EPA-450/5-81-OO2
   Executive  Summary

      Visibility  Benefits
   Assessment  Guidebook
     Benefits Analysis Program

       Office of Air Quality
      Planning and Standards

U.S. Environmental Protection Agency

      Research Triangle Park,
      North Carolina   2771]
          August 1981

-------
              EPA-450/5-81-002
            EXECUTIVE SUMMARY
   VISIBILITY BENEFITS ASSESSMENT GUIDEBOOK
                    by
          Robert D.  Rowe, Ph.D.
          Lauraine G. Chestnut
                 Abt/west
            1410 Grant Street
               Suite C-207
          Denver, Colorado  80203
             (303) 830-0181
              August,  1981
      EPA Contract Number 68-02-3528
           Project  Officer
              Tom Walton
        Economic Analysis Branch
Office  of Air Quality Planning and Standards
    U.S. Environmental Protection Agency
Research Triangle Park, North Carolina  27711
             Abtlwest

-------
              VISIBILITY BENEFITS ASSESSMENT GUIDEBOOK




                          EXECUTIVE SUMMARY








1.0  INTRODUCTION






        Good visibility is something that people value in recreational,




residential, and other situations.  When vacationers make the drive  to the




Grand Canyon, they expect to have a good view of the canyon.  Few people




would be willing to spend the necessary time and money to go there if they




knew that the canyon vistas would be entirely obscured by haze.  For those




who find the view partially obscured or the clarity of the view reduced,




the value of the experience is diminished, although they may still choose




to make the trip.  Similarly, people prefer to have clear views from their




homes.  Whether households have open views from their home that may be




obscured by haze or whether only the color and brightness of the sky




is noticeable, reduced visibility affects people's moods and their enjoy-




ment of their residences.




        There are many reasons why people value and protect good visi-




bility.  Economists divide these reasons into three categories.  The




first is that people value good visibility when they are using a site such




as a residence or a park—this is called activity or user value.  If




people know where they will live and vacation in the future, they will want




to protect visibility in anticipation of future as well as present activity




value.  However, people usually experience some uncertainty planning




future activities and may want to preserve a  certain level of visibility in




order to maintain the option of enjoying that site in the future—this




is called option value.  This may be a particularly important value when




visibility deterioration is difficult to reverse.  A third reason that

-------
people may want to preserve visual quality is simply to know that  it

exists, though the individual does not ever intend to visit the site—this

is called existence value.  This value may be especially important with

regard to unique natural wonders, such as the Grand Canyon, and may be tied

to the philanthropic goal of preservation so that future generations will

be able to enjoy the site.  Changes in air quality that affect any of

these values will be associated with changes in an individual's well-being.

        Visibility is often reduced due to natural causes,  such as fog,

rain, and wind blown particles, but human activities have long been re-

cognized as an important contribution to diminished visual aesthetics.  For

example, in response to recent EPA inquiries,  Federal Land Managers have

indicated that roughly one third of the mandatory Class I Federal areas*

experience, at least at times, undesirable visibility conditions due to

manmade pollution.

        Reflecting the importance of protecting visual aesthetics from

manmade pollution in our scenic areas, the 1977 Clean Air Act Amendments

established a national goal of remedying existing and preventing future

manmade visibility impairment in mandatory Class I Federal areas.   They

mandated the EPA to establish regulations to ensure that reasonable pro-

gress toward this goal would be made.  The EPA report to Congress,  Protecting

Visibility, notes that the determination of what is reasonable progress

toward the national goal requires that the value as well as the costs  of

protecting or improving visibility be considered.
^International parks,  national wilderness  areas,  national memorial parks
 exceeding 5,000 acres,  and national parks exceeding 6,000 acres.

-------
        Visibility regulations based upon the 1977 Clean Air Act Amendments




were promulgated on December 2, 1980.  They indicate that some visibility




impairment will be tolerated only if the costs of reducing or preventing




impairment are considered too high.  However, the size of these costs




is only relevant when compared to the benefits of the improved visibility



to be gained from control, or lost from the lack of control.  Therefore,  in




order to decide whether or not to tolerate existing or additional visi-



bility impacts, some Judgment regarding the size of the benefits must be




made.  Thus, techniques to determine the value of visibility would be an




important aid in determining the appropriate level and type of pollution




controls to be required.




        Although the potential usefulness of estimates of the benefits of



visibility aesthetics has long been recognized,  aesthetic impacts of



pollution have tended to be considered nonquantifiable.  This is no longer



the case.  There are at present several defensible economic methodologies




being applied and developed to estimate the value of changes in visual



quality that may result from changes in air pollution control.   Applications




of these techniques—referred to as visibility benefit analyses—can be




used to estimate the dollar value of changes in visibility.   This can




add to our knowledge of the benefits of pollution control and improve the




benefit-cost analysis of controls.  Current research Indicates  that in most




cases the benefits of protecting visibility are substantial.  This suggests




that the benefits of visibility should be included in the benefit-cost




analysis in order to help determine the appropriate allocation of resources.




        The Visibility Benefits Assessment Guidebook aims to introduce




visibility benefit analysis techniques that can be used to estimate bene-



fits, in dollars, for changes in visibility aesthetics resulting from

-------
alternative levels of air pollution control.  Policy makers can then

compare these benefit measures to the costs of controls.  The Guidebook

saves researchers or reviewers of visibility benefit analyses considerable

time and effort in becoming well-versed by providing a comprehensive review

of the theory, terms, and best practice techniques.  The Guidebook discusses

application of these techniques to evaluating existing or potential

visibility impacts from single sources or regional haze in recreational,

rural, or urban settings.  Documentation of the assumptions, strengths, and

limitations of each method provides inexperienced users with a technical

basis for interpreting the results of visibility benefit studies.  The

results of ten case studies are presented to demonstrate the magnitude of

pollution control benefits.  In many cases, these benefit estimates may be

used to establish initial estimates of damages for other sites.  In short

the guidebook translates technical work for a broad audience,  provides

information about experts in the area and data sources, and offers refer-

ences on specific topics.

        The guidebook is organized as follows:

        Chapter 2.  The conceptual economic framework and definitions of
        monetary benefit measures for visibility Improvements  derived from
        consumer utility theory and demand analysis.

        Chapter 3.  An overview of required inputs from other  disciplines
        to aid the economic analyst in communicating with physical scien-
        tists, psychologists,  and sociologists in performing a complete
        benefit analysis.

        Chapter 4.  The applications, theoretical foundations  and strengths
        and weaknesses of economic techniques currently used to measure
        benefits of visibility and overall air quality changes.   These
        techniques include bidding methods, property value studies,  travel
        cost studies, the ranked attribute approach and others.

        Chapter 5.  A summary of the start to finish research  tasks  for a
        complete visibility benefit analysis with suggested step-by-step
        decision and application guidelines.

-------
        Chapter 6.  Five case studies of actual applications for both  of
        the two predominant techniques—bidding methods and the hedonic
        property value approach—and a comparison of their procedures  and
        results.

        A glossary, annotated bibliography, and an appendix with sample

picture sets and questionnaires for the bidding method are also included.


2.0  CONCEPTS AND MEASURES OF ECONOMIC BENEFITS


        Performing a benefit analysis of changes in visibility aesthetics

 first requires a careful definition of what the benefits are and how

monetary measures of these benefits are defined.  This is especially

important for visibility aesthetics, because the concepts of visibility as

an economic good and of monetary benefits from visibility improvements may

not be obvious.  Therefore, before launching into a description of the

actual benefit estimation techniques, it will be helpful to present the

general definition of benefits that economists use and that the estimation

techniques attempt to capture.

        Consumer demand theory provides a framework for analyzing changes

in an individual's well-being caused by changes in visual quality and for

defining measures of benefits of these changes.  The theory asserts that

individuals derive well-being, or what economists call utility, from the

consumption of goods and services.  Thus,  any change in the level of

consumption that affects someone's utility, such as a change in visibility

at a favorite park, has a value.  This change in utility may be either a

benefit or a damage, depending upon whether the individual's well-being is

enhanced or diminished.  The appropriate monetary measure of benefits for

an improvement in visibility indicates how much the increase in utility is

-------
worth.  The benefit measure therefore equals the change in income  that



would yield the same change in utility as that caused by the improvement  in




visibility.



        Market goods have easily recognized value because when individuals




purchase them they forego purchasing other goods and services that could



also have increased their well-being.  The market price an individual pays




for a good therefore represents a minimum monetary measure of value to the




individual.  Visual quality also has value even though it is not purchased



or sold in a market.  People change their recreation patterns, move their




residences, or simply alter their moods due to changes in the existing



level of visual quality.  Thus, analyzing how individuals react to visual



quality changes may reveal the value they place on these changes in visual




quality.  In other words, a visibility benefit analysis attempts to deter-



mine what individuals would be willing to pay for a change in visual




quality if it were possible to purchase it.




        In order to determine the value or benefits of changes in visi-



bility, economists usually try to estimate or derive a demand curve for




visual quality.  Three related monetary measures of value are used in




visibility benefit analysis, and can be derived from demand curves.  They




are willingness to pay (WTP), expenditures, and consumer surplus.  To



illustrate these measures, a demand curve for an ordinary market good is




given in Figure A and a demand curve for visibility is given in Figure B.




        WTP is the maximum amount an individual would be willing to pay



in order to obtain a given amount of a good or to experience a given level



of visibility.  The maximum WTP for an additional unit is the corresponding



point on the demand curve—B for quantity F and C for quantity G in both

-------
Figures A and B.  Expenditures represent  the  actual amount a consumer


spends, which in a situation of free choice would  never be more than his


maximum WTP.  Expenditures are represented by the  price of each unit times


the quantity of units—the area BDFO for  quantity  F and a constant price B


in Figure A.  Although consumers do spend money in order to go to a park


that has clear visibility or to live in a neighborhood  that has clean air,


consumers do not directly purchase visibility.  Therefore there are no


expenditures for visibility in Figure B.


        In both of the illustrated cases, the consumer  is paying less than


his maximum WTP for all units.  This surplus value,  or  the difference be-


tween the maximum WTP and expenditures, is called  consumer surplus.  Con-


sumer surplus equals the difference between expenditures,  if any,  and the


area under the demand curve up to the quantity of  the good consumed or the


quantity of visibility experienced—the area ADB at  price B and quantity F


in Figure A and the area ADFO at quantity F in Figure B.
                 Rqure A
              Figure B
                           QUANTITY OF
                           MARKET GOOD.
                          QUANTITY OF
                          VlSfflUJTY
     Demand for Market Good
Demand for Visibility

-------
        For changes in prices or quantities of goods, the changes in



consumer surplus are also shown.  If the price of a market good falls from



B to C, the increase in consumer surplus is shown by the area BDEC in




Figure A.  If the quantity of visibility increases from F to G, the



increase in consumer surplus is the area DEGF in Figure B.




        The change in consumer surplus that results from a change in



visibility is the measure of the change in an individual's well-being




(benefits or damages) caused by the visibility change.  When the level




of visual quality changes it affects the individual's willingness to pay



for visual quality or for the associated activity or good that has been



affected.  If visibility is improved,  for example, the individual would be




willing to pay more for either the higher visibility directly or for the



activity or good that is enhanced by the improvement in visibility.   Th£




change in utility caused by the improvement in visual quality is reflected



by the change in WTP.  However, in some circumstances, the individual will




have to pay part or all of his increased WTP in order to obtain the




improvement.  Such an increase in expenditures would leave less income




left to purchase other goods and services; thus,  to the extent that  expend-




itures to obtain or enjoy visibility are increased,  the increase in  utility




from having improved visual quality is partially offset by decreased



utility from other goods and services.  Therefore, it is only the change in




consumer surplus that represents the net benefit to the individual re-



sulting from the change in visibility.




        In general, the change in consumer surplus tells us how much of



an increase in money income would have caused the same net increase  in



utility as that caused by the visibility improvement.  Thus,  the aim of  any
                                   8

-------
visibility benefit analysis is to estimate the changes in consumer  surplus




that would occur in the particular circumstances under study.  By aggre-




gating the changes in consumer surplus for the individuals across all




affected individuals, benefit measures reflecting the total change  in




utility can be obtained for use in benefit-cost analysis.




        There are four refined measures of consumer surplus that each have




slightly different interpretations.  Each of these refined measures is a




theoretically correct measure of benefits, but under different circum-




stances.  In practice they are often quite close in size and therefore the




distinctions are sometimes ignored.  Sometimes more than one of these




measures may be applicable in a given study situation and therefore several




are estimated.  These refined measures are defined and discussed in the




Guidebook.  The important point is that they are measures of changes in




well-being and as such are used as estimates of economic benefits that




result from changes in visual quality.






3.0  TECHNIQUES FOR ESTIMATING BENEFITS






        The techniques used to obtain measures of the value of visual




aesthetics rely upon one of two categories of approaches:  1) asking people




to establish values directly through the use of hypothetical, or what is




known as contingent market, situations posed to survey respondents;  and 2)




using actual market data regarding the relationship between private  markets




and visual quality, combined with special analytical methods, to infer




implicit prices and the demand for visual quality.




        The first group, contingent market approaches, elicits the value of




visual quality through surveys of how respondents think they would behave




if a proposed visibility change were to occur under various  hypothetical

-------
circumstances.  The usual version of this approach is the bidding method.




By this technique respondents are asked to estimate their willingness  to




pay (in terms of entrance fees, utility bills, or whatever payment mecha-




nism is applicable to the situation) to achieve or prevent a given change




in visual quality that is described or pictured for them.  Another contin-




gent market approach is a variation on the travel cost method: respondents



are asked how they would change their travel plans if they knew that




visibility at a particular recreation area were to change.



        The second group, actual market approaches, uses market data in




cases where the selection of a market good may vary with visual quality.



Residential property value studies are a common application of this




approach.  Such studies attempt to derive the demand for visual quality by



determining the effect of different levels of visual quality on the indi-



vidual's willingness to pay for a residence.




        Following are brief descriptions of each of the most frequently



applied, or most promising, contingent market and actual market techniques



for estimating benefits for changes in visibility aesthetics.  Following




these descriptions is a discussion of the necessary considerations in




choosing an estimation technique.






        Contingent Market Techniques






        In bidding methods, respondents are given information on current




and proposed or potential future levels of visual quality at a particular



site.  They are also given hypothetical markets that describe how payments



for changes in visual quality are to be made or received by the respondents.




They are then asked to bid their maximum willingness to pay (WTP) or



minimum willingness to accept compensation (WTA) to obtain or prevent the



change.




                                   10

-------
        If a degradation of visibility is predicted, respondents might be




asked their WTP to prevent this change or their WTA to allow the change.




Similarly, for a proposed visibility improvement, respondents are asked




their WTP to fund the change or their WTA to forego the change.  For




example, respondents might be presented a hypothetical situation—"Local



air quality may be degraded by energy development from the current level,



depicted in picture A, to a lower level, depicted in picture B" (illus-




trated with photographs)—and a market mechanism—"If a surcharge were to




be added to your utility bill that would go to fund equipment that would




prevent visibility degradation, what is the maximum you are willing to pay



each month on your utility bill to prevent the degradation in visibility



from level A to level B?"  Answers are usually elicited by asking the



respondent if he would be willing to pay or accept a given amount.  The




amount is adjusted until the highest amount the respondent would be willing



to pay or the lowest amount he would be willing to accept is determined.



This amount can be used as a direct estimate of the change in the indi-




vidual's consumer surplus that will result from the projected change in



visual quality.



        Aggregate benefits are determined by projecting the average bids



for the respondents across all individuals that would be affected by the



visual quality change.  The results must be adjusted for differences




between the respondents and the total impacted population in income and




other characteristics.  If benefits will continue over a substantial



period of time, future benefits are usually discounted using standard



procedures.




        It is imperative that this approach be carefully designed and



monitored, because the valuations given by respondents may vary with small
                                   11

-------
nuances in its application.  The values that respondents report may



well reflect their "true" values, estimated with a great deal of uncer-




tainty, but the respondents are subject to the influences inherent  in  the




design of the survey.  These influences may decrease the accuracy of the



responses and may cause biases in the valuation process.  Most important,




the contingent market mechanism and the proposed air quality changes must




seem realistic and credible to respondents.  Otherwise, the hypothetical



questions will not elicit accurate answers from respondents (hypothetical




bias).  Tests for the occurrence of bidding method problems and actions to




minimize their impacts on the accuracy of the valuation process have been



and are being refined.  The evidence to date is that if hypothetical




problems can be overcome, the other potential biases may also be eliminated.




        Bidding methods, and other contingent market approaches, also



require that the change in visual quality to be valued be accurately



communicated to the respondent.  Pictorial representations of alternative



visual quality scenarios using photographs, slides, and motion pictures,



coupled with verbal descriptions and maps, have become the standard




mechanism for displaying alternative scenarios of visual quality.   To be



successful, the scenario development must be clearly understood, realistic,




uniform for all respondents and, hopefully, leave the respondent with a




feeling that the situation and his responses are not only credible but




important.  In addition, it must be possible to link the visibility scen-




arios as described to the respondents to the costs of attaining or avoiding



the associated levels of pollution.  This requires knowing the nature and




the sources of the pollutants that are affecting the study area as well as



knowing how people perceive the effects of the pollutants.   A competent



application therefore necessitates interaction with other physical and



social scientists.




                                   12

-------
        The contingent travel cost approaches to determining the value  of




visibility at a recreation site follow the travel cost literature by




recognizing that to use the services of a recreation site, users not only




incur the expenses of entry fees and equipment, but must get to the site.




The cost, or price, to an individual of using the services at a recreation



site will vary according to the travel time and expenses incurred getting



there.  The user's willingness to incur these expenses will be affected by




changes in the characteristics of the site, such as visibility.  The travel



cost approaches attempt to determine the willingness to pay for the quality




of the recreation experience by examining the changes in travel costs



various individuals are willing to incur in order to use various sites as



air quality changes.



        Travel cost approaches have been applied extensively to estimate




the demand for recreation sites using readily available data on visitation




rates, travel distances, and visitor characteristics such as income.




However, estimating the demand for a particular characteristic of recrea-




tion sites, including visibility conditions,  requires data that are less



readily available, and generally requires accounting for all other important




characteristics of the sites as well.




        Such applications of the travel cost  approach are still being



developed.  The contingent travel cost approach is a promising variation.



Sufficiently detailed data are obtained by surveying recreation site users



as to their expected recreation travel behavior given alternative hypo-




thetical scenarios of environmental conditions both at the site and at




competing sites.  A utility maximizing individual is assumed to recreate at



the nearest outdoor facility if all areas are equal and to recreate at the



next nearest area only if there is significant deterioration at the closest
                                   13

-------
site.  By using answers to detailed questions on expected activity rates  at


selected locations, given various site attributes, coupled with the costs of


substituting activities across locations the value of changes in the


levels of visibility can be inferred.


        Other contingent market approaches that are less well developed


for estimating benefits for changes in visual quality than those so far


described, but that may prove to be useful, include the household produc-


tion function approach and ranked attributes.  The household production


function approach is somewhat similar to the travel cost approaches in that


changes in activities and expenditures that are influenced by air quality


are used to generate WTP curves and damage functions for changes in air


quality.  In the ranked attributes approaches, survey respondents are asked


to rank order sites that have different levels of visual quality, prices and


other site attributes.  The respondent does not put a dollar value on


visual quality but values are revealed indirectly by the rankings or


probabilities that a site will be chosen for recreation, given the dif-


ferent entrance fees, travel expenditures, and other costs associated with


using each site.



        Actual Market Approaches



        Property value studies are based on the supposition that if air


quality varies across the area and if people are willing to pay more for a


residence with better air quality, the amount they are willing to pay will


be revealed by the price differences between properties that are similar in


all respects except air quality.  Most such studies have used the hedonlc

              -*
price technique to estimate an implicit price for changes in air quality


that are associated with, and influence, the value of residential property.
                                   14

-------
        Hedonic price theory suggests that the household values a  residen-




tial property because it embodies particular attributes that the household




desires, such as places to sleep and eat, a pleasant environment,  and




access to employment and recreation.  The amounts of each attribute




determine the price that consumers are willing to pay for the residence.



If these attributes can be defined and quantified by consistent units of



measure, then by observing the prices of homes with different levels of




each attribute, the effect of each attribute on the prices of homes can be



determined.



        This relationship between the price of a home and its attributes is




called a hedonic, or implicit, price function.  Its estimation is the first



step of the hedonic price technique.  The hedonic price function enables



the analyst to statistically hold constant all attributes but one in order



to see how changes in the level of that attribute alone affect the prices



of homes.  The hedonic price function can be used to determine how dif-



ferences in visibility would affect prices of residential property if each



residence were identical in every other respect.




        In this way, a marginal implicit price for each attribute can be




derived from the hedonic price function.  The marginal implicit price of an



attribute is the additional amount that must be paid in order to obtain a




home with an additional unit of the attribute.  Deriving a marginal




implicit price for changes in visual quality is a first step toward esti-



mating the household's willingness to pay for changes in visual quality.



        The next step requires observation of how the marginal implicit



price for visual quality changes as the level of visual quality changes



either across locations or across time.  Thus, a WTP function is estimated



which relates the marginal implicit price of visual quality to the level of
                                   15

-------
visual quality at which the household chooses to locate.  It must  also




account for the effects of household income, household size, and other



household characteristics and property attributes that influence the



household's willingness to pay for visual quality.




        Once estimated, the WTF function for visual quality can be used to



derive total benefits for all households that are expected to be affected.



Total benefits are the amount that all households are willing to pay for




the improvement.  If separate WTP functions are estimated for different




household groups (separated according to income, location, or some other




characteristic), the benefits for each group must be summed to obtain total



benefits and some decision must be made concerning how to weight each



group's WTP.  Standard discounting procedures would be applied to aggregate




benefits across time.




        To conduct a property value study, fairly detailed residential




property and household data must be available.  In particular,  air quality




measures that reflect the aspects of air quality that people value and that



are specific to each neighborhood must be available.  This typically



limits the locations and types of impacts that property value studies can



consider.  To date, property value studies have used average concentrations




of specific pollutants or pollution indexes, such as the Pollution Standards



Index (PSI), which are available in most urban areas, as proxies for air




quality in each neighborhood.




        Another actual market approach that has been applied with some



success is the wage and salary approach.  In this approach,  willingness  to



pay for air quality is derived from differences in wages and salaries



between location^ with different air quality.  In order to isolate the



effects of air quality on wages and salaries, other influences  must also be




accounted for.



                                   16

-------
        Choice of Benefit Estimation Technique






        Before a complete benefit estimation study using any of these




methods is undertaken, it would be desirable to determine if results from




previous studies are sufficient to answer the questions currently under




study.  Two levels of accuracy are possible in transferring results.



It may be possible to transfer actual or systematically adjusted results




from past studies to achieve reasonably precise benefit estimates for the



current study—a level of accuracy we call strong transferability.



This requires that the circumstances of the two studies be identical or




different in ways for which the effect upon benefit measures is well known.



If this is not the case, it may still be possible to use previous results



to set upper or lower bounds on benefit estimates for the current study as




a starting point for decision making—a level of accuracy we call weak




transferability.  This requires that the direction, but not the size, of



the differences in the situations be known.  Theoretically, conditions are




established which should ensure valid strong and weak transferability of



results.  Whether these conditions are met is an empirical question that



must be answered for each study area under consideration.  At present, only




a few visibility benefit analysis studies have been completed from which to



transfer results.



        Which benefit analysis technique is appropriate for the study



question at hand must be decided on a case-by-case basis.  It will depend



on the location and nature of the visibility impacts, data availability,



and the budget and time available for the study.  In general, contingent




market approaches are much more flexible in terms of locations and types of




impacts that can be considered, because survey questions can be designed to



suit the particular circumstances under study.  Actual market approaches






                                   17

-------
must rely on data that are already available.  This implies  that  for many




recreation and rural study areas, there may be no choice but to use  a




contingent market approach.  Actual market approaches tend to be  cheaper




and quicker, however, because they do not necessitate any survey  work,




which can be very expensive and time consuming.  No matter which  approach



is used, the researcher needs to be well aware of its strengths and weak-




nesses in order to conduct a credible analysis and in order to understand



the meaning and reliability of the results.




        The most important strengths of the bidding method are that it is




well grounded in economic theory and flexible in application.  The approach



can be applied to obtain separate activity, option, and existence values,



both on-site and at alternative locations.  Any type and location of visual



quality impact can be considered as long as the impact can be conveyed



to respondents in a manner they can understand along with a credible




contingent market mechanism.  Similarly, the contingent travel cost




approach is quite flexible, although it cannot be used to estimate option




or existence values.




         Contingent market approaches, however, presuppose that a particular



change in visual quality can be described to respondents, usually with




photographs and verbal descriptions, in a way that corresponds to what



their perceptions of the actual experience would be.  For example, it is




assumed that a vacationer can look at a photograph of the Grand Canyon



obscured by pollution and accurately predict how experiencing it in reality



would lessen his enjoyment of the canyon and how it might cause him to




alter his plans.  These types of approaches also assume that individuals



are capable of and willing to predict accurately their response behavior to



a hypothetical situation that they may or may not have ever actually
                                   18

-------
experienced.  In this respect, contingent travel cost methods may have an




advantage over bidding methods because they ask questions about decisions




that recreationists do routinely make.




        An important strength of the property value approach and other




actual market approaches is that they are based on households' or indivi-




duals' actual behavior.  This is a significant advantage over bidding




methods that ask people how much they would spend,  but do not require  them




to part with their money.  If the visual quality attribute is successfully




separated from the other property attributes, the hedonic price technique




can be used to determine how much households have actually spent to obtain




various levels of visual quality.




        Actual market approaches, however, presume that individuals




respond in a predictable manner to visual quality conditions that they




encounter in places where they live, work* and recreate.  For example, it




is hypothesized that individuals prefer to live in neighborhoods where the




view is not impaired by air pollution.  These approaches then assume that




the intensity of such preferences is revealed by the individual's behavior




and his willingness to pay for associated market goods.




        A summary guide to method selection is presented in Table 1.  In




addition to these criteria, it is also important to consider that both




contingent market and actual market approaches require several other




analytical and operational assumptions upon which their validity depends.




The analytical assumptions are similar to those used frequently in economic




analysis.  The operational assumptions concern the  conditions necessary to




ensure that the procedure is capturing the desired  value.  These assump-




tions need to be considered in any application of these valuation procedures.




If they are not valid in theory or not met in practice, the validity of the




results generated by the techniques are weakened.




                                   19

-------
                                                                            T«bl« 1

                                                 Selection of Visibility Benefit Estimation Method — A Summary
                Selection
                Criteria
                                     Transfer Resulta
                                    of  Previous  Studies
                                             Bidding Methods and
                                           Other Survey Approaches
                                      Property Values and
                                    Other Market  Approaches
NJ
O
           Major  strength
           Major weakness
            Coat  and  timing
            Locations where  impacts can be
            addressed

            Types  of values
Separability of Impacts
(health vs. aethetics, etc.)

Types of Impacts that can be
addressed
                                    Lowest  cost, quickest results
                                    Lack of prior studies and consistent
                                    empirical analysis
                                    Lowest  cost,  quickest  results
At present;  weak transferablllty for
urban areas and Southwest parks

Constrained by prior studies, primarily
activity value*

Constrained by prior studies, primarily
not separated

Constrained by prior, studies
                                           Flexibility In application to
                                           location,  type of Impact and
                                           values

                                           Reliability of data from hypo-
                                           thetical questions
.Highest costs,  often takes > 1
 year,  seasonal  problems

 All locations,  Class I, urban,
 rural, other

 Activity,  option,  existence
                                                                                          Benefits are separable
                                                                                          Haze or plume, and ratea and
                                                                                          levels of occurence; uniform
                                                                                          or variable
                                    Reliability of data re-
                                    flecting actual behavior
Availability of data for
nonurban locations and re-
quired variations In Impacts

Medium costs, usually < 1
year

Urban, or where sufficient
data exists

Activity, sone option
                                    Separability not yet
                                    addressed

                                    Haze or plume, and rates
                                    end levels of occurence;
                                    must vary across study area
            Evaluation  of current vs.
            future  impacts

            Atmospheric data  requirements
            Economic  and  population
            data
            Reliability of  data and  results
                                    Both,  depending upon prior  studies
                                    Accurate data required to match current
                                    situation to prior studies
                                    Affected population characteristics
                                    Depends upon accuracy of prior studies
                                    and match to current Impacts
                                           May evaluate both current or
                                           future Impacts

                                           Accurate data required for al-
                                           ternative "typical situations"
                                           across area

                                           Affected population characteris-
                                           tics must be cross-checked with
                                           sample characteristics

                                           Some uncertainty due to hypo-
                                           thetical approach
                                    All alternative Impact levels
                                    should occur In study area

                                    Accurate data required .on all
                                    pollution levels at a neigh-
                                    borhood or census tract level

                                    Property values, property use
                                    and characteristics plus pop-
                                    ulation characteristics

                                    Reliable market data, some
                                    uncertainty regarding estima-
                                    tion techniques
             This table is to be  used only  as  a  summary of more detailed comparisons contained in Chapter* 4 and 5 of  this guidebook.  The  same
             criterion can be applied to travel  costs, wage and salary hedonlc techniques, etc., as  they are developed for application to
             visibility benefit analysis.

-------
4.0  OVERVIEW OF THE VISIBILITY BENEFIT ASSESSMENT


        Visibility benefit analysis requires input from many fields in

addition to economics in order to provide useful input for air pollution

control policies.  In general, it requires knowing or determining the

following:

        •  Source of the pollution and its type, intensity, and frequency.

        •  Spatial dispersion of pollutants and type of impact (plume,
           haze).

        •  Scenic content of the affected view.

        •  Atmospheric conditions and transmission of visual images, which
           is affected by lighting, cloud cover, etc.

        •  Human perception of impacts.

        •  Psychological effects of perceived impacts.

        •  Economic valuations of impacts in terms of willingness to alter
           time or dollar expenditures.

        The various research tasks necessary in a typical visibility

benefit assessment can be organized into a six-step process.  This process

should not be considered a formal procedure, but merely a guideline to help

the researcher understand the step-by-step process of performing a benefit

analysis.  There is at present no formal legislative requirement to

perform visibility benefit assessments in any specific situations or in any

particular format.  In addition, any such guidelines should remain flexible

enough to incorporate what is learned through more research and repeated

application.  The six basic steps are as follows:

        Step 1;  Problem Formulation
1.  Describe potentially impacted area.

2.  Identify sources of potential visibility changes.

3.  Define potential pollution changes in terms of specific air quality
    measures.
                                   21

-------
4.  Identify legal foundations and requirements relevant to pollution type,
    source and impact area.

5.  Determine public perceptions of potential emission rates and air
    quality levels.

6.  Determine if further analysis is required, i.e., screen out projects
    whose "worst case" impacts are obviously insignificant in terms of
    physical variables and human perceptions.

        Step 2;   Scenario Development
1.  Determine existing air quality conditions in terms of appropriate air
    quality measures.

2.  Determine the most probable future conditions without the proposed
    change, using selected air quality measures.

3.  Define changes in visibility conditions expected to result from
    proposed action or policy, using selected measures.

4.  Conduct initial physical impact analysis.

        Step 3;   Selecting the Economic Benefit Estimation Technique.
Discern whether the results from previous studies can be transferred
to the study at hand.  If possible, using results from previous studies
will save considerable time and money. If the results from other studies
cannot be accurately or defensibly transferred, the benefit estimation
technique(s) must be selected based upon the following factors:

1.  The location of the impact—urban/rural/national park/other—and
    whether the impact is regional or site specific.

2.  The nature of the impact.  Take into consideration the timing of the
    impact; type of Impact such as haze or plume; frequency, intensity and
    duration of the Impact; and whether it is an aesthetic, health and/or
    materials impact.

3.  Type of values (activity, option, existence) for which benefits
    are to be estimated.

4.  Data requirements and data availability and reliability for each
    technique.

5.  Necessary assumptions of each benefit estimation technique.

6.  Costs and timing requirements for the study effort.

        Step 4;  Application of Estimation Technique.  The bidding method
and property value approach are the two most commonly used to date.  If
applying bidding methods, the following steps are necessary:

I.  Determine the benefit measures (WTP, WTA) to be estimated.
                                   22

-------
2.  Select alternative scenarios to be evaluated and select methodology for
    depicting visibility conditions, usually some type of pictorial
    representation.

3.  Determine survey procedures—who will be interviewed and how and when
    they will be interviewed—and design survey instrument to obtain
    estimates of benefits with a minimum of respondents' time.

4.  Pretest the survey and implement the survey.

5.  Conduct statistical analysis and compute benefit measure estimates.

If applying the hedonic property value approach, the following steps
are necessary:

1.  Select final data—the most important of which are property and
    air pollution data—and make sure all neighborhood amenities are
    incorporated.  Data may be available from U.S. Census, state and local
    government, or private organizations.

2.  Select the study sample and organize data by adjusting data from
    different sources for compatibility.

3.  Perform statistical analysis by estimating the hedonic price function,
    willingness to pay, and benefits or damages of the changes in air
    quality.

        Step 5;  Aggregation of Benefits Across All Affected Populations
Project benefit estimates for the study sample across the entire population
expected to be affected.  Adjustments should be made for differences in
characteristics such as incomes.  Benefits should be aggregated across the
expected time frame of the impact using standard discounting procedures.

        Step 6;  Benefit-Cost Analysis
1.  Identify related emission contrcl costs.

2.  Identify other environmental and non-environmental costs and benefits.


5.0  CASE STUDIES OF VISIBILITY BENEFIT ANALYSIS


        The results of the dozen or more visibility benefit analyses

conducted so far, including both contingent and actual market approaches,

indicate overwhelmingly that the value people place on visibility aesthe-
                                   23

-------
tics is substantial.  Protecting visibility aesthetics is therefore  an




important benefit of air pollution controls that should be considered  in




policy decisions.  The results of the studies reviewed in the Guidebook



are summarized below.  Although the circumstances of each study differ,




the results are broadly consistent, providing evidence that estimation



procedures are valid.




        The five bidding method studies reviewed in the Guidebook are  the




Four Corners study, by Randall et al. (1974); the Lake Powell study, by




Brookshire et al. (1976); the Farmington study, by Rowe et al. (1980 a,b);




the South Coast Air Basin study, by Brookshire et al. (1979;  1980); and



the Grand Canyon/Southwest Parks study, by EPA (1981).  Although the




bidding method has been applied to several dozen other environmental and



resource problems in recent years, these five are the best known applica-



tions to visibility benefit estimation.  With the exception of the South



Coast Air Basin study, all of these studies concern visual range and



discoloration impacts upon recreation areas and local communities induced




by rapid growth in the coal-fired electric-generating industry in the



southwestern United States.




        Each of these studies examines at least one alternative level of



visual quality that occurs or is expected to occur in the study area.  All



of them have used pictorial presentations to describe visual  quality




variations to survey respondents.  The scenario developments  have evolved



from general scenarios in the earlier studies, which defined  air quality




variations simply as "worst" to "best" and lumped several types of impacts



together, to carefully defined scenarios in more recent studies that




attempt to separate visibility aesthetics, power-plant siting, and health-



related impacts and to distinguish between user values and existence
                                   24

-------
values.  Only the Farmlngton study and the Grand Canyon/Southwest  Parks



study, however, have linked the visual quality scenarios to emissions,




allowing the benefits of Improved (or prevention of deterioration  In)




visual quality to be linked to the costs of attaining or preserving clean




air.




        All of the studies obtained a sufficient number of Interviews




for analysis, but not for all subgroups of respondents.  In all cases




the percentages of usable responses were high, but the percentages that




refused to participate and the exact survey procedures were not generally




reported.  Self-selection bias may have occurred because respondents are



likely to be different from those who refuse to participate, especially in



terms of their concern about air quality.  As studies have become more




careful in defining impact scenarios and in isolating values for different



types of impacts, questionnaire length has increased, creating the poten-



tial for response fatigue problems.




        When bias tests were formally designed and implemented, bias



problems were found in most cases.  This has not always been the case for




bidding method applications to other problems.  In some cases the biases



were not large, but the results indicate that respondents do have trouble




putting quantitative values upon visibility and that they are influenced by



the survey and questionnaire design.




        In all of the studies, respondents' incomes have been found to be



statistically correlated with respondents' bids,  but the effect of incomes



on bids appears to be surprisingly small.  For residents, income elas-




ticities (percentage change in bid divided by percentage change in income)



are all less than one, and for recreationists they are very close to zero.



This means that respondents with larger incomes are willing to pay somewhat,



but not much, more than those with lower incomes.




                                   25

-------
             The Four Corners study, conducted In 1972, was concerned with

air and land (strip mining, transmission lines) visual aesthetic impacts  of

coal-fired electric generating plants on residents and recreationlsts in

the Southwest.  Respondents were shown photographs illustrating (A) the

plant without pollution controls, and an unreclaimed strip mine; (B) the

plant with some controls (circa 1972), and a partially reclaimed strip

mine; and (C) the plant with no emissions, and no visible strip mining

impacts.  Residents said they would be willing to pay an average of $50 per

year in additional sales taxes in order to obtain an improvement from A to

B and an average of $82 to obtain an improvement from A to C.*  Total

regional WTP estimates were obtained by adding these results to the res-

ponses of recreationists concerning additional user fees they would be

willing to pay.  For A to B, the regional WTP was about $15.5 million per

year, and for A to C, it was about $24.5 million per year.

             The Lake Powell study, conducted in 1975, considered the

visual impacts of large power plants and visible smoke plumes in view from

a predominantly recreational site.  Recreationists and a few local residents

near Lake Powell were asked what they were willing to pay to prevent

construction of an additional plant if only the plant would be visible

and if the plant and pollution would be visible,  as opposed to current

conditions.  Average bids by users of the recreation area, in terms of

additional user fees per day, were between $.87 and $2.11 to prevent an

additional plant and between $1.75 and $3.38 to prevent an additional plant

and pollution.  The lowest average bids were by residents while the highest

average bids were by remote campers.  Aggregate bids indicate that the
*A11 benefit figures in each case study are reported in dollars current
 the year the study was undertaken.
                                   26

-------
benefits of preventing visibility degradation at that site alone  (there are



also several other recreation areas in the vicinity) were $400,000  to




$700,000 yearly.



             The Farmington study, conducted in 1977, was also concerned




with the impacts of coal-fired electric-generating in the Southwest.  The




focus of this study was the effects of reduced visibility upon local




residents in and near Farmington, New Mexico.  The three visibility levels




illustrated by photographs of two standardized long distance landscape



views from Farmington were (A) visual range of about 120 km, somewhat




better than current conditions; (B) visual range of about 80 km; and (C)



visual range of about 40 km.  The average monthly bid by residents to



prevent deterioration in visual range from 120 km to 80 km was $4.75, and



to prevent deterioration from 120 km to 40 km was $6.50.  Non-resident



recreationists in the area had average bids of $3.00 and $4.00 in additional



user fees per day for the same scenarios.  Aggregate benefit estimates  for




the study area over a 35-year horizon, assuming a 10% discount rate, were



$14.2 million for preventing A to B and $19.2 million for preventing A  to




C.  One of the purposes of this study was to test for biases in the bidding



process.  Hypothetical, starting point, and information biases were




detected.  Additionally, it was found that in contradiction with theoretical



expectations, WTA bids were much larger than WTP bids.   This was attributed




to differences in implied property rights in the two questionnaire proce-



dures .




        The South Coast Air Basin study,  conducted in 1978,  was an appli-



cation of the bidding method and the property value approach to the same



urban area in order to compare the results of these two techniques.  The



scenarios used in the bidding method were illustrated with photographs of
                                   27

-------
two views in Los Angeles showing (A) poor air quality as a visual range  of




about 2 miles, typical in much of the area; (B) fair air quality as a




visual range of about 12 miles, the predominant condition in the area; and




(c) good air quality as a visual range of about 28 miles.  Residents in  12




different communities, each categorized as having poor, fair, or good air



quality, were interviewed.  Average monthly bids in communities with poor



air quality were between $11 and $22 in order to obtain fair conditions.




In communities with fair air quality, average monthly bids to obtain good



conditions were between $5 and $28.  In communities with good air quality,




residents offered average bids of $18 to $67 per month to obtain good air



quality in the entire region.  For all proposed changes, aesthetic,  acute




health, and chronic health components each constituted about one third of



the bid.  Aggregate annual benefits of a 30% improvement in air quality for



the region were estimated between $580 million and $650 million.




        The Grand Canyon/Southwest Parks study,  conducted in 1980,  used



interviews with urban residents at their homes in Los Angeles,  Denver,



Chicago and Albuquerque.  The respondents were asked how much they were



willing to pay to protect visibility at the Grand Canyon and at other parks



in the Southwest region.  Both users and non-users were interviewed in an




attempt to derive both activity and existence values.  Photograph sets



depicted views from the Grand Canyon, Zion and Mesa Verde national parks,




each with five levels of air quality.  The middle level, C,  represented



typical current conditions.  The expected impacts of uncontrolled emissions



were shown in views A and B, and views D and E showed the expected impacts




of more stringent controls than were currently in effect.  The average user



value bid in terms of additional daily entrance fees to improve air quality



at the Grand Canyon from A to C was $2.76; and from A to E was  $5.15.
                                   28

-------
An additional $4.29 per day was the average user value bid to  prevent



deterioration in regional air quality from C to B.  The average preser-



vation value bid, which the authors defined as user plus existence values,




was $5.38 in terms of a residential monthly utility bill increment to




prevent deterioration at the Grand Canyon from C to B, and was an addi-



tional $4.58 to prevent the same deterioration region wide.  These results



indicate the existence values overwhelm activity values, even for park




users.  Aggregate annual benefits for residents of the Southwest region




were estimated to be $470 million for preserving visibility at the Grand




Canyon and $889 million for preserving visibility in the entire region.



Extrapolated to the nation as a whole, these annual benefit estimates were



$3,370 million and $5,760 million respectively.



        The magnitude of the preservation value bids from this study



suggests that the scenarios should be evaluated in terms of their credi-



bility to the respondents.  Future attempts to obtain existence values need




to be designed in order to test for hypothetical bias problems in the



procedure.




        The five property value studies reviewed in the Guidebook are the



Washington D.C. study, by Nelson (1978); the Boston study,  by Harrison and




Rubinfeld (1978); the Denver study, by Bresnock (1980); the South Coast Air



Basin study, by Brookshire et al. (1979; 1980); and the San Francisco Bay




Area study, by Loehman et al. (1980).  They are all based,  at least in




part, on the theoretical framework (described in Section 3.0) for applying



the hedonic price technique to residential property values  to estimate the




WTP for air quality at the residential site.  Each study was conducted in a




major metropolitan area where substantial and detailed data regarding



households, property, and air pollution were available.
                                   29

-------
        All of these studies have applied the hedonlc price  theory  In a




similar way.  However, there were some significant differences In the




specification of the hedonlc price function.  Due to differences In




housing markets and In data availability, different combinations of




property characteristics (In addition to neighborhood air pollution



measures) were used to explain property values.  All of the  studies derived



marginal Implicit prices for some measure of air pollution and estimated



WTF for reductions in air pollution, as previously discussed.




        One important characteristic of the property value studies  con-



ducted so far is that they have all used measures of air pollution, usually



of a specific pollutant, In the hedonlc price function.  Because many




common pollutants in urban areas have both aesthetic and health as well as



some material damage effects, there has been no attempt in these studies to




separate these impacts.  We do not know from these studies, therefore, how




much people are willing to pay to avoid simply the aesthetic impacts of air




pollution.  We know only what they are willing to pay to avoid the combined



impacts.




        In the Washington D.C. study, measures of TSP (total suspended



particulates) and oxldant were used.  Both measures were found to have a



negative relationship with property values.  Based on 1970 data,  the



estimated mean marginal implicit price of TSP was $69 per pg/m3 (micro-




grams per cubic meter) and of oxldant was $15 per .001 ppm (parts per



million).  Total benefit estimates were not derived.



        In Boston, based on 1970 data, the average property value differ-




ences attributable to air pollution between those neighborhoods with the




highest levels of pollution and those within the standards, all other
                                   30

-------
things being equal, was about $5,000 to $6,500.*  The reported mean

marginal implicit price was $1,613 for a one pphm (parts per hundred

million) change in NO .  Both NO  and ISP measures were used, but  the
                     X          A

results of the TSP equation were implausible and therefore the NO^

equation was used for further analysis.  Using the estimated WTP functions,

benefit measures for the reduction in emissions expected by 1990 from

automobile emissions standards were derived.  These averaged $83 per

household per year.

        The Denver study used 1977 data and measures of carbon monoxide

(CO) concentrations, one of the pollutants in Denver that frequently

exceeds national ambient air quality standards.  The mean marginal implicit

price was between $67 and $83 per one pptm (parts per ten million) CO.

This price implies a property value difference of approximately $10,000 to

$12,000 between those neighborhoods with the worst pollution levels and

those that met the standards, all other things being equal.  From the

estimated WTP functions, the benefits of attaining national ambient air

quality standards were estimated at an average of $258 per household per

year.

        The South Coast Air Basin study used 1978 data and measures of

NO2 and TSP.  The benefit estimates derived using each measure were quite

comparable to one another.  The mean marginal implicit price was $2,010
*These estimates of property value differences are ballpark figures given
 for illustrative purposes.  The study reports do not include enough
 information to calculate these estimates precisely; therefore, they are
 roughly calculated using the reported mean marginal implicit price of
 pollution.  Since the marginal implicit price usually increases as pol-
 lution increases, the actual property value differences attributable to
 pollution are understated in the higher pollution ranges.
                                   31

-------
per one pphm of N0_.  This implies an average property value difference




of approximately $5,700 between homes in poor and in fair air quality




locations, and an average difference of approximately $5,300 between homes




in fair and in good air quality locations, all other things being equal.




Average benefits for a regionwide 30% improvement in air quality were




estimated to be between $44 and $49 per month per household.  Aggregated




across all households in the region, this amounted to annual benefits of




$610 to $650 million for improvements from poor to fair air quality and



annual benefits of $340 to $360 million for improvements from fair to good



air quality.  These results are similar, but not equivalent, to those of




the bidding method study for the same communities.  Given the difficulties




in both procedures this is an encouraging result that supports the validity




and defensibility of the benefit measures.




        The San Francisco Bay Area (1980) study attempted to duplicate the



South Coast Air Basin study in order to test the comparability of the




results, but it was found that the study procedures had to be adapted to



the different study area characteristics.  The results of the two studies




are similar, but they do indicate that results cannot necessarily be easily




or directly transferred from one study area to the next.   The final results



of this study were based on measures of ozone and the PSI (Pollution



Standards Index) that included both concentration levels  and frequency of



high level concentration occurrences.  Although such measures make sense,



they are difficult to interpret.  The average annual benefits derived from



the estimated WTP function were $82 per household for a 30% improvement in



the ozone measure.  .Extrapolated to the entire study area,  this  amounted to



about $136 million per year.
                                   32

-------
6.0  FUTURE DIRECTIONS FOR VISIBILITY BENEFIT ANALYSIS






        Results of visibility benefit analyses conducted  to  date have shown




that visibility benefits can be quantified, that the amounts are substan-




tial enough to significantly influence benefit-cost analyses of  air




pollution controls, and that they improve the information available  for




resource allocation decisions.  Nevertheless, much still  needs to be done.




Problems in these applications do exist.  Complete confidence cannot as yet




be placed in the exact values so far estimated, but progress is  being made




in the application of these techniques which warrants continued  efforts in




this area.  First, the public and the public policy decision makers  need to




be made aware that the benefits of protecting visual aesthetics,  just as




the costs, are important and can be measured.  Second, continued  verifi-




cation and refinement of the estimation techniques are needed in  order to




improve the reliability and crediblity of the results.




        The bidding methods need verification through more comparisons  to




actual market results and through follow-up studies to determine  if  actual




behavior is consistent with predicted behavior.  The most important  ques-




tion concerning the bidding methods is whether hypothetical  markets  can be




used to derive accurate measures of visibility benefits.  If  it is deter-




mined that they can, the other application problems,  such as  implied




property rights, detection of false bids, payment vehicles, and selection




of WTP or WTA measures can probably be resolved.




        Hedonic property value studies have some theoretical problems  that




need to be resolved.  Most important is the question concerning the




relationship between'property value differentials and households' WTP, and




whether reliable WTP functions can be identified from the information
                                   33

-------
provided by a cross sectional study of a single housing  market.   Another



area that must be explored before property value  studied can produce



visibility benefit estimates is the identification of measures of aesthetic




impacts that can be separated from measures of health impacts of  air




pollution.



        More research is needed to determine the  applicability of contin-




gent travel cost, household production function, wage and salary,  and




ranked attribute approaches for visibility benefit analysis.  It  may be



that some of these can be applied in situations when other techniques




cannot be used.  Some of these techniques have been quite successfully



applied to other types of recreation demand questions.



        Finally, more visibility benefit analyses need to be  completed in




order that there be a sufficient set of results from which to choose for




transferring results to new study questions.  This will allow estimates of



visibility benefits to be obtained,  in many cases, more quickly and inexpen-



sively than by performing a complete visibility benefit analysis.
                                   34

-------