
4.1
At an 8% Real Discount Rate Until the Year 2000
Baseline
1 - 83 dBA
2-83/80
3-83/78
4-83/80/75
_
$ 7.4
5.5
3.8
0.3
$ 0.5
7.9
6.3
4.6
1.3
$ 1.7
8.9
7.8
6.2
3.2
$ 1.7
8.8
7.4
5.9
2.9
Cost and benefit totals obtained from Tables 9.9 and 9.10, respectively.
using Pratt and Whitney JT8D and JT3D engines. The benefits were
calculated on the assumption that a two-segment (6°/3°) landing
procedure (TSL) had already been instituted for all commercial aircraft,
thereby reducing noise exposure, particularly outside the NEF-40
contour. The report states:
By extrapolating from the EPA data, we were able to determine that if nothing is
done except to implement the proposed two-segment landing approach, approxi-
mately 5.8 million people will be living within the 30 NEF or higher noise
contours by 1978. Retrofitting the entire non-Part 36 fleet by 1978 would result in
a 2 to 3 NEF dB reduction in noise exposure for these people as compared to the
exposure they would otherwise experience. Assuming an average of three persons
per household, an average 1973 property value per household of $21,300, and
using the consensus estimate of 0.5 percent property value loss per NEF dB, the
marginal benefit of retrofit would be a maximum of $617.6-$832.41 million. Since
EPA estimates the cost in 1973 dollars of retrofitting with quiet nacelles to be $800
million, the benefit-cost ratio is 0.772 (COWPS 1975a: 10-11).
Several questions can be raised. The noise depreciation figure used in
the COWPS study is $105 per NEF, which, in light of the research
reviewed in Chapter 7, falls well toward the lower range of the available
noise damage estimates. Had an intermediate figure of $140 per NEF
been used instead, i.e., a 0.7 percent discount, the retrofitting program
-------
Cost-Benefit Analysis: Some Illustrations 203
would have passed the benefit-cost test used in the COWPS study. (It
would have yielded a benefit figure of $821.4 million, as compared with
the $800 million estimated cost.) Nelson's calculation used the higher
$140 figure and assumed retrofitting of the fleet would be completed by
1978 (as does the COWPS study); in addition, he examined the option of
retrofitting without the adoption of TSL. One would then naturally
expect Nelson's assessment of benefits to exceed those of the COWPS.
Yet they are smaller; they amount to $567.5 million while the COWPS
figure is $617.6 million.
The difference in the results of the calculation lies in the estimated size
of the population exposed to noise levels equal to NEF 30+. Nelson used
5.12 million as the estimate of the number exposed in 1975, whereas the
COWPS used the estimate of 6.2 million for 1972 (Safeer 1975:1). It is
disturbing that three-years difference in the program evaluation date
causes very substantial differences hi the evaluation of benefits. The cost
figures differ by much less; they total $800 million in the COWPS study,
whose evaluation point is 1972, and $611 million in the Nelson study,
with its 1975 evaluation point. The major realm of contention is the
benefit side. It seems clear that the data as well as the admissible class of
hypotheses regarding the shape and magnitude of the benefits stream are
uncertain.
Much hinges on the cost effectiveness of a two-stage landing procedure
in reducing noise. In virtually every cost-effectiveness study, TSL is
judged the first option that ought to be adopted. (See Safeer 1975, and
Muskin and Sorrentino 1976.)
But two-segment landings have, in fact, been ruled out of consideration
by the FAA for reasons of safety. The same issues of cost effectiveness,
however, apply to alternative operational techniques for abating aircraft
noise, such as reduced flap/reduced thrust approaches or local flow
control.
This observation also raises serious questions about the acceptability of
the available evaluations of the SAM retrofitting option, which, as
already noted, has always been assumed to be undertaken after the
adoption of TSL procedures. Clearly, such an assumption attributes
lower benefits to a SAM retrofitting program than it would yield if it were
adopted alone.
CONCLUSION
The reader hoping to find here a clear statement about the desirability of
any particular noise abatement program will be disappointed by this
report. Generally, the calculations are too close to permit an unqualified
-------
204 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
judgment, since comparatively minor changes in assumptions or esti-
mates can make the difference between passing and failing the benefit-
cost test. That conclusion in itself is important for it must undermine the
extreme positions that have been taken on this issue.
Although benefit-cost analysis does not yield an unequivocal result for
every program, the procedure can yield more conclusive evaluations of
some noise abatement options. For example, some programs seem to pass
the benefit-cost tests unequivocally. A program of modified operational
procedures for aircraft arrivals and landings, such as local flow control, is
one. Local flow control refers to procedures by which an airplane starts
its landing at a considerable distance from the airport—as much as 100
miles away. The airplane begins a long, slow continuous descent aand
lands under low power and reduced flap settings rather than approaching
the airport under normal speed, circling, and then landing under higher
thrust and more extensive wing flap use. The costs of this program are
minimal, and there are several benefits. The obvious ones are a reduction
in noise, a reduction in exhaust emissions, and a saving in fuel. Indirectly,
there are savings in time, and the reduction in air traffic and congestion
allow take-off procedures to proceed more smoothly and also save fuel,
time, and emissions. In fact, under most sets of assumptions, the costs of
the program are negative.
It is important to reemphasize the Committee's view that the result of a
cost-benefit analysis is but one piece of evidence to be considered in
arriving at a policy decision. Even where imperfections of data or of
method do not permit a definitive cost-benefit calculation, policy
decisions do have to be made. Failure to determine a policy is itself a
decision, albeit one that is in many cases difficult to justify. We believe
that wherever feasible, a cost-benefit calculation should be carried out
because it can contribute to the rationality of the decision process. But its
results should never be used as a mechanical decision rule, both because
of the imperfections of the calculations and because they do not
encompass all of the relevant considerations. But the decision maker will
still have to confront the issues, and they will have to formulate programs
on the basis of evaluations of the nonquantifiable benefits and the other
pertinent considerations that we have emphasized in earlier chapters.
Finally, we must note that we consider it unfortunate that the debate
over aircraft noise reduction has centered on a narrow class of
technological options, such as those examined in our illustrative
calculations, when it is clear that there is an extensive set of alternative
programs that can provide substantial noise abatement, perhaps more
effectively and more efficiently than those usually considered. Among the
alternative methods available to reduce aircraft noise to an amount
-------
Cost-Benefit Analysis: Some Illustrations 205
equivalent to the FAR-36 are noise emission charges; noise quotas that
may be set at each airport, for each carrier's fleet, or for the entire civil
aircraft fleet; a surcharge on B-707/DC-8 flight tickets; or a court ruling
that makes airport authorities liable for an average one-time compensa-
tion of $150 per NEF for each residence exposed to NEF 30+ by the
year 1978 or 1980.4
Until recently, airlines and airport authorities have operated under
conditions that offered them the environs of airports as free dumping
grounds for the disposal of noise, and they have responded to this
condition with vigor and resourcefulness. But noise disposal is not a free
good—for noise does produce substantial harm. Of course, noise
abatement is also a costly process, and that is why we cannot afford a
noise free environment. Yet society can ill afford to permit noise to grow
unchecked or, at least in some cases, even to continue at its present levels.
REFERENCES
Baxter, W.F. and L.R. Altree (1972) Legal aspects of airport noise. Journal of Law and
Economics 15(1): 1-113.
Coase, R.H. (1960) The problem of social cost. Journal of Law and Economics 5:1-44.
Council on Wage and Price Stability (1975a) Comments of the Council on Wage and Price
Stability Regarding Proposed Aircraft Noise Retrofit Requirements. Filing No.38, dated
4/7/75. Washington, D.C.: Council on Wage and Price Stability.
Council on Wage and Price Stability (1975b) Council Comments on EPA Proposal to
Reduce Noise Level Surrounding Airports. Filing No. 109. Washington, D.C.: Council
on Wage and Price Stability.
Council on Wage and Price Stability (1975c) Comments of the Council on Wage and Price
Stability regarding Proposed Noise Emission Standards of Medium and Heavy Trucks.
Filing No. 42, dated 5/9/75. Washington, D.C.: Council on Wage and Price Stability.
Council on Wage and Price Stability (1975d) Additional Comments of the Council on Wage
and Price Stability Regarding Proposed Noise Emission Standards of Medium and
Heavy Trucks. Filing No. 64, dated 7/10/75. Washington, D.C.: Council on Wage and
Price Stability.
Dygert, P.K. (1973) Estimation of the Cost of Aircraft Noise to Residential Activities. PhD
dissertation. Ann Arbor: Mich. University of Michigan.
4It should be noted that if such penalties are announced well before the date at which they
will become effective, it will give time for airport authorities and airlines to undertake the
appropriate noise reduction measures, whereas if the penalty is instituted without notice, it
will merely constitute a once-and-for-all-time transfer from airlines and airport authorities
to homeowners. Such a transfer provides no incentive for noise reduction and is also
inequitable since many current homeowners whose properties lie within an NEF-30 contour
have already been compensated for the noise damage they suffer through the discount in the
value of their property, when they purchased their residence, compared to property in
quieter areas.
-------
206 BENEFITS AND COSTS OF TRANSPORTATION NOISE ABATEMENT
Federal Aviation Administration (1976) FAR Part 36 Compliance Regulation; Final
Environmental Impact Statement Pursuant to Section 102(2X0), P.L. 91-190. Washing-
ton, D.C.: U.S. Department of Transportation.
Illinois Institute for Environmental Quality (1976) Economic Impact Study of the Proposed
Motor Vehicle (In-Use) Noise Regulations. Report of the Task Force on Noise,
Document No. 76/10. Chicago: Illinois Institute for Environmental Quality.
Merewitz, L. (1975) Analysis of Proposed Regulation of Heavy and Medium Duty Truck
Noise. A Submission Prepared as a Result of Comments of December 27, 1974 on the
EPA Noise Dockes 74-1. Detroit: Motor Vehicle Manufacturers Association.
Muskin, J.B. and JA. Sorrentino, Jr. (1976) Externalities in a Regulated Industry: The
Aircraft Noise Problem. Paper presented at A.E.A., Atlantic City, N.J.
Nelson, J.P. (1975) The Effects of Mobile-Source Air and Noise Pollution on Residential
Property Values, DOT-TST-75-76. Washington, D.C.: U.S. Department of Transporta-
tion; PB-241 570/1GA. Springfield, Va.: National Technical Information Service.
Nelson, J.P. (1976) Aircraft Noise, Residential Property Values and Public Policy. Institute
for Research on Human Resources. University Park, Pa.: Pennsylvania State University.
Nwaneri, V.C. (1970) Equity in cost-benefit analysis; a case study of the third London
airport. Journal of Transport Economics and Policy 4(3):235-254.
Paik, I.K. (1972) Measurement of Environmental Externality in Particular Reference to
Noise. PhD dissertation. Washington, D.C: Georgetown University.
Pearce, D.W. and J. Wise (1972) Equity in cost-benefit analysis; A comment. Journal of
Transport Economics and Policy 6(3):324-325.
Safcer, H.B. (1975) Analysis of the Costs, Effectiveness, and Benefits of Aircraft Noise
Reduction Programs. Paper No. 750595. Warrandale, Pa.: Society of Automotive
Engineers.
U.S. Environmental Protection Agency (1974a) Background Document for Proposed
Medium and Heavy Truck Noise Regulations. Surface Transportation Branch, EPA-
550/9-74-018. Washington, D.C.: U.S. Environmental Protection Agency.
U.S. Environmental Protection Agency (1974b) Background Document for Interstate
Motor Carrier Noise Emission Regulations. Office of Noise Abatement and Control,
EPA-550/9-74-017. Washington, D.C.: U.S. Environmental Protection Agency; PB-242
554/4BE. Springfield, Va.: National Technical Information Service.
U.S. Environmental Protection Agency (1976) Background Document for Medium and
Heavy Truck Noise Emission Regulations. Office of Noise Abatement and Control,
EPA-550/9-76-008. Washington, D.C.: U.S. Environmental Protection Agency.
U.S. Office of Management and Budget (1969) Discount Rates to be Used in Evaluating
Time-Distributed Costs and Benefits. Circular A-94. July 1. Revised March 27, 1972.
Washington, D.C.: U.S. Government Printing Office.
Vaughan, R.J. and L. Huckins (1975) The Economics of Expressway Noise Pollution
Abatement. Report No. P-5475. Santa Monica, Calif.: Rand Corporation.
Walters, A.A. (1975) Noise and Prices. London: Oxford University Press.
------- |