-------
Table 3.3
Daily Maximum Hourly Average Concentrations of Various
Pollutants in the South Coast Air Basin
(Arithmetic Average - 1975)
Montebello
Culver City
'Canoga Park
El Monte
Encino
Pacific Palisades
Newport Beach
Irvine
Palos Verdes
Redondo Beach
Huntington Beach
La Canada
Oxidants
pphm)
7.0
5.8
7.5
10.1
7.0
3.0
4.0
4.0
2.0
3.5
3.6
10.2
Nitrogen
Dioxide
(pphm)
12.7
13.0
11.0
12.5
9.1
6.2
6.7
6.8
6.3
8.5
9.0
11.2
Total Sus]
Particul;
(yg/m'
115
88
110
116
105
78
80
75
67
85
115
130
64
-------
CHAPTER IV
THE SOUTH COAST SURVEY QUESTIONNAIRE STUDY
4.1 Survey Instrument Design
Chapter II reviewed the theoretical and conceptual state of the art in
employing the contingent claims mechanisms. The essential questions addres-
sed implicitly in the discussion with regard to aesthetics and health ef-
fects in the South Coast Air Basin are: whether a valuation for an enviro-
mental good can be disaggregated into characteristic parts, the relative
efficacy and consistency of bidding and substitution formats in accomplishing
this task, and whether a survey instrument can be properly designed enabling
the estimation of the overall contingent valuation equation. This chapter
will present the structural design of the survey instrument, the method of
choosing the accompanying photographs, the survey implementation procedure,
and preliminary statistical results from the iterative bidding component of
the survey instrument.
The structural components and the directional flow of the survey instru-
ment are presented in Figure 4.1. Many types of information are sought by
the survey instrument. The first component can be viewed as establishing
baseline information about the respondent. The respondent's current indoor
and outdoor recreational activities, costs of both types of activities,
location of the activities, the frequency and duration of activities, and
the importance of the activities are established. The respondent is held
to a "typical week" time budget for indoor and outdoor activities that was
initially established in the questioning process.— This information was
then entered on the indoor/outdoor activity and cost lists in Tables 4.1 and
4.2.
At this point the interviewer presented information relating to either
aesthetic effects of visibility or health effects in the South Coast Air
Basin. Recalling the earlier discussion about information bias, the alter-
native initiation points for beginning the valuation process were a poten-
tial factor in the final results. That is, in disaggregating an environ-
mental good down into characteristic components, does the order in which
the characteristics are presented affect not only the final summed valuation
of the good but the characteristic parts valuation? In order to test this
hypothesis, information was obtained as presented in Figure 4.2. First, the
sample population was broken into two groups: those mailed a health bro-
chure (as in Appendix B) and those provided no additional aesthetic or
health information other than that presented in the survey instrument. Sec-
65
-------
Information Collective Flow for Survey Instrument
Figure A.I
Indoor
Costs
<
j uaseiine iniormation f
l/
Indoor
Activities
\
Outdoor
Activities
Outdoor
Costs
Expenditures
and Income
Indoor
Activities
1
\
Outdoor
Activities
._. I
J,
In f nr™n f \ nn \f- . — —
I No Substitutions
Bid on Aesthetics
Plus Acute Health
Substitute
Indoor
Activities
' \
Outdoor
Activities
Information
Bid on Aesthetics
Plus Acute Health
Plus Chronic
Health
Substitute
Respondent's
Evaluation
No Substitutions
(Step 1)
(Step 2)
(Step 3)
(Step
(Step 5)
(Step 6)
No Substitutions |
(Step 7)
Indoor
Activities
f
Outdoor
Activities
1 Hornr* I.ivinp I/
(Step 8)
(Step 9)
66
-------
Table 4.1
Outdoor Activity and Cost List
Activity
Outdoor Spectator
Sports
Tennis
BikinR
Beach Activities
General Exercise
Fishing
Swindling
Sailing
Jogging/Walking
Hobbies, Arts 4 Crafts
Outdoor Gardening or
Fixing up House
Golf
Hiking
Camping
v7
Organized Sports Events 1
Individual Sports
Events
Hours
Per Week
A
Other (specify) ; 1
B
C
D
I
Times
Per Week
A
B
C
D
Location
(Map Grid)
.A
B
C
D
Miles
Traveled
A
B
C
D
Direct
Costs
A
B
1
C
D
% Day
Equipment
Replacement
Costs
loportance
-------
Table 4-2
Indoor Activity and Cost List
Activity
Indoor Spectator Events
Indoor Tennis
Raquetball, Handball
Table Tennis
Bowling
Indoor Gardening or
rr, Fixing up House
General Exercise
Organized Sports Eventa
Reading
Television
Movies
Club Activities,
Organlzat Ions
Individual Sports
Swiraning
Visiting Neighbors or
Friends
Other (specify)
v7
Hours
Per Week
A
B
C
D
Times
Per Week
A
B
C
D
Location
(Map Grid)
A
B
C
D
Miles
Traveled
A P
C
D
Direct
Costs
A
B
C
-
D
Z Day
Equipment
Replacement
Costs
Importance
-------
Figure 4.2
Information Sequence in Survey Instruments
1. No Heath Information
Presented
Baseline
Info Sec-
II. Health Brochure
Previously Mailed
A. Aesthetic (l)Aesthetic (2)Aesthetic (3)Total
and Acute Acute, Chronic Check
B. Acute
(l)Acute
Chronic
(2)Acute
Chronic
Aesthetic
(3)Total
Check
C.
(1)
D.
(1)
(2)
(3)
(2)
(3)
-------
ond, these two groups were further subdivided into two additional categories
according to the sequence of information presented in the survey instrument.
Either a single individual was asked valuation questions about air
quality characteristics in an aesthetic affects, aesthetic plus acute
health affects and aesthetic plus acute plus chronic heath affects
sequence or acute health affects, acute plus chronic health affects,
and acute plus chronic health plus aesthetic affects. Data collected in
this manner would allow a statistical test of ordering and initiation
point effects in the overall valuation effort."L'
An iterative bidding format was administered based on a contingency
perturbation from the existing conditions presented to the respondent. This
represented an improvement from the original condition in the resident's
air (i.e., poor to fair, etc.). The bid was established using either a util-
ity bill or a lump sum monthly payment as the vehicle. Further, in order to
be able to observe any individual time discounting, the clean-up period was
set forth as either 2 or 10 years. Additionally, three alternative starting
points of $1, $10, $50 were employed to initiate the actual bidding process.
Finally, some respondents were handed a "life table" that would show the
total amount they would pay as long as they lived in Los Angeles depending
upon their bidu Thus, the iterative bidding format within the survey instru-
ment employed structural characteristics that allowed for eventual testing
of all the potential bias discussed earlier.
After the recording of the maximum bid, the interviewer moved to step
3. This step initially established the following: (1) the respondent had
stated a willingness to pay for improvements in air quality (even if zero);
(2) the respondent had less money overall as a result; and (3) thus (1) and
(2) indicate they value clean air and thus they have traded income for clean
airu Then the respondent was queried as to whether the improved air quality
conditions would alter their current activity patterns in any or all cate-
gories (i.e., time, duration, place and/or type). Thus column B of Tables
A.I and 4.2 were filled out with the time constraint being checked .A'
The beginning of step 4 essentially repeated steps 2 and 3 in procedure,
however, the information content was different. Consider previous bidding
games as a focal point. Typically, the process would involve yet another
perturbation of the environmental good in question. However, we are inter-
ested in attempting to disaggregate the characteristic parts of the enviro-
mental good air quality into aesthetic affects, and acute and chronic health
affectSo Thus step 4, depending on whether step 2 begins with information
on aesthetic or acute affects, presented either acute health affects infor-
mation for the former point or chronic health affects information for the
latter initiation point. The same initial vehicle was employed. If the
"life table" was used earlier it again was made available. The bidding
began with the last maximum stated bid of the previous step. Step 5 again
repeated earlier conditions (i.e., the trade of income for less health
effect) and the outdoor and indoor activity/cost lists were filled in for
column C.
70
-------
At the initiation of step 6, information regarding the last remaining
characteristic of the "good" air quality was provided which was either health
affects for the aesthetic initiation point or aesthetic effects for the acute
initiation point. Then the procedure for the iterative bidding was repeated.
Similarly, siren 7 renres^nted a renetition nf the substitution sections.
Finally, at the termination of step 7, a final review of bidding struc-
ture and the substitution answers x^ere reviewed. The respondent was
allowed any adjustments that were deemed necessary.—
Upon completing the iterative bidding and substitution sections, a
series of general information questions covering socioeconomic information,
property value information of the residence, type of residence (i.e., number
of stories, pool, rooms, etc.), reasons for current locational choice, health
related questions (i.e., heart trouble, medication, etc.), and attitudinal
questions relating to air quality were administered. This is step 8 in
Figure 4.1.
Finally, step 9 involved a respondent's evaluation of the survey (i.e.,
relevant, policy oriented, etc.) and an enumerator evaluation.
4.2 The Photographs Accompanying the Survey
The survey instrument in depicting air quality in the South Coast Air
Basin employed picture sets. This section will discuss the underlying
considerations in constructing the picture set employed in the South Coast
Air Basin.
Visibility is dependent on light. Light is a form of energy, made up
of electromagnetic energy, and is really a form of matter made up of indi-
vidual particles (photons). Light travels in streams and is subject to any
interference in its path. Light waves can bend, spread, interfere with one
another, and react with obstacles. Visibility is the state or quality of
being perceivable to the eye. It is a subjective term in its common usage,
referring to the general clarity of the air. In its more strict use, visi-
bility is defined as the farthest distance that any object of suitable size
can be visually identified without the aid of magnifying instruments. Both
the common and strict definitions of visibility suffer from lack of precise
meaning because of the many variables which are difficult to control. There-
fore, it is important that more precise definitions of visibility be ex-
plored in order to use the concept accurately..
There are three characteristics of a light wave that are of concern:
(1) its intensity, which is related to the height of the wave crests and
indirectly determines brightness of the light; (2) the wave length, which
depends on the distance between crests and largely determines color; and
(3) its polorization, the angular orientation of the crests. These three
characteristics are influenced by what happens when the light waves come in
contact with other matter. In particular, we are concerned with how these
characteristics affect changes in visibility as light waves interact with
particulate matter in the atmosphere.
71
-------
There are two issues in the way light affects visibility. First is the
ability of an object to reflect light in such patterns as define the visual
characteristics of the object. Second is the ability of that reflected
light to reach the observer in such a way as to differentiate the charac-
teristics of an object from the background. First,., let us assume that
every object, except a perfectly black object, reflects light some distance.
Further, if the light reflected from an object reaches an observer and that
object is distinguished from the background, it is said that the object is
visible to that observer. Visibility is not only dependent on light but
upon the distance between the object and the observer. As the distance
increases, less and less light reflected from the object reaches the ob-
server until the object is no longer distinguished from the background.
When the observer can no longer distinguish the object from the background,
the object is said to be beyond the visible range. In summary, the visi-
bility of an object illuminated by light depends upon the apparent contrast
between the object and its background, the ability of the observer to dis-
tinguish the object from its background, the size of the object and the
angle of reflection, and the condition and technique of observing.
Three definitions of visibility are commonly found in the literature.
Visual Range: A dark object is moved through the atmosphere toward the
horizon sky. As the distance between the object and observer increases,
contrast between the object and horizon sky decreases. At some dis-
tance the contrast between object and horizon sky becomes too small
to be distinguished, and the object "vanishes." The distance between
the observer and the object at the "vanishing point" is the visual
range.
Prevailing Visibility: The greatest visibility which is obtained or
surpassed around at least half of the horizon circle, but not neces-
sarily in continuous sectors.
Meteorological Range: The distance at which the contrast of an object
is reduced to the point where the human eye can no longer distinguish
it from the background, or that distance for which the contrast
transmittance of the atmosphere is two percent.
It is possible under a certain set of circumstances to measure visi-
bility by using photography. Stephens (1949) developed a method for mea-
suring photographically the "extent to which visual range has been reduced
by haze."
Briefly, the technique involves photographing (on black and white film)
black objects that are far enough away to be obscured. Then the photo-
graphic densities of the objects and the adjacent sky are measured on the
negative. Calculated from these relative densities are the visual range,
distance of the object, and contrast of the film.
The theory of photographic photometry, used to calculate long-range
visibility, as in Roberts', et. al. (1974) study of visibility measurements in
the Painted Desert, states that if a "black object of sufficient size is
moved through the atmosphere away from the observer, the object will appear
to become brighter as the distance from the observer increases, even though
72
-------
the level of illumination remains constant." The apparent increase in
brightness is the result of light being scattered as it moves toward the
observer by suspended aerosol particles in the air between the object and the
observer. The effect of an apparent increase in brightness illustrates the
reduction of light as it moves through air that contains particulates. It
is this effect that is detected by the technique of photographic photometry.
The apparatus utilized in photometry is very simple. All that is needed
are a camera, a positive gray scale, some means of measuring the distance
from the observation point to the object photographed, and a densitometer.
A densitometer is the devise by which the relative densities on the negative
are measured.
Unfortunately, photographic photometry has various problems which may
cause problems in insuring reliable results. First, for the purposes of this
study, it is crucial to differentiate between visibility reductions due to
natural haze and polluted haze. We are attempting to measure the increase
in haziness (the decrease in visibility) made by pollution. Photometry
simply measures the visible range, without regard to the differentiation of
natural and polluted haze.
There are interrelations among the specifications for the object, the
densitometer, and the camera. The size of the image on the negative whose
density is to be measured depends on the size and distance of the object
and on the focal length of the lens. The minimum size of the image that can
be used depends on the characteristics of the densitometer utilized. To
further complicate the interrelation, since with any ordinary lens the
illumination at the focal plane rapidly decreases toward the edges of the
frame, it is necessary to find what area of the negative is satisfactorily
uniform in relation to the particular camera utilized. This illumination
function is found by trial and error and is beyond the scope of our present
efforts.
Ideally, data should be obtained by photographing distinct objects in
each possible direction once each hour from each sample area and from one
general area.^-' At least two distinct objects should be included in each
observation path to insure that, regardless of the atmospheric conditions,
data can be obtained from the photographs. It is desirable to include views
in all quadrants because of visibility and meteorological differentials
across areas. Observation points were chosen with these criteria.
We define "object" as some unique natural or man-made phenomenon in the
landscape that is distinct from its immediate surroundings. "Observation
path" is the line of sight. It is important to properly identify and locate
objects in the observation path, certainly if accurate measurements are to be
made. It further helps the respondent gain perspective when viewing the
picture set. Proper identification and measurement of objects in each
observation path chosen was accomplished using city and topological maps as
well as visual inspection.
"Observation point" is simply the place in each area from which photo-
graphs were taken. Earlier in this report we noted that one criteria in the
73
-------
selection of some sample areas in the SCAB was a view. Therefore, obser-
vation points in the sample that are with a view were cnosen with'the aim of
representing to respondents within each sample area a scene that they typ-
ically observe. In this way, it was intended that the photographs would
merely serve as a reminder to respondents of the changes in visibility due to
air quality.
Certainly the most important consideration is what is contained in each
observation path and therefore in each picture. Ideally, each observation
path should have at least two readily recognizable objects with which the
majority of respondents are familiar, allowing them to estimate easily visibility
by the contrast of those objects. The observation paths and the objects
therein should be concerned foremost with the portrayal of a visibility
gradient (in our study, "good," "fair," and"poor"), and should be very care-
ful to exclude objects that may trigger bias in the respondent in responding
to something besides visibility (or health affects). For example, a free-
way interchange in the picture may stir up negative feelings in the respondent
even before the respondent considers the impact of changes in visibility.
Such undesirable characteristics in the observation paths may increase the
chance for bias in the valuation procedure.
Another very important consideration was to insure consistency in field
operations. There was a standard operating procedure at. each observation
point. Each photograph was taken with identical equipment. In order to as
closely as possible duplicate the quality of photographs from each location,
each photograph was taken with the same model Minolta SLR camera, 135 mm
lens (used for the photographs the respondents saw), 55 mm lens (to record
on film the local weather conditions for future reference), and the same high
quality professional color film.jj/
Crucial to the photographic data collection effort and the standard-
ization of field operations was for each photograph to be accurately logged.
Thus, each frame of exposed film was logged and each step in the procedure
was carefully recorded so as to minimize discrepancies between observation
points.
For each exposed frame, the researcher kept a record of various char-
acteristics. The time and date of exposure is important in order to co-
ordinate the data the research team collects in the field with the data
collected by the local airports and local air monitoring stations. The F-
Stop (aperature opening) and shutter speed were recorded so as to further
estimate changes in visibility. Since the photographs were going to be shown
to respondents as well as analyzed, it was important to insure the quality
of the photographs. By quality, we mean that each photograph must be an
accurate rendition of the air quality as prevailed during each exposure. In
order to insure a proper exposure, each photograph was "bracketed." That is,
for each photograph one frame was taken with a normal meter reading, then
one frame of the same observation path was "underexposed" (meaning one F-
Stop above normal keeping the same shutter speed as for the normal photo-
graph) , then one frame "overexposed." In this way we were assured that the
best possible representation on film of each observation path was .produced.
74
-------
It should be emphasized that the objective of these photographs is to
portray to respondents changes in visibility due to changes in air quality.
Observation points and observation paths were chosen primarily on the cri-
teria already listed, but once these sites were chosen, pictures could
only be used if the changes in air quality were such as to represent the
range from clarity to visual obscurity that is typical for each area during
the year. Of course, changes in visibility due to changes in air quality is
quite out of our control. Therefore, we could only use those photographs
in the survey instrument portion of this study that were indeed representa-
tive of the typical range of visibility for any one area. Such photographs
could only be obtained if the air quality was "right." The results of this
effort are summarized below.
Photographs were taken from seven sample area observation points and
from one general site observation point. By "general site" we mean some
area or view that would likely be familiar to most of the respondents no
matter where they lived in the South Coast Air Basin.
Figure 4.3 entitled "Los Angeles Observation Paths" depicts the seven
sample areas and the Griffith Park Observatory. The map is scried as shown
and each vector eminating from specific observation points represents fif-
teen miles.
The Griffith site afforded three excellent observation paths: (1) to-
ward downtown Los Angeles, with large buildings approximately five miles
from the observation point; (2) down Western Avenue toward large buildings
approximately four miles away and toward two sets of hills in the back-
ground; and (3) southwest toward large buildings near Beverly Hills.
Recall that we have six pairs of sample areas:
1) Canoga Park* El Monte
2) Culver City* Montebello*
3) Newport Beach* Pacific Palisades
4) Irvine Palos Verdes*
5) Encino* La Canada*
6) Huntington Beach Redondo Beach
Those areas marked with an asterik (*) were chosen as tentative sites
for observation points. A brief description of each observation path from
each site is as follows:
La Canada: (4) northeast across the basin toward mountains; and (5)
northwest through the basin toward the mountains.
Encino: (6) northeast toward large buildings with mountains in the
background; (7) north toward two sets of large buildings at different
distances with mountains in the background; (8) north-northwest toward
large buildings with mountains in background; and (9) west down Ventura
75
-------
Figure 4.3
Observation Paths in the South Coast Air Basin
Los Kngeles
Observation Paths
Canada
Griffith Park
Observatory
A Montebello
, , Miles
012345
Each vector « 15 miles
76
-------
Boulevard toward the mountains.
Canoga Park: (10) north-northeast toward large buildings with moun-
tains in the background; (11) north toward sets of large buildings with
mountains in the background; and (12) west toward a set of large buildings
with mountains in the background.
Culver Cit.y: (13) northwest toward a set of large buildings with
mountains in the background; (14) west toward buildings in Santa Monica; and
(15) southwest toward two large buildings in Marina Del Ray.
Palos Verdes: (16) north toward buildings in Beverly Hills; and (17)
north-northeastern toward large buildings in downtown Los Angeles with moun-
tains (Griffith Park) in the background.
Montebello: (18) south-southwest toward buildings; and (19) southeast
toward Whittier with hills on one side of the observation path.
Newport Beach: (20) northeast toward buildings with two sets of moun-
tains in the background; and (21) east across the Bay toward hills with moun-
tains in the background.
On numerous occasions, photographs were taken from the eight obser-
vation points and the twenty-one observation paths. For each observation
path, of course, the attempt was to photograph "good," "fair," and "poor"
days of visibility. This was successfully accomplished for the Griffith
Park site, but, was unsuccessful for all specif ic. samnl R areas p.xcent Encino.
For the other areas, we were unable to obtain the necessary gradients in the
photographs that would represent the typical range of visibility for each
area.
The photographs used in the asking games for each sample area were the
observation paths from Griffith Park toward downtown Los Angeles and down
Western Avenue. Figures 4.4a-c present the actual photographs in a black
and white version. The visibility for picture set A (poor) was estimated at
2 miles, for picture set B (fair) at 12 miles, and for picture set C (good)
at 28 miles.
The researchers were unable to obtain a poor air quality picture set
for the Griffith Park area with the same light and color characteristics
as the good and fair picture sets, although pictures x^ere obtained for this
location of approximately 2 miles. In consequence, the researchers sub-
stituted a picture set with similar foreground and light and color charac-
teristics taken at approximately the same time in Orange County, California.
4.3 The Surveying Procedures
This section will detail the actual sampling procedures given the
sample plan discussed earlier. The first task was to identify a group
in each paired area to receive a health brochure. The second task was the
actual administering of the survey instrument.
77
-------
Figure 4.4a
(Good)
Photograph Depicting Observation Paths for "Good" Visibility
78
-------
Figure 4.4b
(Fair)
Photograph Depicting Observation Paths for "Fair" Visibility
79
-------
Figure 4,4c
(Poor)
Photographs Depicting Observation Paths for "Poor" Visibility
80
-------
Late in September 1977, a team of University of Wyoming and University of
Southern California personnel contacted by telephone a random sample of the
population in each of the twelve final sample areas. The team was equipped
with reverse telephone directories- (i.e., phone directories listing people
by address instead of by name) which enabled accurate isolation of names and
addresses of potential respondents within the boundaries of each sample area.
Once streets and addresses were located within each sample area, a random
number generating table was utilized to pick the names of potential respond-
ents from each street.—' This was done to insure that no bias would be
introduced into the telephone sampling process.
People were then randomly contacted by telephone until at least thirty
people per area had agreed to cooperate wjLth the research team in the air
quality study. Thus a minimum of 360 people, distributed over twelve sample
areas in the Los Angeles Basin, were to be the respondents in the asking game
portion of the study.
Then in Spring 1978, half of the potential respondents were sent a
health pamphlet entitled "Air Pollution and'"Health," The half of the poten-
tial respondents receiving this pamphlet was to be the group upon which we
would test a learning hypothesis of the asking games. Approximately 180
potential respondents comprised this group.
In early March 1978, a research team comprised of staff and graduate stu-
dents from the University of Wyoming and a similar team from the University
of New Mexico went to Los Angeles to begin the survey instrument portion of
the study. The two teams were divided into four groups in order to sample
each of the twelve sample areas most efficiently.
The first order of business was to contact each of the potential re-
spondents by telephone and set up appointments with them. Although most of
the potential respondents could be reached by telephone, an unexpectedly
high percentage of persons who had previously agreed to cooperate with us
in the study declined the interview. This drastically cut the potential
respondent list and forced alternative methods of sampling.
Because of this setback, the sampling process was broken into three
parts. First, of course, we arranged interviews with those respondents
from our original lists who had said they were willing to cooperate and who
were still interested. Second, we once again utilized the reverse telephone
directories to set up new appointments with people in each sample area.
Third, each sample area was canvassed by members of the research team by a
door-to-door method. By this procedure the sample size was approximately
345 interviews..successfully completed by the research team. Table 4.3
presents the breakdown of the resulting sample by type of survey instrument.
4.4 Preliminary Empirical Results for the Iterative Bidding Portion of
The Survey Instrument Study
This section will present preliminary results of the iterative bidding
format portion of the contingent valuation study. Initial bias tests will
be presented including vehicle bias, starting point bias, and the potential
81
-------
Table 4.3
Survey Instrument Type Breakdown
Questionnaire
Type
Location
La Canada (A ->• B)
La Canada (A -* C)
El Monte (A •* B)
El Monte (A -» C)
Nontebello (A ->• B)
Kontebello (A .+ C)
JCanoga Park (B -» C)
"Encino (B ->- C)
Culver City (B -» C)
Pacific
Palisades (C -f C*)
Redondo Beach (C •* C*)
Palos Verdes (C -* C*)
Huntington
Beach (B -» Cl
Newport Bench (B f C)
Irvine (B -+ c)
TOTALS
Health Pamphlet
Aesthetic
Kale
1
2
1
3
3
1
1
12
Female
1
4
1
3
1
1
1
3
2
4
3
24
Acute
Male
1
1
1
It
1
1
1
3
3
16
Female
1
2
2
2
2
2
1
12
No Health Pamphlet
Aesthetic
Male
2
2
2
5
3
3
6
8
4
5
5
1
5
4
7
62
Female
7
5
5
7
5
5
it
6
8
3
5
7
8
7
7
89
Acute
Male
4
3
6
2
4
2
3
' 5
4
1
4
7
2
4
51
Female
1
4
3
6
4
4
3
5
9
6
4
7
16
2
5
79
Total by Area
by Type
1-°.
14
16
2C
18
20
20
28
30
2]
27
20
40
22
27
345
-------
for sequencing effects. Additionally, some preliminary regression results
will be discussed in Appendix D.
Table 4.4a,b presents the mean total bids by area partitioned by proposed
clean-up date. The results in Table 4.4a range from $47.75 per month for the
Pacific Palisades area to $4.50Per month in the Newport Beach area.I/ This
difference in mean bids between Pacific Palisades and Newport Beach is not
fully understood at this time. As was set forth in the theoretical Chapter
II, these results are not commensurate with economic expectations. This
problem can be investigated when the substitution results are integrated into
the analysis.
Appendix C presents tables for cumulative mean bids by sequence where ei-
ther aesthetic or acute information is presented first by area and differentiat-
ing between a 2 or 10 year clean up time horizon. All other potential effects
such as biases are assumed zero. The results presented in the appendices form
the basis of some simple statistical tests. The tests to be considered are
whether:
1. the area mean bids are significantly different from zero;
2. the aesthetic, acute, chronic, and total bids for the paired
areas are significantly different;
3. the results indicate the existence of starting point, vehicle
or sequencing bias; and
4. the results indicate different bidding behavior when individuals
were offered different completion dates for cleanup.
The results of the t-tests regarding the equality of area mean bids
being statistically different from zero are presented in Table 4.5. Of
interest is whether the results of the test allow the null hypothesis to
be rejected. In all but the one case of Montebello area for the chronic
bid, the null hypothesis is rejected with 90% confidence. Some cases sug-
gest higher levels of confidence. Thus, we can initially infer that in
all areas, the values individuals place on the. three characteristics of air
quality under consideration tend to be non-zero.
In Table 4.6, the equality between bids between the paired areas is
tested for the three characteristics and the total bid. Only two pairs
reject the null hypothesis that the two areas' mean bids are equal: Paci-
fic Palisades/Newport Beach and Culver City/Montebello. The former was for
aesthetic, acute, and the total bid while the latter was for the acute health
bid only. The purpose of this test can be seen in reference to the discus-
sion in the contingent valuation theory section. At issue was the difference
between a bid from the property value study in comparison to the iterative
bidding study. Recall that a contingency proposed to an individual was
moving him along an indifference curve. Assuming that each area represents
a homogeneous set of preferences which differ across areas, the test in
Table 4.6 asks whether the movement in dollar amount is the same across the
paired areas.
83
-------
Table 4.4a
Mi-.HI UUit by Ari-.n by Type*
(C<>M|ilt:Lliiii Date t>C CltMiiupi 2 Yr«.)
Arc.l
El Kin to (A - »)
Ei Horn c (A - C)
la C.in.1il--i U •• H)
U Con.ida (\ • C)
Honccbello (A •, B)
Hontcbcllo (A - C)
ConOfa Pnrk (b -» C)
Culver City (B . C)
Enclno IB • C)
Huntlncton Bcacli (B - C)
Irvine (B •• C)
Mcvporc Bc.iclv ' Hie rL'S|H)'Kj*-iU In
h.is !.<•»•" f-iik- vitli rc»pcc
colU-clJon of bit!-*; .-uirt {
life c.Mc II.-I.H nr li.vi uni
A 1 Iff l:il»Io U(.-|'U-i:. tin-
(or viirltiu!! t-xi'f(-n.-d II (e
Lion In tMA t.iMc h.is been Hint of btrlct aadlttvi:
Ity cffrct. In obc.ttnlnf, the mo .in bidw: (1) no
I».TI(C with re;: eel lo Che bitldinp. SCIIUIMICC ; (2) rxo
i.i.itic vlmthkT lit-.jl Ch p;impliU't I'.lt* or !us not bt-ctl
juiv.nscv ol" i c Interview; (3) n*> <:l f fcr^nt iatlon
i to itie 4lJff rtiit pro|io:icd vrhlcli-« for the
•'•) no lUrfcn- iJatlou h.in lin.n n-nlc ulivthirr a
*St:nul.iril error oE
*S:im|»lv MJru of cue
":.UKk" cotuu ri-.irii* of llu- »'Uciii-rt muntlily bJds
iq..ui;i.
hu mr>.in tiid In all cnsus.
C.IHC Jn All <-.»Hon.
-------
Table 4.4b
HL.III IMUti by An-u by Type*
iU-llon D.uc of Clc-.inun: 10 yra.)
Ar,-a
El ('.jut.- (A - B)
Kl tloiuc (A - C)
U Cnii.ida (A - t)
U Canada (A - C)
Hontcbcllo (h -. B)
Honlubcllo (A - C)
Canoc'i I'nrk (» - C)
Culver city (8 - C)
Enci:io (1) - C)
Huiuinstor. Bi-ach (8 - C)
Uvtnc (B -, C)
Hcvport Bench (3 - C)
Pacific fjlluadcs (C - C«)
Palos Vi.ciics (C - C*)
Kedondo tloicli (C •• C«)
At-sttu-tlc
Bin
1.67
(!.f)7)«
(,)._..
1.17
.(1.54)
3.60
(1.79)
14.43
(7..S5)
2.70
(1.32)
(101
4.38
(1.52)
(8l_
3.48
(1.19)
(HI
U.08
(4.41)
(12).
3.27
(1.93)
(l.D...
10.22
(3 . 30)
(I?)
10.90
(4.00)
liil_
1.15
(0.61)
5.58
(2.14)
'(12)
5.36
(1.24)
(IV)
12.46
(4.68)
(12)
•Van Rica
Aruie !k-.-llt!i
U 1 tl
11.89
(5.28)
L'i)
5.33
(3.07)
ty
15.60
(9.'»0)
(.IPJ
10.7!
(6.85)
_. O)
8.80
(4.73)
02L
1.38
(0.78)
(6)
2.07
(0.94)
OJJ
8.54
(?.54)
— —iLU.
5.36
(3.19)
(U)
10.79
(2.83)
tZ£I_
10.54
(3.92)
(12)
2.00
(1.09)
.___ (1C)
14.83
(3.45)
(12)
13.09
(5.12)
(")
6.96
(4.05)
(12)
(S./ao,,a)
Chronic ik'.ilth
BU
1.11
(1.11)
1.17
(O.S3)
(6)
9.50
(8.96)
(10)
0.43
(0.43)
(7)
1.70
(0.67)
(10)
0.75
(0.53)
(8)
0.48
(0.33)
ill)
4.08
(2.10)
(12)
1.45
(0.65)
(11)
6.84
(2.3S)
(19)
1.94
(0.92)
(12)
3.45
(2.06)
(10)
59.67
(39.58)
(1?)
2.73
(1.80)
(11) ._
4.42
(1.73)
.(12) ..
Total
Bid
14.67
(5.55)
(9)
10.67
(2.97)
(6)
28.90
(52.73)
(10)
25.57
(7.86)
(7)
13.20
(5.69)
(10)
6.50
(2.28)
(S)
6.03
(1.26)
(11)
23.71
(8.35)
(12)
10.09
(3.61)
28.42
(5.91)
_yjy
23.38
(5.96)
_oy_
6.60
(2.89)
(10J
80.08
(41.34)
(12)
21.18
(5.40)
. X'U .
23.83
(9.05)
(12).__
M'lic implicit .issii;a;it ion ^ this table In?; bi-cn tli.n
of h!rt:= (or r.icli air qti.iliiy cld'CU 1« ol»t:\liil-.ij*, il\c T.
olU-cUoa of hiiiy; ;i i»il ('•) no <\ 1 ( fcn-ut J-iL l<
Jfc CtbK- l-:is or h.Tb m»L It•.•*•» »;liuvn Lo the r<-M-
li!f ciMc iN-plrta ilit- "stoc'h" countcrpartti oC
or v.ir louy L-x]K-ccod 1 If c;"i>.Tnn,
of otrict adriUivtty
.in bic-j: (1) no
:; (2) no
; nut tictn
•iul.lt Ion
the
cioulUly bida
*SiniuUu-d t-rror of the mean bid in all caeca.
-------
Table 4.5
Results of the t-tests Regarding the Equality of Area Mean Bids to Zero*
H : The mean bid is equal to zero**
o
H : The mean bid is greater than zero
Name of Area
El Monte
El Monte
La Canada
La Canada
Montebello
Mon tebello
Canoga Park
Culver City
Encino
Type of
Contingency
(A - B)
(A + C)
(A -+ B)
(A -* C)
(A •+ B)
(A -* C)
(B •* C)
(B + C)
(B -<• C)
n
20
13
17
17
19
19
19
28
28
Aesthetic Bid
Reject H at 95%
0
Reject H at 95%
o
Reject H at 95%
0
'
Reject H at 95%
o
Peject H at 99%
0
Reject H at 90%
0
Reject H at 99%
Reject H at 99%
°
Reject H at 99%
o
Acute Health Bid
Reject H at 95%
o
Reject H at 95%
o
Reject H at 90%
o
Reject H at 95%
o
Reject H at 99%
o
Reject H at 99%
o
Reject H at 99%
o
Reject H at 99%
o
Reject H at 99%
Chronic Health Bid
Reject H at 95%
0
Reject H at 95%
o
Accept H
o
Reject H at 95%
o
Reject H at 95%
o
Accept H
o
Reject H at 95%
0
Reject H at 99%
o
Reject H at 93%
(continued)
-------
Table 4.5
(continued)
Name of Area
Huntington Beach
Irvine
Newport Beach
Pacific Palisades
Palos Verdes
Redondo Beach
Type of
Contingency
(B ->• C)
(B •* C)
(B + C)
(C ->- C*)
(C -*• C*)
(C ->• C*)
n
38
27
20
20
19
26
Aesthetic Bid
Reject H at 99%
o
Reject H at 99%
0
Reject H at 99%
Reject H at 95%
o
Reject H at 99%
o
Reject H at 99%
0
Acute Health Bid
Reject H at 99%
o
Reject H 'at 99%
o
Reject H at 95%
o
Reject H at 95%
o
Reject H at 99%
o
Reject H at 99%
o
Chronic Health Bid
Reject H at 99%
o
Reject H at 99%
o
Reject H at 95%
0
Reject H at 90%
o
Reject H at 95%
0
Reject H at 99%
o
*The bids for each air quality effect are assumed to be strictly separable. In obtaining the mean bids,
no differentiation is made with respect to: (a) different bidding sequences; (b) vehicle used; (c) starting
bid; (d) health pamphlet versus no health pamphlet; or (e) life table versus no life table (f) completion date of
cleanup.
**0ne-tail test t T~-— where p = area mean bid for a certain air quality effect
//—
s = sample standard deviation
n = sample size
-------
Table 4.6
Results of the Bid Equality Tests of the Paired Communities*
H : The two mean bids are equal
o
H : The two mean bids are unequal
Paired Areas
Pacific Palisades
Newport Beach
Canoga Park
El Monte
Irvine
Palos Verdes
Encino
La Canada
Huntington Beach
Redondo Beach
Culver City
Montebello
N
20
20
19
33
27
19
28
34
40
26
28
38
Aesthetic Bid
Reject H at 99%
Accept H
Accept H
o
Accept H
Accept H
o
Accept H
Acute Health Bid
Reject H at 99%
o
Accept H
o
Accept H
o
Accept H
0
Accept H
o
Reject H at 95%
0
Chronic Health Bid
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Total Bid
Reject H at 95%
o
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
o
CO
CO
*The statistical test employed was a two-tailed t-test of the null hypothesis (H ) that the mean bids for
each of the paired communities were equal. The test also used a pooled variance estimate in the celculation of
the test-statistic. The information in the major cells of the table reports the level of significance for the
statistical tests. Rejection of H at the reported significance level means that the test failed to reject H at
a higher level of significance. Only three significance levels were tested: 90%, 95%, and 99%. Failure to feject
H means that H could not be rejected at the 90% level or greater. The purpose of the above tests is to check if
tnWe is a stat2stically significant difference between the mean bids of the areas within the same pair. Throughout
this analysis, the bids for each air quality effect are assumed to be strictly separable. In obtaining the mean
bids no differentiation is made with respect to: Ca) bidding sequence; (b) vehicle used; (c) starting bid;
(d) health pamphlet versus no health pamphlet; or (_&} life table versus no life table (f) completion date of cleanup.
-------
In an iterative bidding format, various types of biases might be
introduced by the structure of the survey instrument. In this s.tudy, the
types of biases selected for examination were vehicle bias, starting point
bias, and information sequence Bias.
A test of means was conducted between the monthly utility bill and the
lump sum payment mechanism for the areas- by characteristic bid and for the
total bid. Table 4.7 presents the results. The null hypothesis set forth
was that the mean bids were equal irrespective of the Bidding vehicle. For
Montebello, Canoga Park, Encino, Huntington Beach, Newport Beach, Pacific
Palisades, Palos Verdes, and Redondo Beach, the null hypothesis is accepted
for the total bid. However, for Irvine, Culver City, La Canada, and El Monte,
we reject the null hypothesis, at least at the 90% confidence level, for the
total bid. No obvious reason exists at this point in time for this result.
The principle problem area then appears to be in the aesthetic bids.z.'
A second bias of concern is that of starting point bias. Recall from
previous discussions that starting point bias results from the final bid
being definitely related to the starting bid, i.e., the higher the starting
point, the higher will be the final bid, thus suggesting a type of inform-
ation bias. Table 4.8 presents the results of a test for starting point
bias. The structure of the test was as follows. Three starting points of
$1, $10, and $50 were employed in the survey instrument. This results in
three potential comparisons of starting points for the resulting mean bids:
(1) $1 to $10; (2) $1 to $50; and (3) $10 to $50. The null hypothesis was
whether the total mean bids were equal within the three combinations of mean
bids ignoring all other potential effects. For the $1 to $10 pair, the null
hypothesis of no effect was rejected in La Canada and Encino. The. $1 to
$50 pair was rejected for La Canada and Montebello. Finally, the $10 to $50
pair was rejected only for Redondo Beach.
To fully understand why the isolated cases indicate starting point bias,
a greater understanding would require consideration of other systematic
effects in the data set. However, preliminary evidence based on Table 4.8
suggests that starting point bias is not a major problem for all of the
iterative bidding results.
Another area of consideration is the question of sequencing of inform-
ation affecting the bid structure not only for the air quality characteris-
tic bids, but also the final bid. Recall that bids were collected according
to the following sequences:
1. aesthetic, aesthetic plus acute, and aesthetic plus acute plus
chronic, or,
2. acute, acute plus chronic, and acute plus chronic plus aesthetic.
The question of sequencing is whether the ordering of the bidding process
effects the size of the bid. For instance, will individuals bid a different
amount for aesthetic effects if it is first, as in (1) above, compared to
being last as in (2) above. Similarly, will the acute bids vary? Addition-
ally, we are interested in whether the orderings presented in (1) and (2)
39
-------
Table 4.7
Results of Che t-tests for che Equality of the Mean Bids
by Sample Area by Bidding Vehicle5'1
H : The two .mean bids are equal
o
H : The two mean bids are unequal
Name of Area
El Monte
La Canada
Montebello
Canoga Park
Culver City
Encino
Huntington Beach
Irvine
n^*
20
22
21
1
17
U
18
9
n2**
13
12
17
12
11
14
22
18
Aesthetic Bid
Accept H ***
0
Accept H
Accept H
o
Accept H
o
Reject H at 95%
Accept H
o
Accept H
o
Accept H
o
Acute Health Bid
Accept H
0
Accept H
o
Accept H
o
Accept H
o
Accept H
Accept H
o
Accept H
o
Reject H at: 90%
o
Chronic Health Bid
Accept H
o
Accept H
o
Accept H
o
Accept H
0
Accept H
0
Accept H
o
Accept H
o
Accept H
o
Total Bid
Reject H at 90%*>'<**
o
Reject H at 90%
o
Accept H
o
Accept H
o
Reject H at 90%
o
Accept H
o
Accept H
o
Reject H at 95%
o
(continued)
-------
Table 4.7
(conciiuied)
Name of Area
Newport Beach
Pacific Palisades
Palos Verdes
.
Redondo Beach
"1**
13
11
12
'
17
nz**
7
9 '
7
g
Aesthetic Bid
Accept H
o
Accept H
0
Reject H at 95%
Accept H
o
Acute Health Bid
Accept H
o
Accept H
Accept H
o
Accept H
o
Chronic Health Bid
Accept 1!
0
Accept H
0
Accept H
o
Accept H
o
Total Bid
Accept H
0
Accept H
0
Accept H
o
Accept H
. o
"Throughout the questionnaires, rvo different vehicles of payment are employed. Some Respondents are
proposer! to pay their bids in separate monthly payments, and some others in additions to their utility bills.
This tabJe is prepared to check if the choice of the payment vehicle has any statistically significant effect
on the mean of the bids.
The bids for each bidding stage are assumed to be strictly separable. In obtaining the mean bids no
differentiation is made with respect to: (1) starting bid; (2) bidding sequence; (3) health pamphlet versus
no health pamphlet; and (4) life table versus no life table-(5) completion date of cleanup.
**n^: Sample size of the interviews in which the respondent was proposed to pay his bids in separate
monthly payments.
**n2: Sample size of the interviews in which the respondent was proposed to pay his bids as additions to
his utility bills.
***The tests are done for <* •= O.Q1, * = 0.05, and =; = 0,10^ "Accept HO" means HQ is accepted for 1-* ~ Q.90
and higher, i.e., for « <_ 0.10. "Reject Ho at X%" means Ho is rejected at the given X% but is accepted at the
next higher 1-= value, i.e., if x% - 90%, then Hn is accepted at l-« - 95%.
-------
Table A.8
Test of Means for Starting Point Bias*
H : Mean bids are equal
o
H : Mean bids are unequal
Name of Area
El Monte
La Canada
Montebello
Canoga Park
Culver City
Encino
Huntington Beach
Irvine
Sample
n
8
8
15
13
13
12
16
16
12
7
7
6
9
9
11
9
9
11
8
8
15
7
7
12
Sizes
"2
15
9
9
12
9
9
12
9
9
6
6
6
11
8
8
11
8
8
15
15
15
12
8
8
Starting Point Pairs
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
Totals
Accept H **
Accept H°
Accept H°
o
Reject H at
Reject 11° at
Accept H°
o
Accept H
Reject H° at
Accept H°
o
Accept H
Accept H°
Accept H°
o
Accept H
Accept H°
Accept H°
o
Reject H at
Accept H°
Accept H
o
Accept H
Accept H°
Accept H°
o
Accept H
Accept H°
Accept H°
o
95%
90%
90%
90%
(continued)
92
-------
Table 4.8
(continued)
Name of Area
Newport Beach
Pacific Palisades
Palos Verdes
Redondo Beach
Sample
n.
1
9
9
6
7
7
6
2
2
9
10
10
10
Sizes
n
2
6
5
5
6
7
7
9
8
8
10
5
5
Starting Point Pairs
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
1-10
1-50
10-50
Totals
Accept H
Accept H°
Accept H°
o
Accept H
Accept H°
Accept H°
o
Accept H
Accept H°
Accept H°
o
Accept H
Reject H° at
Reject H° at
o
i
95%
90%
*The purpose of this table is to check if there is any significant
influence of the starting bid offered by the interviewer on the total bid
of the respondent. In calculating the mean total bids: (1) no differentiation
has been made with respect to the sequence that the air quality effects are
presented; (2) no differentiation has been made whether a health pamphlet has
or has not been sent to the respondent in advance of the interview; (3) no
differentiation has been made with respect to the different proposed vehicles
for the collection of bids; and (4) no differentiation has been made whether
a life table has or has not been shown to the respondent during the interview.
A life table depicts the "stock" counterparts of the elicited monthly bids for
various expected lifespans.(S) no differentiation has been made with respect to
the different dates of cleanup.
**Accept H •+ H is accepted for 1 - = = 0.90 and higher; i.e., for
= < 0.10. ° °
93
-------
will give different total bids. Ideally, the sequencing or ordering of bid-
ding information will not affect the results. In an attempt to test for
sequencing affects, two separate tests of -means were"conducted• The first
test involved a comparison by area By bid type of the mean values of the
observed bids against the derived Bids. If an assumption of additivity is
made in the bids, then we can obtain an aesthetic observed bid (from 1 above)
and a derived aesthetic bid (from 2 above). The question is then whether
the two bids differ.—' That is, does- the order in which we obtain bids affect
the magnitude for the bid. Table 4.9 presents the results of this test, ^or
aesthetic bids, El Monte (A -> B) ,±±> La Canada (A -*• C), Canoga Park, Encino,
Huntington Beach, Irvine Palos Verdes, and Redondo Beach the null hypothesis
was rejected. The null hypothesis was rejected for the acute bids in La
Canada (A -*• C), Culver City, Encino, Huntington Beach, Newport Beach, and
Palos Verdes.
The -null hypothesis was rejected for chronic bids for La Canada (A •> C)
and Newport Beach. Finally, the null hypothesis was rejected for the total
mean bids only in Newport Beach and Pacific Palisades. What can be con-
cluded from this set of results? First, the test does not completely resolve
the issue of sequencing. In some cases, the mean bids that were observed
are statistically different under the assumption of linear additivity. Sec-
ond, keeping the first point in mind, we note that the total bid does appear
to be insensitive to the bidding across different orderings of character-
istics of the environmental good air quality.
A second test to further investigate the extent of sequencing effects
is to compare each step of the bidding process irrespective of the subject
(i.e., acute or aesthetic information) of the bid. The null hypothesis is
then to compare the mean values of step 1, the mean differences in values
of step 2 from step 1, the mean difference in values from step 2 to step 3,
and the total bid. Table A.10 presents these results. For the first bid-
ding step, only Palos Verdes had the null hypothesis rejected. The null
hypothesis for the second bidding step was rejected for Pacific Palisades,
Newport Beach and Irvine. For the third bidding step only El Monte was
rejected. Finally, the null hypothesis was rejected for Pacific Palisades
and Newport Beach.—' What can we conclude about sequencing from this test?
First, again no definitive statement can be made regarding the existence or
non-existence of sequencing. The results suggest that regardless of the
information being bid upon, the step size (i.e., bid difference from the last
step) is independent of the information underlying the bid. Second, ir-
respective of the bidding order, the total Md is insensitive to orHer effects.—''
The results of the t-tests comparing the effects of different completion
dates of cleanup for each area are presented in Table 4.11. Additionally,
Table 4.12 presents similar results for each of the paired areas. The null
hypothesis of this test was that the bids are equal no matter the completion
date for the cleanup. The null hypothesis was rejected only in isolated
cases such as Canoga Park in Table 4.11. The implication of this result is
that individuals appear not to view the magnitude of their bid being signi-
ficantly determined by the proposed cleanup date.
94
-------
Table <..9
Results of the t-tests for che Equality of the Mean Bids
for Observed versus Derived Bids by Sample Area*
(Two-tail test; Pooled Variance Estimate)
H : The two mean bids are equal
o
H : The two mean bids are unequal
Name of Area
El Monte
El Monte
La Canada
La Canada
Montebello
MontebeJ Id
Canoga Park
Culver City
Enclno
Type of
Contingency
' (A - B)
(A -* C)
(A -» B)
(A - C)
(A -, B)
(A -» C)
(8 - C)
(B •* C)
(8 -» C)
-
No. of
"l
7
10
11
11
10
12
9
13 1
15
Oba.
nz
13
3
6
6
9
6
10
15
13
Aesthetic Bid
Reject H at 90X
o
Accept H
o
Accept: H
o
Reject H at 90%
o
Accept H
o
Accept H
o
Reject H at 952
o
Accept H
0
Reject H at 99%
o
Acute Health Bid
Accept H
Accept H
0
Accept H
o
Reject H at 902
0
Accept H
o
Accept H
0
Accept: H
Reject H at 90% i
o .
Reject H at 99X
0
Chronic Health Did
Accept H
o
Accept H
o
Accept H
0
Reject H at 95%
o
Accept H
0
Accept H
o
Accept ii
o
Accept H
o
Accept H
Total Bid
Accept H
o
Accept H
o
Accept H
0
Accept H
r 0
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
o
(continued)
-------
Table 4.9
(continued)
Name of Area
Huntington Beach
Irvine
Newport Beach
Pacific Palisades
Palos Verdes
Redondo Beach
. Type of
Contingency
(B •» C)
(B -» C)
(B -* C)
(C - C*)
(C * C*)
(C •* C*)
No. of
"l
16
18
14
10
11
14
Obs.
"2
22
9
6
10
8
12
Aesthetic Bid
Reject H at 90%
o
Reject H at 95%
o
Accept H
Accept il
o
Reject H at 99%
o
Acute Health Bid
Reject H at 90Z
0
Accept H
Reject H at 95*
o
Accept H
0
Reject H at 90%
o
Reject H at 95% j Accept HQ
'
Chronic Health Bid
Accept 11
o
Accept H
o
Reject H at 952
o
Accept H
o
Accept H
o
Accept II
o
Total Bid
Accept H
o
Accept 11
o
Reject H at 95%
O
Reject H at 90%
o
Accept H
o
Accept H
o
*Across the questionnaires, the effects of air quality are introduced in two different sequences: (1) Aesthetic •+ Acute Health -«•
Chronic Health; (2) Acute Health -1- Chronic Health •* Aesthetic. The bids for each effect are assumed to be separable. The purpose of
the above tests is to check if there is any significant influence of the sequence of presentation of the air quality effects on the
mean bids for each effect. For example, the mean aesthetic bid obtained by the first sequence for some area is compared with the mean
aesthetic bid obtained by the second bidding sequence for the same area. These tests of significance are repeated for each mean bid
and for each area to find out the "sequencing effect" on bids. ' In obtaining the mean bids, no differentiation is made with respect to:
(a) vehicle used; (b) starting bid; (c) health pamphlet versus no health pamphlet; and (d) life table versus no life table., (f)
pletion date of clean-up.
-------
Table 4.10
Results of the t-tests for Comparing the Sequencing Effects
in Each Step of the Bidding Process
(Two-tailed test; Pooled Variance Estimates)
H : The bids are equal
O
H : The bids are unequal
Name of Area
El Monte
La Canada
Montebello
Canoga Park
Culver City
Encino
Huntington Beach
Irvine
"1*
17
22
22
9
13
15
16
18
V
16
12
16
10
15
13
24
9
First Bid
Accept H **
0
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
0
Accept H
o
Accept H
o
Second Bid
Accept H
0
Accept H
o
Accept H
o
Accept H
0
Accept H
o
Accept H
o
Accept H
o
Reject H at 90%
o
Third Bid
Reject H at 90%
o
Accept H
O
Accept H
o
Accept H
o
Accept H
o
Accept H
0
Accept H
0
Accept H
o
Total Bid
Accept H
0
Accept H
o
Accept H
o
Accept H
0
Accept H
o
Accept H
o
Accept H
o
Accept H
0
(continued)
-------
Table 4.10
(continued)
oo
Name of Area
Newport Beach
Pacific Palisades
Palos Verdes
Redondo Beach
V
14
10
11
14
V
6
10
8
12
First Bid
Accept H
o
Accept H
o
Reject H at 95%
Accept H
o
Second Bid
Reject H at 95%
o
Reject H at 90%
0
Accept H
o
Accept H
0
Third Bid
Accept H
o
Accept H
o
Accept H
Accept H
o
i • •
Ttotal Bid
Reject H at 95%
o
Reject H at 90%
o
Accept H
o
Accept H
o
*This table presents the results of the t-tests done to determine whether or not there is a significant
difference between the means of the first bid, the difference between the first and second bids, and the
second and third bids irrespective of bidding sequence (i.e., whether the Aesthetic or Acute bid was asked
first in the questionnaire).
**n = chose questionnaires which ask Aesthetic question first, n = those questionnaires which ask. .Acute
question first.
-------
7-ihle A.11
ReaulCs of th« t-testa for Conparing the Effecta
of Different Conplet ion Dates of Cleanup
in Each Step of the Bidding Process a«b
(Two-tailed test; Pooled Variance Estimates)
H : The bida nr« of]ual.
H : The bids arc unequal.
Area
Zl Konte
El Monte
La Canada
La Canada
Hontcbello
Hontcbcl io
Canoga Park
Culver City
EDC tOO
Hontington Beach
Irvine
Nevporc Be-ich
Pacific PjJlsades
Pa Los Verdes
Redondo Beach
. Type of
Contingency
A •' B
A * C
A •* B
A -> C
A * B
A ' C
B -* C
B - C
Nunber of
Observations
V
10
7
7
10
9
11
8
16
B - C 17
B » C
B -» C
B •• C
C - C*
C -• C«
C - C*
19
15
10
8
8
14
V
9
6
10
7
10
s
11
12
11
20
12
10
12
11
12
Mean Bids
Aesthetic Bid
Accept H
Accept H
o
Accept H
o
Accept K
Accept H
o
Accept H
o
Accept H
o
Acct-'pt H
o
Accept H
Accept H
o
Accept H
0
Accu-pt H
o
Accept H
0
Accept H
0
Accept H
0
Acute Health Bid
Accept H
o
Accept H
0
Accept H
o
Accept H
o
Accept H
Q
Accept H
o
Reject H at 99Z
0
Accept H
o
Accept H
0
Accept H
o
Accept H
0
Accent H
0
Accept H
o
Accept I!
o
Accept H
o
Chronic iie.il th Bid
Accept H
o
Accept H
o
Accept. 15
o
Accept H
o
Accept H
o
Accept t!
O
Accept 11
o
Accept H
o
Ac cup I H
o
Accept H
Accept IJ
0
Accep: H
o
Accept K
O
Accept H
o
Accept H
o
Tot.il Bid
Accept H
o
Accept H
o
Accept H
o
Accept* H
o
Accept H
o
Accept H
o
Reject H at. 951
o
Accept K
0
Reluct H at 90*
o
Accept H
o . ,
Accept H
Accept H
o
Accept H
o
Accept H
o
Accept H
0
The bids for each bidding stage are assumed to be strictly separable. In obtaining the juean bids, no differentiation la Bade
with respect to: (1) bidding sequence; (2) starting bid; (3) bidding vehicle; (4) health pamphlet veraus no health pamphlet; and
(5) life table versus no life table.
b
The tests are done for « - 0.01. - - Q.05, and - " 0.10. "Accept H " Hi-ana H la accepted for « - 0.01, « - 0.05, and - - 0.10.
"fc«J«ct H at XI" aeans that H is rejected at XZ, but is accepted at the°nei!t higher I - * value; e.g., If XI • 901. then H Is
redacted for « • 0.10 but Is Accepted for « • 0.05 ond « - 0.01. • °
•o^ . aaaple tiz, Of interviews with propoted completion date of cleanup of 2 years.
*»2 " »Hple a-iM of interviews with propose* coapletiea dace et eleoaup of 10 yeare.
-------
Table 4.11
Results of the t-cesis for'Copipnring the Effects
of Different Completion D.ito.s of Cleanup
in Each rtcp of the iUJciin;; Process ai°.c
(Two-tailed test; Pooled Variance Estimates)
H : The bids are equal.
H ; The bids are unequal.
Paired Areas
El Monte-
Canoga Park
Montebelio-
Culver Citv
La Canada-
Enclno
Huntlngton Beach-
Redondo Beach
Newport Beach-
Pacific Palisades
Number of
Observations
V
25
35
34
33
18
Irvine- j 23
Palos Verdes i
V
26
30
28
32
22
23
Mean Bids
Aesthetic Bid
Accept H
o
Accept H
Accept H
r 0
Accept H
o
Accept H
o
Accept H
0
Acute Health Bid
Accept I!
0
Accept HQ
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Chronic Heal th Bid
Accept H
o
Accept 11
0
Accept H
o
Accept H
o
Accept K
o
Accept H
o
Total Bid
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
o
Accept H
0
Aggregate Data Set j 169
161
Accept H
o
Accept H
0
Accept H
0
Accept H
0
In obtaining the results* the total number of Interviews in each paired area is divided into two parts with respect to
their proposed completion dates of cleanup (i.e., 2 years versus 10 years) and the t-testp are done to test whether this
difference has any significant influence on the mean bids.
The tests are done for « - 0.01, " - 0.05, and « - 0.10. "Accept H " means U is accepted for all three of the
« levels. ° °
The bids for each bidding stage are assu.-r.ed to be strictly separable. In obtaining the Bean bids no differentiation la
made with respect to: (1) bidding sequence; '(2) starting bid; (3) bidding vehicle; (4) health pamphlet versus no health
pamphlet; and (5) life table versus no life table.
*n - sample size of interviews with proposed completion date of cleanup of 2 years.
*n - eample alze of interviews with proposed completion data of cleanup of 10 y&ara.
-------
The previous results presented rely in many cases on small sample sizes
for the statistical tests due to the partitioning required. Recall that
several types of bias as well as game structure questions had to be examined.
In view of the small sample sizes for a few areas, additional questionnaires
were administered. Table 4.13 presents the mean Bid results of these ad-
ditional interviews. Before integrating into the basic data set, it was
decided to test whether the "new" data was significantly different from the
"old" data in terms of mean values. Results of the tests are presented in
Table 4.14. Culver City for the total bid category is the only significantly
different result from the "old" data set. This is due to one of the indiv-
iduals bidding an exceptionally larger sum than others as noted in the foot-
note in Table 4.13.
101
-------
Table 4.13
Mean Bids by Area by Type for the "New" Data Set*
Area
Canoga Park (B->-C)
Canoga Park (B-vC)
Culver City (E->C)
Encino (B->C)
Encino (B->£)
Pacific Palisades
(C->C*)
Pacific Palisades
(C-vC*)
Completion
Date of
Cleanup
2 years
10 years
10 years
2 years
10 years
2 years
10 years
Aesthetic B
3.20
(1.29)**
(10)***
3.00
(2.00)
(5)
6.25
(6.25)
(4)
6.67
(6.67)
(3)
4.17
(1.54)
(6)
6.00
(3.70)
(6)
16.67
(23.58)
(9)
Mean-
Acute
H.B.
9.18
(2.12)
(10)
9.00
(2.92)
(5)
27.50
(8.29)
(4)
13.33
(7.26)
(3)
6.67
(3.07)
(6)
9.17
(2.71)
(6)
12.89
(5.75)
(9)
Bid ($ /Month)
Chronic
H.B.
8.00
(3.27)-
(10)
10.00
(5.70)
(5)
28,75
(16.25)
(4)
1.67
(1.67)
(3)
8.33
(6.41)
(6)
6.67
(2.98)
(6)
6.67
(2.89)
(9)
Total
Bid
20.30
(3.41)
(10)
22.00
(8.15)
(5)
62.50****
(21.07)
(4)
21.67
(12.02)
(3)
19.17
(6.76)
(6)
22.67
(6.31)
(6)
35.11
(13.35)
(9)
*The bids for each bidding stage are assumed to be strictly separable. In
obtaining the mean bids no differentiation has been made with respect to 1)
bidding sequence, 2) starting bid, 3) bidding vehicle, 4) health pamphlet
versus no health pamphlet, and 5) life table versus no life table. A life
table depicts the "stock" counterparts of the elicited monthly bids for
various expected life spans.
**Standard error of the mean bid in all cases.
***Sample size of each case in all cases.
****Individual total bids for Culver City were as follows:
I: $100
II: $ 75
III: $ 25
IV: $ 50
Aes.
0
0
0
25
Ac.
25
50
10
25
Ch.
75
25
15
0
102
-------
Table A.IA
Results of the t-tests for Comparing the Equality of the Mean Bids Obtained from
the "Old" and the "New" Data Sets in Each Step of the Bidding Process3'
(Two-tailed test; Pooled Variance Estimate)
H : The bids are equal.
H-' : The bids are unequal.
Completion
Type of Date of
Area Contingency Cleanup
Canoga Park
Canoga Park
Encino
Encino
Pacific Palisades
Pacific Palisades
Culver City
B-K:
B-K:
B-K:
B-K:
C-K:*
C-K:*
B-vC
2
10
2
10
2
10
10
years
years
years
years
years
years
years
Number of
Observations
Nl N2
8
11
17
11
8
12
12
10
5
3
6
6
9
A
Mean Bid
Aesthetic
Bid
Accept
Accept
Accept
Accept
Accept
Accept
Accept
Ho
Ho
Ho
Ho
Ho
Ho
Ho
Acute Health Bid
Accept
Reject
Accept
Accept
Reject
Accept
Reject
"o
HQ at 95%
Ho
Ho
H at 95%
H0
H at 95%
Chronic
Health Bid
Accept HQ
Accept H
Accept H
Accept H
Accept H
Accept 1!
Accept H
Total Bid
Accept
Reject
Accept
Accept
Accept
Accept
Accept
Ho
HQ at 90%
Ho
Ho
Ho
Ho
Ho
The bids for each bidding stage, are assumed to be strictly separable. In obtaining the mean bids, no differentiation
is made with respect to: 1) bidding sequence, 2) starting bid, 3) bidding vehicle, A) health pamphlet versus no health
pamphlet, and 5) life table versus no life table. The "old" and "new" data sets are a result of additional inter-
viewing that was carried out to supplement the sample size in a few areas.
O.OS but is accepted for <* = 0.01.
The tests are done for = = 0.01, <* = 0.05, and « = 0.10. "Accept H " means H is accepted for <* = 0.01, « = 0.05,
and = = 0.10. "Reject H at X%" means that H is rejected at X2 , but is accepted at the next higher 1-= value;
e.g.., if X% = 95%, then H is rejected for « = 0.10, and
N-: Sample size of interviews from the "old" data set.
N^: Sample size of interviews from the "new" data set.
-------
FOOTNOTES; CHAPTER_IV
— Some questions have been raised about employing a "typical week." as
the time period upon which to base, the analysis. However, in a previous
study valuing wildlife, various time units- were employed (last trip, typical,
dairy formats) for hunting experience. No statistical difference was found
in the activity responses. See BrooRsfiire, Randall, et. al. (1977).
21
— Note when the acute initiation point was employed the picture sets
were not made available until the bidding on the aesthetic characteristic
was begun.
3/
— Note holding the respondent to the original time constraint for their
current situation implies no leisure-work tradeoff possibilities. Arguments
can be'presented for or against this assumption. However, we note that in
Blank, et. al. (1977) that when this tradeoff was offered as part of the
substitution format, few respondents did make the trade.
A/
— In a few cases the respondents bid their maximum willingness to pay
initially rather than on the characteristic points.
— The equipment used was a 135 mm Minolta single lens reflex camera
(SLR), 135 mm telephoto lens, 55 mm lens, tripod, Kodak Vericolor II profes-
sional color film, log sheets and a small ice chest.
— It should Be noted that film is affected a great deal by even mild
variations in temperature, especially heat. Thus, it is critical for the
film to be protected before, during, and after use. The film used for our
study was kept in a small portable ice chest until used; once the roll or
film was used, it was put back into its air tight container and then back
into the cooler until ready for processing. As a further aid in protecting
the film, the researchers used rolls of twenty rather than 36 frames in an
effort to minimize the time the film was exposed to heat.
— Appendix E details the exact streets in the sample plan.
o /
— Note the Pacific Palisades bid is for a C -> C* contingency, thus
implicitly employing a bid for a basin-wide improvement, not involving this
location directly.
9/
~~ Further examination of vehicle bias will require breaking out
aesthetic observed versus aesthetic derived bids and conducting a t-test.
—- Note this applies to acute and aesthetic bids, however, the chronic
bids are entirely derived.
— The (A -»- B) and (A -> C) notation refers to type of contingency moves
for residents of the A (poor) air quality area. The principle reason for
administering two types was- to avoid an overly long survey instrument.
104
-------
12/
— This is the identical result noted in the first sequencing test
which.by the structure of the tests must be the case.
— A follow-up on this thesis would be a te
against total for step 2 against total for step 3.
— A follow-up on this thesis would be a test of the total step size 1
105
-------
Chapter V
THE SOUTH COAST PROPERTY VALUE STUDY
5.1 Overview
Many different methodologies for valuing non-market or environmental
goods and services have been proposed. However, none of these approaches
is universally accepted and debate remains over which methodology is most
appropriate. New valuatioq methods such as the contingent valuation approach,
are marked by uncertainty and criticism from both professional and r.on-
professional audiences, and thus require replication and evidence of
internal consistency in order tp demonstrate validity
The purpose of the research on property values presented here is to
provide the necessary comparison for the contingent valuation approach
which is described in detail above. This is accomplished through an
analysis of the housing market within the sample plan communities of the
South Coast Air Basin located in Los Angeles and Orange Counties. Specifi-
cally, this research asks if households will actually pay for cleaner air
in the form of higher property values for homes in clean air communities
and if this willingness to pay is comparable to the hypothetical willingness
to pay expressed in the survey instrument.
Valuation of reductions in urban air pollution concentrations based
upon housing value differentials is the most common form of the hedonic
price procedure as developed by Rosen (1974), the basis of which is Lan-
caster's (1966) consumption theory. This procedure assumes that access
to environmental (dis)amenities is capitalized in property values. This
assumption is based on the premise that households are willing to pay a
premium for an otherwise identical home located in a clean air area versus
that located in a polluted area. The capitalization can be. discovered by
isolating the impact of air quality in two alternative ways: (1) by
developing a sample pairing system which minimizes the variation in housing
and community variables other than air quality and comparing housing values;
or (2) by regressing housing value data on air quality and other variables.
In the latter method, the resulting empirical relationship is the basis for
a determination of the value of the environmental good.
Previous housing value studies have concentrated on the regression
procedure. The first significant empirical study of air quality and pro-
perty values was done by Ridker and Henning (1967), The authors applied a
least squares regression model to cross-sectional data (compiled by census
tract) for the St. Louis area. In order to fully specify their model of
property values, variables corresponding to housing, location, neighborhood,
106
-------
political jurisdiction and individual characteristics were included with air
pollution measures as independent variables. A significant negative rela-
tionship was found between the. sulfation level (annual geometric mean) and
median property values. Further, a property value increase of between $83
and $245 was associated x^ith each .25/mg./100cm2/day reduction in the
sulfation level. This translates into total benefits of approximately $83
million if sulfation levels are reduced by .25 nig., or to an ambient
concentration level of .49 mg.
This research was followed by a similar study by Anderson and Crocker
(1971) who analyzed the impact of air pollution on both renter and owner
occupied properties for St. Louis, Kansas City and Washington, B.C. As in
the Ridker-Henning work, the basic unit of observation was the census tract and
cross-sectional data which was employed in a non-linear regression model. The
Anderson-Crocker results confirmed the Ridker-Henning finding of a negative
and statistically significant relationship between air pollution (annual
arithmetic mean concentrations of sulfur oxide and suspended particulates)
and property values. The same result was also found for rental property.
Deyak and Smith (1978), in an effort to generalize these empirical
conclusions utilized an updated data base (1970 census) gathered from
representative SMSA's. Their results provided added support for the findings
of Ridker-Henning and Anderson-Crocker. However, in andtaer paper, Smith-
Keyak (1975), using data on owner and renter occupied central city housing
in eighty-five cities, which also formulated a residential location model
that included location pubiic services and tax effects, found that air
quality did not significantly affect property values. This conclusion was
in accordance with the results found by Steele (1972) and later" Wieand
(1978). Both authors found no statistically significant relationship to
exist between air pollution and property values. The Wieand findings are
especially surprising since he employed essentially the same data base as
Ridker-Henning. The major change was substituting monthly rent per acre
•in place of median property value as the dependent variable.
These results indicate that an analysis of housing markets can yield
information on the value of non-market goods. However, they also demon-
strate the fragility of the methodology. That is, all assumptions outlined
in Chapter II must be met and extreme care is required in model specification
and interpretation of the results.
The analysis undertaken here encompasses three separate but related
approaches, with benefits from reduced air pollution in bid equivalent terms
(e.g., terms comparable to the contingent valuation results) specified at
each level. The first approach involves a straightforward comparison of
average housing values in the sample pair communities, standardizing only
for house size (square feet of living area). The resulting differential
in sale price between paired communities, which are theoretically identical
except with respect to ambient air pollution concentrations, is then
attributed to the disparity in air quality. It should be noted that this
methodology relies quite heavily on the successful operation of the sample
plan. That is, the variation across pairs in all other housing and neigh-
borhood characteristics (excepting air quality) must be minimal if the sale
price differential assigned to air pollution is to reflect accurately
107
-------
individuals true willingness to pay for clean air.
In the second procedure we utilize an econometric estimate of the
impact of air quality on housing values to determine benefits of reduced
air pollution. This portion of the study corresponds to the traditional
econometric analysis of the housing market and is an attempt to estimate
a linear relationship between a home's sale price and its supply of housing
and community attributes. The value of an improvement in air quality is
then deduced from the resulting hedonic housing value equation.
The final approach is a further refinement of the above methods and
consists of a multi-step procedure which makes allowance for air pollution
abatement to be valued differently by households with varying income and
initial pollutant concentrations. This methodology was developed recently
in a paper by Harrison and Rubinfeld (1978). The first step is to estimate
a hedonic housing value equation, similar to the second approach, but
allowing 'for non-linearities where appropriate in the functional form. The
second step is to calculate the marginal willingness to pay for individuals
in each of the sample communities for a small change in air quality. The
third step is to estimate a marginal willingness to pay equation as a
function of income and other houshold variables. By integrating individual
marginal willingness to pay estimates, we at least partially overcome the
problems pointed out in Section 2.1. Finally, we employ this latter rela-
tionship to determine benefits of air quality improvements.
Each of the three appro-aches as described above can be viewed as a
part of a systematic analysis of housing market data in the communities
which comprise the sample plan. Further, each procedure yields pollution
abatement benefit estimates which can be used to compare to the contingent
valuation experiment. In addition to its usefulness as a comparability
exercise, this housing value analysis has advantages over previous studies
in that data is drawn as part of the sample plan which by its nature con-
trols for many exogenous factors not wholly explained in the standard treatment.
This, for example, tends to explain why our statistical "fit" is superior
to previous studies. However, it should be kept in mind that sampling
is therefore appropriate for comparison to the contingent validation
experiment but is non-random and may lead to biased estimates of basin-wide
damages.
The remainder of this chapter is organized as follows. Section 5.2
describes the data base and sources utilized in the study. In Section 5.3,
the three appraoches and their associated results are presented. Section
5.A concludes'the analysis.
5.2 Data Characteristics
The area under investigation is defined as the South Coast Air Basin.
However, this study utilizes data for the sample plan communities only (see
Chapter IV). In this regard, the sample chosen for study is not entirely
random but rather a function of a pre-testing scheme. This may not Be a
major restriction on either the methodology or results since the paired
communities are representative of the entire spectrum of living conditions
in Los Angeles and Orange Counties^ It sfiould also be noted that this study
103
-------
is restricted to single family residences and the results are therefore
only possibly applicable to other housing types (niohile homes s apartments,
condominiums). Further, we concentrate on the owner market to the exclusion
of other markets (rental, leasing, etc.).
Focusing upon the paired communities then, the data base was constructed
to enable the impact of air quality differentials on housing sale price to
be isolated. Thus, the dependent variable in the analysis is the sale price
of owner occupied single family residences. The independent variable set
consists of variables which correspond to three levels of aggregation:
house, neighborhood, and community. The data base contains 719 independent
observations. Table 5,1 describes further the data employed in the study.
The housing characteristic data, obtained from the Market Data Center
(a computerized appraisal service centered in Los Angeles), pertains to
homes sold in the January, 1977 - March, 1978 time period and contains
information on nearly every important structural and/or quality attribute.
Table 5.2 provides summary statistics for many of the housing characteristic
variables for each of the sample communities. It should be emphasized that
housing data of such quality (e.g., micro level of detail) is rarely
available for studies of this nature. Usually outdated data which is overly
aggregate (for instance census tract averages) is employed. These data
yield functions are relevant for the "census tract" household and are only
marginally relevant at the micro level. However, in this study it was
imperative that data comparable to that used in the contingent valuation
experiement be utilized. That is, since pollution abatement benefit estimates
were calculated at the household level in the contingent study, it was
necessary to generate similar estimates based on comparable data in this
validation exercise.
In addition to the immediate characteristics of a home, other variables
which significantly affect its sale price are those that reflect the
condition of the neighborhood and community in which it is located. That
is, the local tax and public goods expenditure rates, school quality, ethnic
composition, crime rates, proximity to employment centers (and in the South
Coast Air Basin, distance to the beach), and measures of the ambient air
quality which have a substantial impact on sale price. Therefore, in order
to capture these impacts and to isolate the independent influence of air
quality, these variables are included in the econometric modeling.
The measures of air quality used in the empirical analysis were obtained
from California Air Resources Board publications (1977). This agency is
responsible for monitoring pollution levels in the Basin. The South Coast
Air Shed, because of the existence of a large number of both mobile and
stationary sources combined with meteorological and topographical conditions
which limit the area's ability to disperse pollutants, has a long history of
pollution problems. A relatively complete regional network of monitoring
stations has been developed. This allows the measurement of ambient air
pollution levels rather than concentrations on isolated hotspots. A detailed
exposition of air pollutants by area was given in Chapter III.
In conclusion, we view the data b'ase ass-embled for the housing value
study as appropriate for comparability testing of the contingent valuation
109
-------
Table 5.1
Variables Used in Analysis of Housing Market
Variable
Definition (assumed effect on
housing sale price)
Units
Source
Dependent
Sale Price
Independent-Housing
Sale Date
Age
Bathrooms
Living Area
Pool
Fireplaces
Independent-Neighborhood
Distance to Beach
Sale price of owner occupied single
family residences.
Month in which the home was sold
(positive indicator of inflation)
Age of home (negative indicator
of obsolence and quality of
structure)
Number of bathrooms (positive
indicator of quality)
Living .area (positive indicator
of the quantity of home)
Zero-one variable which indicates
the presence of a pool (positive
indicator of quality)
Number of fireplaces (positive
indicator of quality)
Distance to the nearest beach
(negative indicator of relative
proximity to main recreational
activity)
($1,000)
January 1977=1
March 1978-15
Years
Number
Square feec
Zero=no pool
One=pool
Number
Miles
Market Data
Center
Market Data
Center
Market Data
Center
Market Data
Center
Market Data
Center
Market Data
Center
Market Data
Center
Calculated
(continued)
-------
Table 5.1
(continued)
Variable
Definition (assumed effect on
housing sale price)
Units
Source
School Quality
Ethnic
Population Density
Housing Density
Distance to
Employment
TSP
School quality as measured by student
percentile scores on the California
Assessment Test-12th grade math
Ethnic composition-percent white in
census tract(s) which contain sample
community (positive).
Population density in surrounding
census tract (negative indicator
of crowding)
Housing density in surrounding
census tract (negative indicator
of crowding)
Weighted distances to eight
employment centers in the South
Coast Air Basin (negative indicator
of proximity to employment)
Nitrogen dioxide concentrations
Concentrations of total suspended
particulates
Percentile *100
Percent *100
Local School files
in sample communities
1970 Cfnsus
People per square 1970 Cenfus
mile
Houses per square 1970 Census
mile
Miles/Employment Calculated
Density
Parts per hundred California Air
million (pphm) Resources Board
Micrograms per ^ California Air
cubic inecer (iig/m ) Resources Board
(continued)
-------
Table 5.1
(continued)
Variable
Definition (assumed effect on
housing sale price)
Units
Source
Independent-Community
Public Safety
Expenditures
Crime
Tax
H-1
vo
Expenditures on public safety per $/People
capita (positive indicator of
attempt to stop criminal activity)
Local crime rates (negative indicator Crime/People
of peoples' perception of danger)
Community tax rate (negative measures $/$l,000 of
cost of local public services) home value
1976-77 Annual
Report Financial
Transactions
Concerning Cities
of California
FBI (1976)
1976-77 Annual
Report Financial
Transactions
Concerning Cities
of California
-------
Table 5.2
Average Housing Characteristics
City
Canoga Park
El Monte
Culver City
Montebello
Orange
Whittier
Redondo Beach
Huntington Beach
Pacific Palisades
Newport Beach
Palos Verdes
Irvine
Encino
La Canada
Population
Sale
Price ($)
43,914
34,273
82,916
63,957
70,368
67,647
64,817
77,214
257,383
141,473
165,016
83,054
209,158
153,804
99,719
Sale Price/
Sq. Ft.
40.299
32.1
58.03
43.48
46.89
41.64
58.6
53.239
91.05
68.5
64.98
50.97
70.95
59.91
58.07
Living Area
(Sq. Ft.)
1089.68
1067.68
1428.73
1470.95
1500.58
1624.67
1104.18
1450.32
2826.67
2065.41
2539.44
1629.49
2947.84
2567.17
1717.1
Number of
Bathrooms
1.16
1.18
1.54
1.67
1.98
1.64
1.28
1.95
3.14
2.43
2.72
2.13
3.04
2.6
1.99
Number of
Fireplaces
.227
.18
.56
.67
.73
.95
.30
. 71
1.78
1.20
1.24
.95
1.44
1.45
.86
Age of Home
(Years)
32.9
35.6
26.1
28.1
16.4
32.3
27.0
13.3
28.1
15.5
13.5
4.4
16.2
33.3
19.2
The property value study includes two more communities in the data base then did the contingent
valuation study: Orange and Whittier.
-------
experiments. The reasons are threefold. First, the housing characteristic
data is extremely detailed at the household level of aggregation, and
extensive in that a relatively large number of observations are considered.
Second, we have assembled a variety of neighborhoods and community variables
which enable the isolation of the air quality influence on housing values.
Third, the air pollution data is comprehensivei
5.3 Empirical Analysis
As outlined in the introduction, each of the three stages of empirical
analysis undertaken in this study constitutes a separate attempt to capture.
the monetary impact which air quality differentials have on housing values.
Once discovered, these monetary estimates of the air quality effects are
translated into the value of improving air quality in the South Coast Air
Basin. These calculations are later utilized to test the validity of the
contingent valuation experiment.
The following household benefit equation is used and shows the inter-
relationships or common characteristics of the three approaches;
N N
AHB = [( E AQI.*NH *AD )/ £ NH ]*CRF (5.1)
1=1 X 1=1
where
AHB = average annual benefits per household for a reduction in air
pollution concentration.
AQI = air quality improvement in area i (poor-fair, fair-good).
NH = number of homes in area i affected by the air quality change.
AD = average home sale price differential attributed to a one unit
improvement in air quality.
CRF = capital recovery factor. This is the rate necessary to transform
an initial capital investment into a series of equivalent annual
charges including payment of both capital and interest. In this
study the CRF is assumed to be .0995 which corresponds to a
.0925 interest rate and a payback period of 30 years.
N = number of specific areas affected by air quality improvement. In
this study N is restricted to two as benefits are calculated for
upgrading the air quality in the poor areas to fair and in the
fair areas to good (see Chapter III).
The number of homes in the affected areas (see Table 5.3) and the air
quality improvement are common to each of the three methodologies. Table
5.4 illustrates both the present air quality classifications-, the reductions
in nitrogen dioxide (NO ), and total suspended particulatea (TSP) which are
required to achieve significant improvement as measured by the relative
quality indicators. This analysis was not able to effectively separate out
114
-------
Table 5.3 , . _
Number of Homes in South Coast Air Basin
bV Air Quality Categories
Number of Homes in South Coast Air Basin
Air Quality (Los Angeles and Orange Counties)
Poor 1,056,325
Fair 804,823
Good • • 228,772
Total 2,089,920
Table 5.4
Air Quality Definition - NO
Air Quality
(Arithmetic Average 1975 - pphm) Classification
> 11 Poor
9-11 Fair
< 9 Good
12.38 ,,„. Average Pocr
9CC^-^™ 1 r- •
.55 Average Fair
6.9 Average Good
Air Quality Definition - TSP
Air Quality
(Arithmetic Average 1975 - yg/m ) Classification
> HO Poor
90-100 Fair
< 90 Good
118.4 > ^-- Average Poor
100.0 ^. Average Fair
78.8 Average Good
115
-------
the independent effects of each, of the pollutants due to collinearity in the
data set. Therefore, all calculations employ only one. pollutant measure as a
proxy for the general air quality situation. Nitrogen dioxide is usually
that variable, however, resulta are also presented for TSP for purposes of
comparison.
The final element of equation (5.1), the sale price differential
attributable to differences in ambient air quality (which can be inter-
preted in this context as the willingness to pay for improved air quality)
is determined by the particular approach. The first method determines this
parameter in the simplest manner, through comparison of sale prices in the
paired communities, and relies almost exclusively on the sample plan. The
second is intermediate in complexity and employs traditional property value
analysis to determine the monetary effect of air quality differentials. The
third approach is the most involved, attempting to account for variation in
preferences in the determination of willingness to pay through statistical
means-. In this manner the air quality impact on individuals is explicitly
specified. These methods are addressed in detail in the following three
subsections.
Before discussing in detail the empirical investigation and correspond-
ing results, a few notes pertaining to the theoretical underpinnings of the
analysis are in order. First, the capitalization of environmental goods
into housing values can be captured through such empirical work only if
certain assumptions concerning the economic behavior of individuals and
the functioning of the housing market are accepted. These are: (1) con-
sumers must perceive differences in housing and neighborhood characteristics,
expect them to remain unchanged and act on these perceptions; (2) housing
markets should function reasonably well and be in short run equilibrium;
(3) environmental quality must be exogenously determined and differences in
environmental quality must be capitalized only in housing prices; and (4)
all relevant hedonic price functions should be continuous with first and
second derivatives that exist (e.g., there must be sufficient variation in.
both housing and neighborhood characteristics, including air quality, to.
permit, continuity). Second, it should be noted that this housing market
analysis is consistent with and indeed a substudy within the general
theoretical treatment developed in detail above.
Paired Sample Methodology
The paired sample approach is limited by the ability of the sample plan
to pair communities which are virtually identical in every respect including
air quality. No explicit effort is made to account for differences in sale
price induced by other housing or neighborhood characteristics. This is
admittedly a naive approach, yet it produces two positive outputs. First,
because this methodology only implicitly compensates for many home and
community variables, the resulting benefit estimates can be considered an
upper bound on the population's willingness to pay for reduced air pollution.
Second, if these benefit numbers closely parallel those of the other, more
refined econometric methods can be considered successful.
Table 5.5 presents average sale prices in each of the sample communities
and air quality regions„ These figures are standardized for house size with
116
-------
Table 5.5
Sale Price Differentials Attributed to Air Quality
Paired Sample Methodology
Community
Canoga Park
El Monte
Culver City
Montebello
Orange
Whittier
Enclno
La Canada
Redondo Beach
Huntington Beach
Pacific Palisades
Newport Beach
Palos Verdes
Irvine
Air
Quality
Fair
Poor
Fair
Poor
Fair
Poor
Fair
Poor
Good
Fair
Good
Fair
Good
Fair
Avg. Home
Sale Price -
1717 Sq. Ft.
Home
69,193
55,116
99,645
74,655
80,516
71,491
121,826
102,868
100,790
91,412
156,342
117,608
111,573
87,622
Sale Price
Differential
«)
25.5
33.5
12.6
18.4
10.3
32.9
27.3
Avg. Sale Price
Differential
Poor-Fair, Fair-
Good ($) (%)
19,371 (25%)
27,498 .(28%)
117
-------
each assigned the population average 1717 square foot home. The illustrated
sale price differentials are calculated utilizing the community with the
poorer air quality as- the base. Further, the figures indicate that the value
associated with an improvement in air quality from poor to fair is approxi-
mately 25% of the average poor community home price (for the same sized
home). A similar upgrade in classification from fair to good is valued at
about 28% of the value of the average fair region home. Translated into
monetary terms these figures represent approximately $19,000 and $27,000
per home, respectively.
As all other components were specified previously, these price figures
complete the information required to compute annual household benefits using
the paired sample approach. These benefits are presented in Table 5.6. As
illustrated, the capitalized benefits are in excess of $39 billion for an
approximate 25-30 percent reduction in urban air pollution concentrations
(poor-fair, fair-good). This translates into nearly $4 billion in annual
terms' evaluated at 9.95 percent capital recovery factor, which is equivalent
to how much individuals would be willing to pay for cleaner air in the form
of higher house payments. In order to transform these figures into bid
equivalent terms, they are weighted by the total number of affected homes
and the days in a year. Thus, based on the paired sample methodology, each
household is willing to pay $4.50/day or $135 per month for the stated air
quality improvement. Although it is expected that this method produces
high benefit estimates, the above figure seems a reasonable amount when one
considers the variety of impacts (health, aesthetic, etc.) associated with
deteriorated air quality.
Although these willingness to pay figures seem interesting and
reasonable, this methodology possesses a number of obvious shortcomings
which may negate their significance. These can be classified as follows,
First, this 'methodology attributes the entire differential in average sale
price to the variation in air quality. This explicitly neglects a variety
of other possible differences which could account for the disparity in sale
price (although at least partial compensation for these factors is incor-
porated in the sample plan). That is, this approachs at least to some
extent, is using air quality to proxy for many relevant neighborhood and
community variables. Isolation of the independent influence of air quality
may not be complete.
The second problem with this methodology is that each household,
regardless of its characteristics, is assumed to place an identical value
on the reduction in air pollution. Thus no allowance is made for household
differences which would imply varying valuations. In the next subsection
we employ traditional property value analysis and attempt to solve the first
category of shortcomings. However, the latter problem is not effectively
addressed until the following subsection.
Econometric Approach - Linear Equation
The underlying structure of the econometric approach, is a single
equation empirical model which purports to explain the variation in sale
prices of homes located in the South Coast Air Basin. The basic operational
118
-------
Table 5. 6
Benefits - Paired Sample Methodology
Change In Air Quality
Capitalized Benefits
(Billion Dollars)
Annualized Benefits
(Billion Dollars)
R = .0925, CRF = .0995
Poor to Fair
Fair to Good
20.46
19.34
2.04
1.92
Total
39.8
3.96
Capitalized Benefits
($)
Annualized Benefits ($)
R = .0925, CFJ- = .0995
Per Home
Per Home Per Day
Per Home Per Month
21,385
2.30
4.50
135.00
119
-------
tool is ordinary least squares (OLS). The procedure is to regress the full
independent variable set (see Table 5.1 for a complete description of these
variables and their expected relationship to housing price) on the vector of
home s-ale prices. The result of this econometric analysis of housing
market data is a hedonic housing value eq-uation. The estimated coefficients
of such an equation provide information on the relative significance and
value of each of the independent variables. That is, the coefficients
specify the effect that a unit change in a particular independent variable
has on sale price.
In reference to the air quality variable, this procedure allows one to
focus on its impact while separating out the influence of other extraneous
variables. Therefore, this analysis- yields two outputs concerning the
relationship of air quality differentials to housing price. First, the
relative significance of air quality variations is determined and second,
the estimated coefficient pertaining to air quality implicitly measures its
monetary value.
The initial objective then is to estimate a linear hedonic housing
equation which best fits the data. However, there exist a number of
empirical problems which could prevent efficient estimation of the desired
relationship. For instance, two problems which generally arise in property
value studies are misspecification bias (.the independent variable set is
incorrectly specified) and multicollinearity (members of the independent
variable set are highly correlated). Either of these may produce biased
estimates. Furthermore, it is essential that these biases be avoided since
the estimated coefficients become the basis for the benefit calculations.
Misspecification can be adequately countered by including in the equa-
tion all relevant independent variables without including variables which
have no a_ priori (on theoretical grounds) relationship with the dependent
variable. The data set used in this study is relatively complete, in that
it contains a large number of housing and neighborhood characteristics.
Further discussion of this subject, however, is postponed until the next
section where experiements which demonstrate the effect of specification
error are performed.
Multicollinearity is a common problem in studies of this nature.
This occurs since many of the independent variables are integrally linked.
and therefore possess extremely high correlation coefficients. For instance,
with respect to housing characteristics, living area, number of rooms,
number of bedrooms, etc.; they are so interconnected (each representing size
of home) that least squares estimation techniques cannot determine the
independent impacts that these variables have on housing values. Therefore,
living area was chosen as the proxy variable for house size. Note that
home quality is measured by the inclusion of fireplaces, pools, and number
of bathrooms.
Similarly, the air pollution variables showed a high degree of correla-
tion. Again the estimation procedures were unsuccessful in separating out
the independent influence of each of the pollutants. Thus only one pollution
variable, usually NO was utilized as a proxy for the general state of air
120
-------
variables and accessibility to beaches. However, this collinear relation-
ship was effectively broken fay the structure of the sample pairings. Thus,
the simple correlations between NO,,, TSP, and distance to beach do not ex-
ceed .37.
Our concern about multicollinearity among the neighborhood and com-
munity variables was also warranted when it was found that housing density
and population density were so highly correlated (.96 simple correlation)
that they were essentially measuring the same phenomena. The solution was
to allow only one of these variables in any equation. These empirical
problems aside, we next proceed to discuss the estimated hedonic housing
value equations. However, it should be re-emphasized that the estimation
was accomplished within the bounds of these empirical difficulties.
The equations which provided the best statistical fit of the data are
presented in Table 5.7. The relatively high R^'s (: .83) indicate that a
large proportion of the variation in housing price is explained by the
variation in the independent variables. Except for two aberrations all
coefficients possess the expected sign and are statistically significant.
The exceptions are age, which is positive related to house value and
significant, and local tax rates which have the anticipated relationship
but are statistically insignificant. The former may occur since age may
not be an adequate measure of housing condition since many older homes in
the Los Angeles are of high quality. The insignificance of local tax rates
seems puzzling. However, this is probably a result of the linear functional
form since in the next subsections we find that taxes become significant
when non-linearities are introduced. Furthermore, the age variable assumes
the proper relationship in the non-linear equations.
Further examination of Table 5.7 gives added insight into the deter-
minants of house value. The air pollution variables both perform as expected
and are highly significant. In addition, the coefficients on the pollution
variables are quite similar (-316.89 for TSP and -260.4 for N02 when N02 is
converted to ug/nP units) signifying that each, as a proxy for pollution,
has a similar impact on house price. The stability of the coefficients on
the non-pollution variables (they are virtually identical) is also striking.
This result suggests that households are averse to pollution generally
rather than to any single pollutant.
The quantitative significance of a unit change in any of the independent
variables is obtained by examining the coefficient values. For instance,
an increase of 100 square feet of living area would cause a $2866.8 increase
in the house price. Likewise, the coefficient on sale date shows that sale
prices were'increasing by nearly $l,900/month over the study period.
Employing this same type of analysis, benefits from a reduction in either
N02 or TSP can be calculated. Therefore, using N02 as the proxy, an im-
provement in air quality from poor to fair infers capitalized benefits of
$14,445/home while a change from fair to good is valued at $13,526/home.
As in the previous methodology, these figures together with data previously
generated (number of affected homes, etc.) become the basis for calculating
average annual benefits [see equation (5.1)]. These benefit computations
are completed in Table 5.8 for both N02 and TSP (in parentheses).
121
-------
Table 5.7
Estimated Econometric Equations (Linear)*
Dependent Variable = Home Sale Price in Dollars
Independent Variable
Sale Date
Age
Living Area
Bathrooms
Pool
Fireplaces
Distance to Beach
Distance to Employment
Crime
School Quality
Ethnic Composition
Housing Density
Tax
Public Safety Expenditure
TSP
N°2
Constant
R2
Sum of Squared Residuals
Degrees of Freedom
NO Equation
1897.8
(7.0041)
313.3
(2.8236)
28.665
(13.516)
21.856
(9.2552)
10213
(3.216)
14107
(7.1613)
-436.55
(-.19769)
-22597.
(-9.635)
-564090.
(-2.7727)
208.91
(2.7353)
4178.3
(2.7697)
-5.5248
(-1.9503)
-8.7207
(-.68288)
59.189
(6.7578)
-5104.3
(-4.8851)
-324820
(-2.2395)
.832
496900
703
TSP Equation
1944.6
(7.0946)
192.41
(1.7823)
28.558
(13.272)
22.378
(9.3117)
11375
(3.5566)
13187
(6.6648)
761.27
(-3.4148)
-18370
(-6.3776)
-674680
(-3.0476)
171.94
(2.1777)
7442.9
(5.5327)
-7.9192
(-2.6061)
-3.0441
(-.22507)
56.278
(5.6769)
-316.89
(-2.7845)
-
-652150
(-5.0917)
.828
508200
703
*t-statistics are in parentheses.
122
-------
Table 5.8
Benefits r Linear Econometric Methodology
NO (TSP)
Change- in Air Quality
Capitalized Benefits
(Billion Dollars)
Annualized Benefits
(Billion Dollars)
R = .0925, CRF = .0995
Poor to Fair
Fair to Good
15.3 (6.2)
10.9 (5.4)
1.52 (.61)
1.08 (.54)
Total
26.2 (11.6)
2.6 (1.15)
Capitalized Benefits
($)
Annualized Benefits ($)
R = .0925, CRF = .0995
Per Home
Per Home Per Day
Per Home Per Month
14077. (6233)
1401. (620)
3.84 (1.70)
115.20. (51-.0)
123
-------
The benefit figures dictate discussion on two counts. First, the large
discrepancy between the N02 based and TSP based benefits is a result not of
the respective regression coefficients but rather from the fact that present
NO- concentrations are much higher and demonstrate greater variability than
those for TSP. For instance, the average poor community has an ambient NC^
level of 242.7 yg/nr whereas the TSP concentration for a similar community
is 118.4 pg/m-\ Also, the gap between the quality indicators (poor-fair,
fair-poor) and therefore the required improvement is much greater for NC^
than TSP in percentage terms. Thus, since TSP concentrations are both
lower and more ubiquitous than N0~ concentrations, the benefits on reducing
the latter are correspondingly higher.
Second, the linear econometric methodology yields benefit estimates
which are somewhat lower (~ 15% for NO^ calculations, ~ 62% for TSP cal-
culations) than those presented for the paired sample approach. This is an
expected occurrence since the linear econometric study explicitly accounts
for the variation in non-pollution variables through statistical means.
Therefore, this method can be considered an improvement over the previous
examination of mean housing values.
However, this approach is not without its associated problems.
Specifically, there has been much discussion in the property value liter-
ature that benefits based on a linear equation coefficient tend to overstate
the true willingness to pay for air pollution reductions (see Section 2.1
and references 1, 4, 5, 10, 11). That is, it is generally accepted that
the air pollution coefficient may be employed to value marginal change but
its applicability for total benefit calculations (non-marginal changes)
requires that further assumptions be made. For example, the linear equation
method contains the implicit assumption that every reduction in air pollu-
tion is valued identically by all households. This neglects variations in
average benefits which may accrue to particular population groups identified
by income or susceptibility to present pollution concentrations. In effect,
household preferences are assumed to be identical. This limits the
acceptability of the linear econometric approach. In the next subsection we
further refine this approach and address remaining issues.
I/
Econometric Approach - Non-linear Equation
The non-linear methodology is a multi-stage procedure, the objective of
which is to determine the benefits of air pollution abatement while allowing
different values for various individuals. In essence, this method addresses
the major criticisms of the previous approaches but must effectively assume
the mathematical form of individual preferences. The first step involves
the estimation of a hedonic housing value equation. This is similar to the
previous analysis except that in this case we do not arbitrarily restrict
the functional form to be linear. Non-linearities are to be expected in
an analysis of housing market data because: (1) the market may not be in
long-run equilibrium; (2) there may exist disequilibrium supply conditions;
or (3) there are indivisibilities among housing and neighborhood character-
istics. Therefore, in this step an attempt is made to find the functional
form which provides the best statistical fit for the data.
124
-------
The second step is to determine the marginal willingness to pay for
small changes in the air pollution data. This is done by taking the
derivative of the hedonic housing value equation (obtained from the initial
step) with respect to air pollution and evaluating for each of the four-
teen sample communities. Calculated for each community, this derivation
yields the within community average household willingness to pay for marginal
improvements in air quality. Determination of the marginal willingness to
pay is accomplished at the community level of aggregation based on the
assumption that the individual households within the community are com-
pletely homogeneous.
In the third step the marginal willingness to pay figures (obtained
in previous step) are regressed on a set of community characteristics
(income and present pollution level) in order to estimate a marginal
willingness to pay schedule. The resulting estimated equation provides
information on how various communities identified by these characteristics
value reductions in air quality. Thus, differentiation along community
preference can be accounted for. For instance, it is a widely held belief
that marginal willingness to pay increases with income. This hypothesis
is tested in this step.
The final step employs this latter estimated relationship to determine
the home sale price differential attributable to the previously specified
air quality improvements. Mathematical integration of the relevant mar-
ginal willingness to pay equation (a function of the stated community
characteristics) accomplishes this task. This final information component
is then inserted into equation (5.1) to derive average household benefits
in bid comparable terms.
The results of the hedonic housing value equation estimation are pre-
sented in Table 5.9. As measured by R^, the non-linear functional form
performs somewhat better than the linear equation. In the N02 equation all
independent variables conform to our a_ priori expectations concerning the
relationship to sale price and all except ethnic composition are statisti-
cally significant at the 5% level ( 11 _>_ 1.645). A similar statement holds
for the TSP equation except that crime replaces ethnic composition as the
only insignificant variable. In their respective equations, the air pol-
lution variables are highly significant. Note also that squared pollution
terms were utilized in the estimation. It was found that these performed
better than either the first-order or cubic terms. However, the performance
difference was not significant. Therefore, further analysis (benefit cal-
culations, etc.) based on the equations containing the first or third order
terms was completed and is discussed below.
The non-linear specification prevents straightforward analysis of the
quantitative impact of a unit change in an independent variable since the
effect depends upon the level of all other variables. However, if NC>2 and
the other variables are assigned these mean values than a unit improvement
in N02 (one pphm) is valued at $2,010.
Before proceeding to the next procedural step, a few comments concerning
the effect of misspecification bias are in order. That is, we conducted
experiments to see what would happen to the coefficient on air pollution if
125
-------
Table 5.9
Estimated Econometric Equations*
Dependent Variable - Log (Home Sale Price in $1,000)
Independent Variable
Sale Date
Age
Living Area
Bathrooms
Pool
Fireplaces
Distance to Beach
Distance to Employment
Crime
School Quality
Ethnic Composition
Population Density
Log (Tax)
Public Safety Expenditures
(TSP)2
(NO )
Constant
R2
Sum of Squared Residuals
Degrees of Freedom
NO Equation
.018439
(10.108)
-.0027044
(-3.5185)
.00019976
(14.024)
.14777
(9.2661)
.089959
(4.2096)
.10355
(7.8325)
-.014037
(-9.1443)
-.26979
(-11.663)
-2.2798
(-2.3574)
.00099327
(2.0286)
.0081532
(1.2523)
-.000067145
(-7.8422)
-.030991
(-1.8253)
.00032792
(5.1487)
-.0010374
(-2.6935)
4.2297
(6.2304)
.877
22.62
703
TSP Equation
.018924
(10.427)
-.0031401
(-4.1178)
.00019688
(13.896)
.15285
(9.6443)
.092764
(4.389)
.099225"
(7.5833)
-.013132
(-9.1824)
-.23201
(-9.1314)
-1.5245
(-1.5444)
.0010087
(2.0792)
.027307
(4.5564)
-.000061627
(-7.2705)
-.046438
(-2.7565)
.00028288
(4.8582)
-.000015702
(-4.1798)
—
2.3602
(3.8836)
.878
22.29
703
126
-------
certain neighborhood variables were omitted from 'the equation. For example,
if distance to beach is excluded then the air pollution coefficient
increases from .0010374 to .0034176. Similarly, if population density is
omitted then the pollution coefficient increases to .0024284. In each of
these cases the air pollution term serves as a measure of pollution and
other neighborhood disamenities as well. These specification errors would
eventually result in biased benefit estimates. Therefore, a fully specified
equation is crucial.
The estimated equations shown in Table 5.9 yield the marginal
willingness to pay for improvements in air quality by taking the derivative
with respect to the relevant air pollution variable. This procedure
supplies information on the amount of money the average household in each
community would be willing to pay for small changes in pollution levels.
This information, in conjunction with community average income and pollution
levels, are the basic inputs to the third methodological step - estimation
of the willingness to pay equation. Table 5.10 presents two formulations
of this equation for N02- The first assumes a linear relationship while
the second postulates a log-log form. As is indicated by the coefficients
both income and pollution are positively related to marginal willingness to
pay. Thus, higher income communities in poor air quality regions have the
greatest willingness to pay. Similar results were discovered for the TSP
based equations but they are not presented.
Given this analysis it then becomes possible to complete the multi-
step procedure and calculate: (1) the average sale price differential
attributable to changes in air quality; and (2) benefits derivable from
these changes in per home, per day units, through use of equation (5.1).
The first calculation is accomplished by integrating the xjillingness to
pay equations (assigning the income variable its mean value) over the
range of air quality improvement._/ In this manner, the reduction in
pollution consistent with the poor to fair improvement is valued at $5,793/
home for the linear N(>2 willingness to pay equation and $6,134/home for the
log-log NC>2 equation. The values which correspond to the fair-good change
are $4,244/home and $4,468/home, respectively. If TSP is used as the mea-
surement criteria then poor-fair is valued at $6,053/home (linear) and
$6,033/home (log-log) while fair-good is valued at $5,677/home (linear)
and $5,964/home (log-log).
The above figures are translated into average benefits illustrated in
Table 5.11 through application of equation (5.1). As can be seen from
examination of Table 5.11, daily household benefits calculated using the
multi-step procedure range from $1.40/day/home to $1.48/day/home or $42.00
and $44.40 per month, respectively for N0?. These are considered our
"best" estimates since the technique used as their specification at least
addresses known methodological problems. We correspondingly place the
most faith in them. Further, the TSP based calculations remain fairly
constant at about $1.60/day/home, so the daily household willingness £0 pay
to achieve the specified air quality improvements are relatively insensitive
to the pollutant used in the willingness to pay equation. The TSP results
are also insensitive to the specification of the hedonic housing equation,
the first link in this methodology. That is, whether the first or third
order TSP term was used in this equation (rather than the squared term)
127
-------
Table 5.10
Estimated Willingness to Pay Equations (NO )*
Dependent Variable = Marginal Willingness to Pay in Dollars
Independent Variable Coefficient t-statistic
Constant -1601.3 -2.7622
Income** .050051 8.2662
NO level 162.67 3.7832
2
R - .864
Degrees of Freedom = 11
Dependent Variable = Log (Marginal Willingness to Pay in Dollars)
Independent Variable Coefficient t-statistic
Constant -6.4845 -5.7025
Log (Income**) 1.1473 13.092
Log (N00) .87283 6.1051
R » .942
Degrees of Freedom = 11
*These equations are based on the hedonic houring value equation which
utilizes (NO )2 as the air pollution measure.
**The income variable is defined as average community income and in
dollars.
128
-------
Table 5.11
Benefits - tyulti-Step Econometric Methodology*
(A) NO (TSP) - Linear Willingness to Pay Equation
Change in Air Quality
Poor to Fair
Fair to Poor
Total
Capitalized Benefits
(Billion Dollars)
6.12 (6. A)
3.42 (4.6)
9.56 (11.0)
Annualized Benefits
(Billion Dollars)
R = .0925, CRF = .0995
.61 (.637)
.34 (.458)
.95 (1.095)
Per Home
Per Home Per Day
Per Home Per Month
Capitalized Benefits
(?)
5136 (5910)
Annualized Benefits (?)
R - .0925, CRF = .0995
511 (588)
1.40 (1.61)
42.00 (48.30)
(B) N02 (TSP)
Change in Air Quality
Poor to Fair
Fair to Poor
Total
- Log-Log Willingness
Capitalized Benefits
(Billion Dollars)
6.5 (6.4)
3.6 (4.7)
10.1 (11.1)
to Pay Equation
Annualized Benefits
(Billion Dollars)
R = .0925, CRF = .0995
.645 (.64)
.355 (.47)
1.0 (1.1)
Per Home
Per Home Per Day
Per Home Per Month
Capitalized Benefits
($)
5427 (5964)
Annualized Benefits ($)
R = .0925, CRF = .0995
540 (593)
1.48 (1.63)
44.40 (48.90)
*Note that in the estimated hedonic housing equation (step 1) the
second order pollution terms were used.
129
-------
had little effect on the eventual benefit calculations. However, this was
not the case for NC^- In this instance, daily household benefits fluctuated
from a low of $. 87/day/home or $26.10 per-month I(NC>2) used in hedonic
housing equation and linear willingness to pay equation] to a high of $2.09/
day/home or $62.70 per month (first order NO,., term used in housing equation
and linear willingness to pay equation).
In comparing these figures to the simpler property value approaches,
we again find an adjustment downward as the methodology becomes more re-
fined. This is consistent with our conjecture that the paired sample
approach would yield upper bound benefits. This result also provides
further support for the hypothesis that the linear econometric approach
overestimates the total willingness to pay for pollution reductions. This
overestimation can be partially corrected by employing the final procedure
posited here.
In conclusion, we have attempted to describe and utilize a multi-step
approach to the determination of air pollution abatement benefits. Each
of the steps is linked to those that preceed it. Therefore, benefit
calculations are a function of a hedonic housing value equation, the
resulting marginal willingness to pay data, and an estimated willingness
to pay schedule which yields the sale price differential attributable to
air quality. Finally, our "best" estimates of daily household benefits
was $1.40/day/home calculated using the second order NO term in the hedonic
housing equation and a linear willingness to pay equation. However,
benefits could easily range from $.87/day/home to $2.09/day/home.
5.4 Summary
This paper began with the premise that valuation of non-market com-
modities constitutes a socially desirable objective on efficiency and
equity grounds. However, no methodology, which is generally accepted,
exists to accomplish this goal. Therefore, any new experimental valuation
technique requires validation. The analysis undertaken here is an attempt
to satisfy this requirement for the contingent valuation approach.
This study can be viewed as a systematic investigation of housing
market data within the communities which comprise the sample plan. It
consists of three separate approaches. The first, the paired sample
methodology, is primarily based on the sample plan. In this procedure we
attempted to determine the benefits derivable from air quality improvements
through a comparison of sale price averages in the paired communities.
This approach was found to be beset with a number of problems, yet the
upper bound of $4.50/home/day for a poor-fair, fair-good improvement was
determined.
The second approach,, a linear econometric methodology, was an attempt
to utilize traditional property value analysis to develop benefit estimates.
The ordinary least squares regression technique was the basic tool used to
estimate a linear hedonic housing value equation. The benefit calculations
derived from this equation were considered an improvement over the paired
sample approach since explicit account was made for a number of housing
130
-------
and neighborhood variables. Thus, this analysis provided more refined
benefit figures but they were still considered an overestimate since no
allowance was made for varying valuations of air quality changes dependent
on houshold characteristics.
In the final approach, multi-step econometric methodology, we addressed
the criticisms which plagued the earlier approaches and developed more
refined benefit estimates. Our best estimate of willingness to pay for the
specified air quality change (about a 30% reduction in average ambient levels)
was approximately $1.40/day/home ($42.00 per month). This amount is based
on a hedonic housing equation which allows non-linearities [including using
(NOo) as a proxy for air pollution] and either a linear or a log-log
willingness to pay equation. However, this figure is not precise and
therefore we put the possible range of benefits at between $.87/day/home
($26.10 per month) and $2.09/day/home ($62.70 month).
131
-------
FOOTNOTES: CHAPTER V
~~ This analysis follow., closely the procedure developed by Harrison
and Rubinfeld (1978).
2/
— The formula used in these calculations is:
J Pollution before
(WTP.)d Pollution
Pollution after
where WTP = f(income, pollution).
i
132
-------
Chapter VI
PRELIMINARY COMPARISONS BETWEEN PROPERTY
VALUES AND ITERATIVE BIDDING RESULTS
The South Coast Air Basin experiment consisted of an attempt to value
air quality through examination of differences in property values and
through an interview survey instrument to measure willingness to pay. Six
pairs of neighborhoods were selected for comparative purposes. The pairings
were made on the basis of similarities of housing characteristics, socio-
economic factors, distance to beach and services, average temperature, and
subjective indicators of the "quality" of housing. Thus, for each of the
pairs, an attempt was made to exclude effects on property values other than
differences in air quality.
While the sample paired methodology was an attempt to establish com-
parability betxv>een results of the research designs, certain cautions should
be kept in mind. These additional assumptions are that:
1. an implicit hypothesis exists such that there is a directional
consistency between the types of biases of the two research
designs;
2. in a theoretical sense, each research design is measuring the
same "good;"
3. the groups being sampled are identical within the paired areas;
4. the time frames from which the valuation estimates are derived
are assumed constant (i.e., equilibrium versus non-equilibrium
contexts for individuals and markets); and
5. a problem exists in assigning proper weighting for a set of
diverse samples.
With these difficult qualifications in mind, let us turn to a pre-
liminary comparison of results obtained from the property value and sample
survey results. Table 6.1 provides some extremely preliminary results on
monthly valuations by households of an arbitrary improvement in air quality
in the Los Angeles Basin of approximately 30%. For the paired comparisons
property value study, the estimate per household with no adjustments for
household differences except in an areal and subjective sense (see Chapter
III), is approximately $135 per month. Extrapolated^the results to the basin
as a whole yields an annual benefit from an improved air quality improvement
of 30%, a value of approximately $4 billion dollars.
133
-------
Table 6.1
Alternative Estimates of Monthly Bids by Household,
Total Benefits for Air Quality Improvement
in the South Coast Air Basin
(Approximate 30% Improvement in Ambient Air Quality)
Average ($) bid per house-
hold per month
Annual benefits (selected
areas and groups of the
South Coast Air Basin)
in billions of $'s)
Property Value Study
Paired
Communities
$135
$3.96
Linear
Regression
$51-115
$1.25-2.6
Non-
Linear
3-Step
$42*
$.95
Survey Study
Mean.
Bid
$29**
$.65
Preliminary
Regression
Results
$26***
$.58
*Best estimate, possible range, $26-63 per month.
**Based on maximum total bid with an adjustment for years to achieve
improvements in air quality.
***Based on maximum total bid equation wich an adjustment for the amount
of air pollution information available to the household.
134
-------
The other extreme is an estimate of, -its value of improved air quality
per month by household utilizing the preliminary results in Appendix D
from the survey. This value is approximately $26 per month per household.
This yields a rough estimate of annual benefits from an approximate 30%
improvement in air quality of slightly more than $.5 billion dollars.
Further, intermediate estimates are calculated on the basis of various
economic assumptions delineated in Chapters II and V. By making various
assumptions with regard to the change of air quality in the Los Angeles
Basin, other estimates of improvements can be derived. For example, if it
can be presumed that the various areal groups, when bidding from a
reasonably poor air quality to a reasonably good air quality, were bidding
on the basis that their area would be totally cleaned up, an alternative
estimate of the mean bid on an annual benefits basis is $1.07 billion
dollars. This is comparable with the linear estimate derived from the pro-
perty value study. For the reasons given earlier, these researchers believe
that, depending on assumptions, a range of willingness to pay for both
studies anywhere from a low of approximately $20-30 to a high of approxi-
mately $140-150 per month per household is obtained.
It appears from these preliminary results and comparisons that con-
tingent valuation studies will tend to give a lower valuation of air quality
improvement than observing at the margin what happens in an extremely
volatile property market. However, only after substantial in-depth
statistical examination and comparability checks between the two studies
will the researchers be able to state unequivocably how these valuations
may turn out. The results compiled in this study suggest that survey
instruments, when compared to property value techniques, provide a rea-
sonable mechanism to obtain environmental quality benefit estimates. The
survey approach has the advantages that: (1) data can be collected at
low cost on specific environmental problems (the investigator is not tied
to the availability of existing data sets); (2) benefit measures can be
disaggregated across individuals and sources of benefits from various
characteristics such as aesthetic experiences and perceived health can be
obtained; and (3) a voluntary consumer statement of willingness to pay
gives some justification in and of itself for expenditures on air quality
and perhaps more generally on environmental quality programs.
As a final caution, it should be kept in mind that the South Coast Air
Basin studies were conducted in an area where individuals have both an
exceptionally well-defined pollution situation that they have encountered
and a well-developed hedonic price-property value market for clean air.
The effect of clean air on property values, and in turn, on the degree to
which people are aware of increased housing prices in high air quality
areas appears to be exceptionally well specified at this time in the
South Coast Air Basin. Note further that 1970 property values on the basis
of several studies have shown a much weaker association with air quality
than those that were obtained utilizing the 1977-78 air quality data set
applied here. We feel that this change reflects a substantial shift in
tastes and concern over air quality for this regional population. Therefore,
it should be recognized that the results of this experiment may well not
be generalizable to other situations where the environmental commodity,
135
-------
i.e., air quality, is not so well specified, either through actual market
prices or human perception.
136
-------
APPENDIX A
This appendix presents the bas-ic survey instrument employed in the
South Coast Air Basin study. As discussed, the initiation point for the
survey instrument was either acute health effects or aesthetic effects.
Three basic areas existed (good, fair, Bad), Thus the following combinations
existed for survey instrument types.
1.
2.
3.
4.
5.
6.
7.
Format
Format
Format
Format
Format
Format
Format
Format
for
for
for
for
for
for
for
for
A
A
A
A
B
B
C
C
area
area
area
area
area
area
area
area
moving
moving
moving
moving
moving
moving
to C*
to C*
to B" (Aesthetic)
to B (Acute)
to C (Aesthetic)
to C (Acute)
to C (Aesthetic)
to C (Acute)
(Aesthetic)
(Acute)
The structure of the different combinations was identical.
1 is presented for illustrative purposes.
Combination
137
-------
Table A.I
INDOOR ACTIVITY AND COST LIST
Activity
Indoor Spectator Events
Indoor Tennis
Raquetball, Handball
Table Te-nnis
Bowling
Indoor Gardening or
Fixing up House
General Exercise
Organized Sports Events
Reading
Television
Movies
Club Activities,
Organizations
Individual Sports
Swinnint?
Visiting Neighbors or
Friends
Other (specify)
V
Hours
Per Week
A i B
C
D
Tines
Per Week
A
Bi C
1
D
Location
(Hap Grid)
A ' B
1
|_
)
\
C I D
|
Miles
Traveled
A P , C i D
i
1
I
' 1 !
1
Direct
Coses
A
i
B
1
i
C
D
I Day
Equipment
Replacement
Coses
\
I
1
1 :
i
Importance
-------
Table A.2
OUTDOOR ACTIVITY AND COST LIST
Activity
Outdoor Spectator
Snorts
Tenni s
Biking
Be.sch Activities
General Exercise
Fishing
Sw i m.T. i n £
Sail in"
Jocelnr./'.Val'Kinc
Hobbies. Arts & Crjfts
Outdoor C.-rdening or
Fixing up House.'
Golf
Hiking
N/
Hours
Per Week
A
Canping
Organized Sports Events '
Individual Sports
Events
Other (spccliv)
I
B
C
D
l ;
'
1
Times
Per Week
A
B
C
D
|
I
Location
(Mao Grid)
A
B
C
D
Miles
Traveled
t.
B
C
D
1
Direct
CostE
A
B
C
D
i
% Day
Equipment
Replacement
Costs
Importance
-------
GAME FOR A AREA MOVING TO R, AESTHETIC Questionnaire //
Interviewer //
INTRODUCTION
1. In a typical week how much day and night leisure time do you have available?
This includes both weekdays and weekends. By leisure, I mean the time you
do not spend eating, sleeping, or working to earn a living. hours
2. Has air pollution influenced where you have chosen to live? Yes[ ] No[ ]
3. Has air pollution influenced where you have chosen to work? Yes[ ] No[ ]
4. Would you classify the air quality in the area where you live as:
Good[ ] Fair[ ] Poor[ ]
5. Would you classify the air quality in the area where you work as:
Good[ ] Fair[ ] PoorI ]
5a. What is your occupation?
6. Are you aware of any health hazards due to air pollution? Yes[ ] No[ ]
7. How long have you lived in the Los Angeles area? ___
8. How much longer do you plan to live in the los Angeles area?
9. Do you think automobile emission standards should be: Increased[ ]
Decreased[ ] Kept the Same[ ]
ADMINISTER INDOOR AND OUTDOOR ACTIVITY AND COST LISTS (TYPICAL WEEK) CHECK
THE TOTAL TIME CONSTRAINT
Bidding Game for Residents of Area A
Here are three photographs representing average levels of visibility
for the three different regions of the Los Angeles Area shown on this
map. Picture A represents poor visibility; Picture B represents fair
visibility; and Picture C represents good visibility.
Public officials are strongly considering the possibility of trying to
reduce the levels of emissions throughout the Los Angeles Area. Such
action could require additional funds which might be generated by (a
monthly charge, an extra charge in your utility bill) for as long as
you live in the Los Angeles area. These funds will be used to help
finance air quality improvements in the Los Angeles area.
140
-------
Aesthetic
The Los Angeles area has some very "Beautiful background scenery.
Because of automobile and industrial emissions', there is a haze which
reduces and distorts the ability to see this scenery. This means that
many people have to leave Los Angeles and travel long distances to be
able to enjoy views which could be visible from their homes if these
emissions were reduced.
As indicated by the map, you live in an area which has been classified
as having poor air quality relative to the rest of the Los Angeles
area. Picture A represents the visibility level which typically occurs
in your area. I am only interested in how you value being able to see
long distances.
If the level of emissions could be reduced in the Los Angeles area so
that visibility conditions would be represented by Picture B instead of
A, not only in the B area but also in your area, and if the air would
be cleaned up to this level in (2, 10) years, would you pay (a monthly
charge, an extra charge in your utility bill) of ($1, $10, $50) for as
long as you live in the Los Angeles area?
[ ] DO FOLLOWING ONLY IF CHECKED
LIFE TABLE: Here is a table that might help you. It shows the total amount
you would pay for as long as you live in Los Angeles for various amounts of
monthly payments.
RECORD MAXIMUM BID
Would you consider moving to a new location in the los Angeles area
if air quality were like Picture C everywhere? Yes[ ] No[ ]
IF YES:
Where would you move? (GRID LOCATION ON MAP)
141
-------
Bidding - Aesthetics and Acute Health
The questions I asked earlier were only concerned with, your perception
of visibility. This section deals with, not only visibility hut with
short term health effects which may be aggravated by air pollution.
Some pollutants in high concentrations cause eye irritation for many
individuals. Studies have shown that about half of the population
experiences eye irritation under conditions represented by Picture A;
about one-fourth experiences- eve irritation under conditions represented
by Picture B'; while no one experiences- these irritating effects when
conditions are represented by Picture C.
Since vou reside in area A, which has Keen classified as having poor
air quality, there is reduced visibility as well as irritating health
effects as compared with B'. If the level of emissions could be reduced
'in the Los Angeles area so that visibility and irritating health effects
were represented By Picture B not only in the B area but also in your
area, and if the air would be cleaned up to this level in (2, 10) years,
would you pay a (a monthly charge, an extra charge in your utility bill)
of (START BIDDING WITH PREVIOUS MAXIMUM BID) for as long as you live
in the Los Angeles area?
[ ] DO FOLLOWING ONLY IF CHECKED
Life Table: Here is a table that might help you. It shows the total amount
you would pay for as long as you live in Los Angeles for various amounts of
monthly payments.
RECORD MAXIMUM BID
REVISIONS IF NECESSARY
Would you consider moving to a new location in the Los Angeles area
if air quality were like Picture C everywhere? Yes[ ] No [' ]
IF YES:
Where would you move? (GRID LOCATION ON MAP)
142
-------
Substitutions
Aesthetic; Original Position A; Movement to B
NON-ZERO BIDS:
You stated that you would pay $ per month for as long as you live
in the Los Angeles area if visibility improved from A to a condition
like that shown in B, if this could be accomplished in ('2, 10) years.
If you actually paid this amount in order to help finance air pollution
control programs, you would have less money to spend overall. However,
because you said you would pay some amount of money, you are indicating
that clearer air is something you value. Consequently, when conditions
like B are achieved, even though you have paid money to help improve
the visibility, this does not mean that your standard of living is
worse than before because now you will be living and recreating in a.
less polluted area.
If the visibility conditions were to improve from A to B in your area,
would the improved conditions change the pattern of your leisure
activities? This could be changes in time per week, location, or
frequency. Yes[ ] No[ ]
IF NO, GO TO NEXT BIDDING GAME
IF YES, THEN:
A. Administer indoor and outdoor activity and cost list
B. Check time Constraint
ZERO BIDS:
Although you told me that you would not pay anything to have visibility
improve throughout the Los Angeles area to like that shown in Picture B,
would the improved conditions change the pattern of your leisure
activities? This could be changes in time per week, location, or
frequency. Yes[ ] No[ ]
IF NO, GO TO NEXT BIDDING GAME
IF YES,.THEN:
A. Administer indoor and outdoor activity and cost lists
B. Check time constraint
143
-------
Substitutions
Aesthetic + Acute; Original Position A; Movement to R
NON-ZERO BIDS:
With the extra information on possible short term health, effects when
conditions are like A, you said that you would pay $ per month
for as long as you lived in tne Los Angeles area if conditions im-
proved from those associated with Picture A to conditions shown in
Picture B, and if this could be accomplished in (2, 10) years. As
before, Because you said you would pay some amount of money, you are
indicating that clearer air is something you value. Consequently, when
conditions like A are achieved, even though you have paid mqnay to help
improve the visibility and to lessen short term health effects, this
does not mean that your standard of living is worse than before because
now you will be living and recreating in a less polluted area.
If conditions- improved so that Picture B were representative of the
entire area, with no visibility problems or irritating effects, would
the improved conditions change the pattern of your leisure activities?
Yes[ ] No[ ]
IF NO, GO TO NEXT BIDDING GAME
IF YES, THEN:
A. Administer indoor and outdoor activity and cost lists
B. Check time constraint
ZERO BIDS:
Although you told me you would not pay anything to have visibility
conditions and short term health effects improve throughout the area
to like those shown in Picture B, would the improved conditions change
the pattern of your leisure activities? Yes[ ] No[ ]
IF NO, GO TO NEXT BIDDING GAME
IF YES, THEN:
A. Administer indoor and outdoor activity and cost lists
B. Check time constraint
144
-------
Substitutions
Aesthetic + Acute + Chronic; Original Position A; Movement to K
NON-ZERO BIDS:
Given the information that continued exposure to levels of air pol-
lution like those shown In Picture A could actually reduce your life
expectancy, you said you would pay $ per month for as long as you
lived in the Los- Angeles- area if conditions improved from those in A to
those shown in IS, and if this- could lie accomplished in (2, 10) years.
Once again, I would like you to think of this expenditure as leaving
you as well off as before you paid the money, since you are now living
and recreating in a less- polluted area.
If conditions- improved so that Picture B were representative of the
entire area, with no visibility problems- or short and long term health
effects-, would the improved conditions change the pattern of your lei-
sure activities Yes[ ] No[ ]
IF NO, PROCEED TO GENERAL INFORMATION SECTION
IF YES, THEN:
A. Administer indoor and outdoor activity and cost lists
B. Check time constraint
ZERO BIDS:
Although you told me that you would not pay anything to have visibility
or short and long term health effects improve throughout the Los
Angeles area to like those shown in Picture B, would the improved con-
ditions change the pattern of your leisure activities? Yes[ ] No[ ]
IF NO, GO TO GENERAL INFORMATION SECTION
IF YES, THEN:
A. Administer indoor and outdoor activity and cost lists
B. Check time constraint
-------
Bidding - Aesthetic, Acute Health, and Chronic Health Effects
The quality of the air may also affect your long term health. There is
evidence that high concentrations of emissions as- represented in
Pictures A and B have lasting effects- upon the respiratory and circul-
atory systems in addition to eye irritation and reduced visibility.
Evidence shows- that, on the average, people who live in areas with con-
centrations- like those in Picture A can expect a reduced lifespan of up
to 2 years compared with people who-live in conditions represented by
B, and up to 3 years when compared with people who live in conditions
represented by Picture C.
If the level of emissions- could Be reduced in the Los Angeles area so
that visibility, short and long term health conditions would be repre-
sented By Picture B instead of A, not only in the 5 area but also in
your area, and if the air would be cleaned up to this level in...("2, 10.)
years, would you pay (a monthly charge, an extra charge in your utility
bill) of (START BIDDING WITH PREVIOUS BID) for as long as you live in
the Los Angeles area?
[ ] DO FOLLOWING ONLY IF CHECKED
Life Table: Here is a table that might help you. It shows the total amount
you would pay for as long as you live in Los Angeles for various amounts of
monthly payments-.
RECORD BID
REVISION IF NECESSARY
Is there some other payment scheme besides (a monthly charge, an extra
charge in your utility bill) that you would prefer? Yes[ ] No[ ]
IF YES:
What would it be?
Would you consider moving to a new location in the Los Angeles area
if air quality were like Picture C everywhere? Yes[ ] No[ ]
IF YES:
Where would you move? (GRID LOCATION ON MAP)
146
-------
GENERAL INFORMATION SHEET: WOULD YOU PLEASE FILL OUT THE FOLLOWING?
1. Age
2. Sex: MaleJ J FemaleJ ]
3. Marital status: Single! ] MarriedJ .]
4. Number of persons- In your household?
5. Your education: years. Highest degree obtained:
High School[ 1 College! ] Vocational! ] Advanced! ]
6. Address of employment:
7. Location of employer(s) (GRID LOCATION ON MAP)
8. Are there any environmental hazards associated with your job, such as
noise, health, or sight? Yes! •] No! 1 IF YES: What are these hazards?
9. What is the percentage of your work time indoors? %
10. In a typical work week how much time do you spend on the job? hours
11. If you received our pamphlet last week, did you read:
[ ] 0-5 pages
[ ] 5-10 pages
[ ] more than 10 pages
If you live in area A or B:
How much would you pay for this same house (apartment) today if it
were located in an area where the air pollution levels were like those
shown in Picture C? $
If you live in area C:
Do you believe that any part of the value of your home is because you
live in a relatively unpolluted part of Los Angeles? Yes[ ] No[ ]
IF YES: How much (% or dollars)
If you live in area B or C:
Would you consider moving if the air pollution problem were as bad as
A throughout the entire area? Yes[ ] No! ]
IF YES: Where would you most likely move? ('GRID LOCATION ON MAP)
147-
-------
How much do you think it would cost to clean up air pollution in the Los.
Angeles area to a. condition like that shown in Picture C?
$
If all citizens- were billed equally, how much do you think, it would cost
each person in order to achieve conditions like that shown in Picture C?
$
148
-------
1. Characteristics of. home:
Living Area: square feet
Number of Rooms:
Number of Bedrooms
Number of Bathrooms:
Other Rooms: ("PLEASE CHECK)
[ ] Den
[ ] Family room
[ ] Dining room
[ ] Enclosed porch
[ ] Attic
[ ] Basement
% Easement finished
Tl"Utility room
[ ] Other
Scenic View: YesI ] No [ ] IF YES: Specify
Number of Stories: (INCLUDE BASEMENT)
Remodeled: Yes[ ] No[ ] Don't know[ ]
IE YES: Specify previous style and date
2. Equipment: (PLEASE CHECK)
[ ] Dishwasher
.[..] Disposal
[ ] Central Air Conditioning
[ ] Trash Compactor
[ ] Central Heating
Pool: Yes[ ] No[ ]
IF YES: Circle whether heated, enclosed, or other (specify)
Fireplace: Yes[ ] No[ ]
Age of home: years (when constructed)
149
-------
IF YOU LIVE IN AN APARTMENT, GO TO QUESTION 4
3. a). Year of purchase:
b) Could you please indicate what the purchase price was: $_
c) What are your monthly payments: $
d) What do you feel your home is worth, in today's- market? $_
e) What are your property tax payments- per year? $
f) How long have you been living in this house? years
GO TO QUESTION 5
4. a) How long have you been living in this apartment: years
b) Would you indicate your monthly rent? $
5. What are your insurance payments per year? $
6. What do you pay monthly for general upkeep around your home (apartment)?
7. Why have you chosen to live in this area? RANK IN ORDER OF IMPORTANCE,
WHERE ONE IS MOST IMPORTANT. CHOOSE TOP FIVE.
[ ] Attractiveness of area in general
[ ] Close to work
[ ] Close to recreation activities
[ ] Close to friends
[ ] Close to schools
[ ] Close to services
[ ] Close to transportation routes
[ ] Air quality
[ ] Affordability of home
[ ] Low crime rate
[ ] Prestige of area
[ ] Quiet neighborhood
[ ] Other
8. What are your average expenditures per month for food: $
9. What are your average expenditures per year for clothing? $
10. Please mark the box corresponding to your annual household income.
I ] 0-$5,000 I ] $3Q,QOO-$35,aOO
[ ] $5,000- $10,000 I ] $35,000-$40,000
[ ] $10,000-$15,000 [ ] $40,000-$50,000
[ ] $15,000-$20,000 I ] $50,000-$60,000
[ ] $20,000-$25,000 [ ] $60,000-$80,000
[ ] $25,000-$3n,000 I ] Over $80,000
150:
-------
1. Do you own or share in the. ownership of a motor vehicle? Yesf J No[ J
Type of Vehicle(s) Model Year
Model Year
Model Year
2. How many licensed drivers are in your family?
3. How many miles- per gallon do you get for each car in the city?
mpg
mpg
How many miles do you and your family typically travel in your automo-
bile per week? miles
4. How many hours do you and your family spend in a typical week commuting
to:
Work or school hours
Shopping hours
Recreational activities hours
5. Do you participate in a car pool? Yes [ ] No [ ]
6. How much do you spend each month on:
a) Gasoline costs $
b) Maintenance costs $
c) Public transportation fares $
d) Insurance payments $
7. Did you.take a vacation within the last year where you were away from
home for more than 4 days? Yes[ ] No[ ]
IF YES: About how much were your expenditures on this trip? $
151
-------
1. Have you ever had any of the following? (PLEASE CIRCLE.)
a) High blood pressure a) Asthma
b.) Heart trouble f) Chronic nervous trouble
c) Strokfc g) Cancer
d) Chronic bronchitis h) Tuberculosis
2. Have you ever had trouble with the following? (PLEASE CIRCLE)
i) Pain in the heart or tightness or heaviness in the chest
j) Trouble breathing or shortness- of Breath
k) Frequent headaches
1) Constant coughing or frequent heave chest colds
m) Frequent eye irritations
n) Allergies
.o) Nose and throat irritation
3. Are any of these (the above) conditions aggravated (or made worse) by
heavy air pollution? Yes[ ] No[ ]
IF YES: Which ones? (LIST LETTERS CIRCLED)
4. Do .you suffer from any other diesases which could be made worse by poor
air quality? Yes[ ] No[ ] Specify
5. Do you or any member of your family have any physical disabilities which
limit your activities? Yes[ ] No[ ]
6. Would conditions like those in Picture C, if they occurred' over the
entire area?
a) Make your life more
pleasant Not at all[ ] To some degree! ]
Greatly[ ]
b) Require you to spend
less money on drug
items or doctor's Not at all[ ] To some degreef ]
fees Greatly[ ]
c) Make it easier to Not at all[ ] To some degree[ ]
do your work? Greatly[ ]
7. Do you enjoy doing your leisure activities more during the ^ay or during
the night? Day Night Makes no difference
8. Do you smoke? Yes I ] No[ J IF YES: How many packs per day?
9. Do you take medication regularly? Yes[ ] NoJ ]
IF YES: Monthly expense on this medication? $ /month
152
-------
10. How much do you spend monthly on medical problems associated with, air
pollution effects? $ /month.
11. How much, do you spend yearly for doctor's fees? $ /year
12. How much, do you spend yearly on medical and life insurance.?
$_J /year
13. Have you purchased any items to reduce your exposure to air pollution
(such as carbon filters)? Yes[ ] Not ]
IF YES: What items?
153
-------
1. Which, one, of theae statements applies to you? (CHECK. ONE)
[ ] I have not Been bothered by air pollution.
[ J I have been somewhat bothered by air pollution.
[ ] I have been bothered quite a lot by air pollution.
2. Do you believe that air pollution in Los Angeles: (CHECK ONE)
[ ] Has become worse since you have lived here.
[ ] Has stayed about the same since you have lived here.
I ] Has gotten better since you have lived here.
3. What do you think should be done about air pollution? (CHECK ONE)
[ ] Ignored
[ ] Reduced
4. Please rank the following problems in terms of importance (most to
least) as the major issues facing the community. (CHOOSE TOP FIVE)
[ ] Juvenile delinquency [ ] Nuclear energy
[ ] Communicable disease [ ] Alcoholism
[ ] Unemployment [ ] Water pollution
[ ] Air pollution [ ] Energy
[ ] Car accidents [ ] Congestion
[ ] Crime [ ] Other
5. Do you believe that air pollution in the Los Angeles area:
[ ] Cannot be reduced below its present level
[ ] Can be reduced below Its present level
[ ] Can be almost completely eliminated
6. What do you think the words "air pollution" mean to most people in the
Los Angeles area? Do they mean:
a) Frequent bad smells in the air YesI J Nol ]
b) Too much dirt and dust in the air YesI ] Nol ]
c) Frequent haze or fog in the air Yes[ ] Nol 1
d) Frequent irritation of the eyes YesI ] No[ ]
e) Frequent nose or throat irritation YesI ] Nol ]
f) Other YesI J
7. Have you read or seen anything in the newspaper recently about air
pollution? YesI ] No! ]
8. When you read the newspaper, do you generally choose to read articles
on air pollution? YesI ] Nol ]
154
-------
9. Do you consult the daily air pollution index before engaging in any
activities? Yes.f ] No [ J
IF YES: What kind of activities?
155
-------
1. If you received our health, pamphlet, what do you think about it?
2.
[ ] Did not read
[ ] Very informative
[ ] Hard to understand
[ ] Scientific mumbo-jumbo
[ ] Made me more concerned ab.out
health", effects
1 ~] Had no influence on me
Here is a list of words and phrases. Select two which describe how you
feel about your participation in this survey.
[ ] Stimulating
[ ] Just tolerable
[ ] A waste of time
[ ] Educational
[ j Boring
[ ] An invasion of privacy
'[ ] Interesting
[ ] Kind of fun
[ ] Hard to take seriously
Here is a different list of words and phrases.
how you feel about the questionnaire.
Select two which describe
[ ] Relevant
[ ] Credible
[ ] Likely to influence air quality control
[ ] Unrealistic
[ ] Pretty flakey
[ ] Unlikely to have any effect on air quality control
[ ] Irrelevant
Finally, here is another list of words and phrases. Select one from
each column to describe how you feel about your answers to the question-
naire.
Column 1
[ ] Quite accurate
[ ] There was no way I could
come up with accurate
answers.
[ ] Accurate in a "ball park"
kind of way.
THANK YOU FOR YOUR COOPERATION.
Column 2
[ ] A fairly good guide for
valuing air quality
[ ] A good guide for valuing
air quality.
[ ] A poor guide for valuing
air quality.
156
-------
APPENDIX B
The following represents the health pamphlet that was sent to half of
the respondents who were contacted by phone and agreed to participate in
the study.
157
-------
AIR POLLUTION AND HEALTH
This pamphlet will try to answer some questions about air pollution
and human health. How do the major pollutants affect the body? What is
known scientifically about these effects? What kinds of real life studies
have been carried out to test facts learned in the laboratory? This
infgrmation is provided so that you can draw your own conclusions about
the health effects of air pollution.
Nearly every day in the Los Angeles area a chemistry of air and sun-
light gives rise to toxic gases known as photochemical oxidants. These,
together with carbon -monoxide, sulphur dioxide, nitrogen oxide, hydrocar-
bons, aldehydes, and ketones, make up the haze, account for its aroma, and
may impair human health.
Environmental standards like those in Table B.I
attempt to protect the public. When the concentration of any pollutant
exceeds the standard, acute, short term, irritating symptoms may be noticed.
These acute effects, such as chest tightness, eye irritation, slowing of
response time, and attention loss, are not experienced by everyone, but
people with pre-existing heart condition and lung disease are particularly
vulnerable.
In addition to acute health effects, chronic effects of long term
exposure to low and average levels of the oxides, aerosols, particulates,
and other elements of the haze are a particularly challenging question.
Does air pollution cause influenze sometimes, or does it merely make it
more of a problem? Links between oxides of nitrogen and cancer have been
investigated. Finally, it is possible that years of continuing exposure
could have some influence on total lifespan.
158
-------
The Major Pollutants
Carbon Monoxide
The major sources of carbon monoxide pollution are automobiles, trucks,
buses, and, to the habitual smoker, cigarettes. Peak hourly readings of
carbon monoxide from 1963 to 1970 averaged 10.8 ppm* in the Los Angeles
area, and exposure in die California Central Valley was about half this
figure for the same period. [ll]
Carbon monoxide has a very direct effect on the human body. Entering
the lungs, it diffuses into the blood, where it is absorbed by red blood
cells and displaces and competes with oxygen. Carbon monoxide reduces the
oxygen carrying capacity of the blood. Low concentrations cause tiredness
and listlessness.
The heart is doubly affected. Its oxygen supply is reduced, but at the
same time it must exert more effort to increase its output if body oxygen
transport is to be maintained. And so, at relatively low concentrations of
carbon monoxide (10-50 ppra for one hour exposure), [191 patients with heart
disease may experience adverse effects.
During heavy muscular exercise, the oxygen consumption rate of the
body increases to as much as 20 times the rest rate. Consequently, carbon
monoxide exposure reduces maximum exercise performance.
In controlled experiments with humans, researchers have projected the
maximum levels and exposure times shown in Table B.2. Normal healthy
individuals are unlikely to experience any of the above effects until the
threshold concentration is 21-72.5 ppm. [19] Individuals with emphysema,
bronchitis, and asthma are more sensitive, perhaps experiencing effects at
17.5-52.5 ppm, and heart patients are extremely sensitive to carbon
monoxide, as noted above. [19] All these effects are acut^e, occurring at
high concentrations. The effects of exposure over long periods to low
carbon monoxide levels are not known at this time.
Sulphur Dioxide
Los Angeles has not had a deadly pollution episode such as those
observed in Belgium; Donora, Pennsylvania; London; or New York, but an
ingredient of Los Angeles air pollution, sulphur dioxide, is held respon-
sible for high death tolls in these places.
A well known air pollution episode occurred December 1-5, 1930 when
several hundred persons became ill in the Meuse Valley, Belgium. There
were 63 deaths. It was estimated that sulphur dioxide and sulphuric acid,
which may have reached a level of 9 ppm, were the chief causes of illness.[30]
*ppm denotes the number of pounds of pollutant for each million pounds
of air.
159
-------
During late October, 1945, Donora, Pennsylvania, was blanketed by dense fog.
Forty-three percent of the population was affected. Twenty persons died.
Ten percent of the residents were severely affected. Again, sulphur
dioxide was held to be responsible.
In London, England during December, 1952, the world's worst air pol-
lution incident occurred, causing about 4,000 more deaths than would be
expected in the Greater London Area for a month's period. Marked increases
in deaths both from lung and heart disease were observed. Detailed invesgi-
gations of 1,280 post-mortem reports of persons who had died before, during,
or shortly after the episode indicated all such fatalities could be explained
by previous health problems among the victims. The elderly and persons with
already existing lung and heart disease were most susceptible. During this
time in London, daily sulphur dioxide and smoke measurements were from two
to four times higher than typical winter levels. [30]
'Sulphur dioxide, as is well known, has an odor. It is readily soluble
in water and, when breathed, is absorbed quickly in the upper airx^ays of the
nose. In Table B.3 are recorded laboratory observations of throat and lung
effects from sulphur oxides. In air with small dust particles, sulphur
dioxide is partially converted into sulphuric acid, which may be a severe
problem of its own. The Los Angeles area currently has relatively low
levels of sulphur dioxide.
Photochemical Oxidants
Along with carbon monoxide, gasoline engines produce nitric oxide*
and hydrocarbons. Secondary products of these emissions, photochemical
oxidants - ozone, nitrogen dioxide, and peroxyacetylnitrate (PAN), may be
more toxic than the original compounds.
Early morning car traffic produces exhaust with large quantities of
nitric oxide and hydrocarbons. In the presence of sunlight these products
react, converting nitric oxide into nitrogen dioxide but low nitric oxide
levels. Then nitrogen dioxide breaks down into ozone during the afternoon.
Late afternoon automobile traffic again emits large amounts of nitric
oxide, which reacts with ozone, removing most of the ozone.
Ozone is among the most poisonous of gases. Relatively insoluble in
water, when inhaled, ozone can damage the central airways and other pas-
sages of the lung.
Health studies of certain occupations have provided understanding of
the effects of exposure to oxidants. A 51 year old welder who was working
in a poorly ventilated area, developed a kind of pneumonia which lasted
for six days.l9j A crane operator working above a tank into which ozone
was bubbled developed a dry cough and frontal headache after two hours
*Nitric oxide is also a byproduct of natural gas combustion and the
processing of nitric acid industrially.
160
-------
Table B.I
Pollution Levels and Standards
(Parts Per Million)**
National
Standard for
one hour
exposure
Average peak
hourly level
in Lennox,
1973-75
Average peak
hourly level
in Costa Mesa
Harbor,
1973-75
Average peak
hourly level
in Pasadena,
1973-75
Ozone
0.03
CL04
0.05
0.11
Carbon
Monoxide
40.0
12.1
9.53
Nitrogen
Dioxide
0.99
0.07
0.13
Sulphur
Dioxide
0.50
0.06
0.03
0,03
Source: Three Year Summary of Califorjaia_A_ij__r^
Air Analysis Branch and EDP I'anacement Section (January 1977),
State of California Air Resources Board.
*State of California hourly standard.
""'Parts per million denotes the number of pounds of pollutant found
in a million pounds of air.
161
-------
Table B.2
o
Carbon Monoxide Effects
(National Standard: 40 parts per
million/one hour exposure)
Concentration ppm
Exposure Time
Effects
Acute Effects
50
53
100
500
1,000
2 hours
1.5 hours
0.5 to 2
hours
1 hour
Shortened average time to
a heart attack among
individuals with heart
disease^
Shortened average time to
a heart attack among
individuals with heart
disease.
Loss of physical and
mental coordination among
healthy subjects.
Mild to throbbing headache
among healthy subjects.0
Vomiting, unconsciousness
and death among healthy
subjects.
Sources: Leung, Goldstein and Dalkey, Final Report: Human Health Damages
from Mobile Source Air Pollution, 197^, California Air Resources
Board.
W.S. Aronow and H.W. Isbell, "Carbon Monoxide Effect on
Exercise-Induced Angina Pectoris," Annals of Internal Medicine 79
(1973), 392-395.
J. Koch-Weser, "Common Poisons," in Harrison (ed.), Principles of
Internal Medicine, Ch. 166 (1970), 652-653.
162
-------
Table B.3
Sulphur Dioxide Effects
(National Standard: .5 parts per
million/one hour exposure)
Concentration ppm Exposure Time Effects
Acute Effects
1 0.5 hour Choking sensation in some
individuals.
5 3-10 About 80% of healthy
minutes individuals will have
difficulty in breathing.
5-10 — Deep gasping feeling,
severe choking in some
individuals.
Source: National Pollution Control Administration, Air Quality for
.Sulphur Dioxides, 1969, U.S. Department of Health, Education
and Welfare, Public Health Service, Consumer Protection and
Environmental Health Service.
163
-------
exposure. In attempting to leave the crane, the employee nearly lost
consciousness. After administration of oxygen, he improved, and 48 hours
later showed no adverse symptoms. [1 I]
Other studies have confirmed these effects. A long term study over
the period October, 1961, to June, 1964, recorded daily symptoms in student
nurses in good health from two Los Angeles nursing schools. The nurses
kept diaries for 868 days on appearance of cough, chest discomfort, and
headaches. For the same period hourly peak concentrations of photochemical
oxidants, carbon monoxide, and daily temperature were measured at stations
within two miles of the schools.
Cough and chest discomfort increased with higher hourly concentrations
of ozone. Headaches had some association with ozone levels but less than
other symptoms. Eye discomfort, not a direct effect of ozone, although
often associated wtih photochemical oxidants, was the most strongly noted
symptom. When the oxidant level reached 0.5 ppm, a third of the nurses
reported eye irritation. Temperature, carbon monoxide, and nitrogen
dioxide levels did not explain the results found. Because all participants
were young, healthy adults, relatively free from chronic disease, the
effects on elderly persons or on those with chronic heart or lung disease
could be expected to be more severe. [22]
During the period 1963-1970 peak hourly readings of oxidants averaged
.104 ppm in the Los Angeles area, and again readings in the California
Central Valley were about half this figure. [19]
According to a panel of experts,
"the oxidant threshold in normal individuals ranges from 0.05 to
0.20 ppm. The threshold concentration is lowered among young
and old individuals, and also those with underlying disease. Those
with respiratory and chronic obstructive diseases are most sen-
sitive to the photochemical oxidants, and the threshold levels
for these population groups range from 0 to 0.20 ppm." [19]
In addition to discomfort and aggravation of existing lung disorders,
ozone and other photochemicals can cause changes in behavior. Automobile
accidents, for example, were recorded in each daylight hour of each weekday
in the "high smog" months of August through November for two years. A
relationship between Los Angeles oxidant concentrations and the number of
car accidents was found. [ 22 ] Attention span and visual performance were
reduced. Lethargy is reported as well as difficulty concentrating. 0.8 ]
Because lung function is impaired, evidence suggests photochemical
smog increases individual vulnerability to acute throat or lung infections.
Studies with experimental animal populations have reported changes in the
makeup and working of the lung as well as lung-tumor acceleration. (25l
Whether ozone is a cancer causing agent is currently an important topic
for research. Human white blood cells exposed to ozone exhibited chromosome
breakage and genetic abnormality. [12] In Table. B.4 a summary of ozone
effects is given.
164
-------
Table B.4
Summary of Fxperimental Data on Ozone Effectsc
(National Standard: .OB parts per
million/one hour exposure)
Concentration ppn
Exposure Time
Effects
Acute Effects
0.15-0.30
0-37-0.70
0.25
0.5
0.8-1.7
1.0-2.0
2 hours
Sources:
"ye irritation due to some
photochemical products.
Cough, nose and upper
throat irritation, chest
soreness, chest tightness,
symptoms made worse by
exercise, headache in 50%
of normal subjects.
Less than 6% of asthmatics
may have attacks "hen this
level is reached.
Formation of fluid in the
lungs among healthy
subjects.
Lung congestion.
Incapacitating illness
anonp, normal subjects.
Leung, Goldstein and Dalkey, Final Report: _Human Health Damage_s
from '"obile Source_Air PollutionTT^lT California Air
Resources Board..
G.E. Schoettlin and E. Landau, "Air Pollution and Asthmatic
Attacks in the LA Area,'' Public Health Report 76 (1961),
545-543.
165
-------
Nitrogen dioxide, a by-product of auto emissions, has effects similar
to ozone but at higher levels of concentration. Originally, exposure to
nitrogen dioxide was known as "silo-filler's syndrome," since extremely
high condentrations of nitric oxide and nitrogen dioxide are generated
within farm silos. Documented deaths from a kind of pneumonia and acute
throat and lung ailments were traced to this type of exposure.
Like ozone, nitrogen dioxide's low water solubility allows it to
penetrate deeply into the lung, where it damages tissue. At low concen-
trations, it impairs breathing. At higher levels it increases the risk
of an individual having a throat or lung ailment. At 25-100 ppm, it causes
acute (but quickly remedied) symptoms of pneumonia and bronchitis.
A study of the environmental health effects of nitrogen dioxide was
conducted in four residential areas, each containing three elementary
schools, in greater Chattonooga, Tennessee. One area, close to a large
TNT-plant (which processes nitric acid), had high nitrogen dioxide and low
particulate exposure. Another area had high suspended particulate and low
nitrogen dioxide concentrations. The other two areas were "clean" and
used for comparisons. Careful monitoring of particulate matter, nitrates,
sulphates, and gaseous nitrogen dioxide concentrations was conducted in
1968 and 1969 in these four areas.
Two possible health effects of nitrogen dioxide exposure were investi-
gated: (1) difficulty breathing in elementary school, children; and (2)
increased frequence of respiratory illness in family groups. It was
established that second grade school children in the high nitrogen dioxide
area were consistently higher than those in the two control areas during
the study period. However, the researchers could not establish a relation-
ship between chronic bronchitis and the levels of nitrogen dioxide. [26]
Tn the period 1963-1970, nitrogen dioxide levels in the los Angeles
Area averaged .28 ppm and therefore constituted a potential health hazard
given the estimates of effects in Table B.5
The air pollution health problem of the Los Angeles Area is far more
complex than brief accounts of the hazards of carbon monoxide, sulphur dioxide,
and two photochemical oxidants can indicate. For one thing, literally hun-
dreds of hydorcarbon compounds are present in the Los Angeles air, each with
its own characteristics and products. A group of secondary organic aerosols
may be responsible for adverse health effects and undoubtedly contribute to
visibility loss. Little, however, is known of the mechanism and health impact
of these compounds.
However, there is evidence that the pollutants discussed have health
effects at levels experienced within Los Angeles and other urban areas.
These pollutants cause irritation and stress within the lungs and heart.
Acute effects range from eye, nose, and throat irritation, headache,
chest tightness, difficulty in breathing to the aggravation of bronchitis,
asthma, emphysema, other lung ailments and heart disease. A relationship
between episodes of sulphur dioxide and particulate pollution and increased
166
-------
Table B.5
Nitrogen Dioxide Effectsa
(National Standard: .05 parts per
million/one hour exposure)
Concentration ppm
Exposure Time
Effects
Acute Effects
0.7-2.0
4-5
4-5
6-40
greater than
25
150-200
greater than
200
10 minutes
10 minutes
1 hour
5 minutes
1 hour or
less
Difficulty expelling air
from the lungs increased
by 15% and difficulty
breathing air into the
lungs increased by 50% in
normal subjects.
Half hour after exposure,
difficulty breathing,
increased by 117 to 92%.
Decrease in oxygen in the
blood.
Increased difficulty
breathing by 24% in
normal subjects.
Bronchiolitis and pneunonitis
in normal subjects.
c
Disintegration of the lung.
Lunp, fluid formation and
death.
Sources; Leung., Goldstein and Dalkey, Final Report: Human Health Damages
from Fobile Source Air Pojlution, 1975, California Air resources
Board.
D.V. Bates, "Air Pollution and the Hunan Lunp,,': American Review of
Respiratory Disorders. 105 (1972), 1-13.
°H.E. Stokinger and D.L. Cottin, "Biologic Effects of Air Pollutants,1'
in A.C. Stern, ed., Air Pollution and Its Effects, Ch. .13,
New Yorlr: Ararlcmin Press£l958~)~j 445-546. '
167
-------
death rates is accepted. The chronic effects of photochemical oxidants
lower general resistance to infections of the respiratory tract and lung
since they cause damage to the lung. Behavioral changes associated with
carbon monoxide, ozone, and nitrogen dioxide have heen documented. Activity
levels are depressed and overall work ability is impaired through visual and
chemical intervention.
These air pollutants may also shorten the lifespan by aggravating
existing health problems, particularly those problems associated with the
respiratory tract.
168
-------
BIBLIOGRAPHY
1.. Abe, Matsuo, "Effects of Mixed NO -SO Gas on Human Pulmonary Functions,"
Bulletin, Tokyo Med. Dent. UniversiEy 14 (1967), pp. 415-433.
2. Anderson, E.W., R.J. Andelman, J.M. Strauch, N.J. Fortain, and J.H.
Knelson, "Effects of Low-Level Carbon Monoxide Exposure on Onset and
Duration of Angina Pectoris: A Study in Ten Patients with Ischemic
Heart Disease," Ann. Intern. Med. 79, 46, 50 (1973).
3. Aronow, W.S. and M.W. Isbell, "Carbon Monoxide Effect on Exercise-Induced
Angina Pectoris," Ann. Int. Med. 79 (1973), pp. 392-395.
4. Bates, D.V. , "Air Pollution and the Human Lung," American Review qf_
Respiratory Disorders 105 (1972), pp. 1-13.
5. Bates, D.V., G.M. Bell, C.D. Burnham, M. Hazucka, J. Mautha, L.D. Penselly,
and F. Silverman, "Short-Term Effects of Ozone on the Lung," Journal
of Applied Physics 32 (1972), pp. 176-181.
6. Beard, R.R. and G.A. Wertheim, "Behavioral Impairment Associated with
Small Doses of Carbon Monoxide," American Journal of Public Health 57
(1967), pp. 2052-2022.
7. "Behavioral Toxicology Looks at Air Pollutants," interview with C.
Xintaras, Environ. Sci. Tech. 2 (1968), pp. 731-733.
8. California Department of Public Health, Clean Air for California, Initial
Report of the Air Pollution Study Project, Berkeley (March 1955, March
1956, and February 1957).
9. Challen, P.J.R., D.E. Hickish, and J. Bedford, "Investigation of Some
Health Hazards in Inert-Gas, Tungsten Arc Welding Shops," British
Journal Ind. Med. 15 (1958), pp. 276-282.
10. Committee on Medical and Biologic Effects of Environmental Pollutants,
Division of Medical Science, Assembly of Life Sciences, National
Research Council, Ozone and Other Photochemical Oxidants, National
Academy of Sciences, Washington, D.C. (1977).
11. Coordinating Committee on Air Quality Studies, National Academy of
Sciences, National Academy of Engineering, Air Quality and Automobile
Emission Control, Vol. 3, The Relationship of Emissions to Ambient Air
Quality, U.S. Government Printing Office, Washington, D.C. (1974).
12. Fetner, R.H., "Ozone-Induced Chromosome Breakage in Human Cell Cultures,"
Nature 194 (1962), pp. 793-794.
13. Haagen-Smit, A.J. and L.G. Wayne, "Atmospheric Reactions and Scavenging
Processes, Ch. 6," in A. C. Stenn, ed., Air Pollution and Its Effects,
New York: Academic Press (1968), pp. 149-186.
169
-------
14. Hammer, D.I., V. HasselbJLad, 11. Portnor, and P. Wehrle, "Los. Angeles
Student Nurs.e Study, Daily Symptom Reporting and Photochemical Oxidants,"
AEH 28 (1974), pp."255-260.
15. Hazacha, M. , F. Silverman, C. Parent, S'. Field, and D.V. Bates, "Pul-
monary Function in Man After Short Term Exposure to Ozone," Arch,
Envinson, Health 27 (1973), pp. 183-185.
16. Kelly, F.J. and W.E. Fill, "Ozone Poisoning," AEH 10 (1965), pp. 517-519.
17. Koch-Weser, J., "Common Poisons," in Harrison (ed.), Principles of
Internal Medicine, Ch. 166, 6th Edition (1970), pp. 652-653.
18. Lagerwertt, J.J., "Prolonged Ozone Inhalation and Its Effects on Visual
Parameters," Aerospace Med. 34 (1963), pp. 479-486.
19. , Leung, Steve, Elliot Goldstein, and Norman Dalkey, Human Health Damages
from Mobile Source Air Pollution: Final Report, California Air
Resources Board (March 1975).
20. Lowry, T. and L.M. Schumann, "Silo-Fillers Disease Syndrome Caused by
Nitrogen Dioxide," JAMA 162 (1956), pp. 153-160.
21. Mert, T.H., H.A. Bender, H.D. Kerr, and T.J. Kulle, "OBservations of
Abberations in Chromosomes of Lymphocytes from Human Subjects Exposed
to Ozone at a Concentration of 0.5 ppm for 6 to 10 Hours," Mutat. Res.
31 (1975), pp. 299-302.
22. Proceedings of the Conference on Health Effects of Air Pollutants,
Assembly of Life Sciences, National Academy of Science, National
Research Council, Oct. 3-5. Prepared for the Committee on Public
Works, U.S. Senate (1973).
23. Ramirez, R.J. and A. R. Dovell, "Silo-Fillers Disease: Nitrogen Dioxide
Induced Lung Injury, Long Term Follow-up and Review of the Literature,"
Ann. Int. Med. 74 (1971), pp. 569-576.
24. Ruffin, J.B., "Functional Testing for Behavioral Toxicity: A Missing
Dimension in Experimental Environmental Toxicology," J. Occup. Med.
5 (1963), pp. 117-121.
25. Schoettlin, C.E. and E. Landau, "Air Pollution and Asthmatic Attacks in
the LA Area," Public Health Report 76 (1961), pp. 545-548.
26. Shy, C.J., J.P. Creason, M.E. Pearlraan, K.E. McClain, F.B. Benson, and
M.M. Young, "The Chattanooga School Study: Effects of Community
Exposure to Nitrogen Dioxide, II, Incidence of Acute Respiratory
Illness," Journal of the Air Pollution Control Association 20(9)
(September 1970), pp. 582-588.
27. Shy, C.J., J.P. Creason, M.E. Pearlman, "The Chattanooga School Children
Study-II, Incidence of Acute Respiratory Illness," Journal of the Air
Pollution Control Association 20 (1970), pp. 582-588.
170
-------
28. Stokinger, H.E. and D.L. Cot tin, "Biologic Effects of Air Pollutants,"
in Air Pollution and Its Effects, A.C. Stern (ed.)> New York: Academic
Press, Ch. 13 (1968), "pp. 445-546.
29. U.S. Department of Health, Education and Welfare, Public Health Service,
Consumer Protection and Environmental Health Service, Air Quality
Criteria for Sulphur Oxides (January 1969).
30. U.S. Department of Health, Education and Welfare, Pulbic Health Service,
Consumer Protection and Environmental Health Service, Air Quality
Criteria for Sulphur Oxides (January 1969).
31. Ury, U.K., "Photochemical Air Pollution and Automobile Accidents in Los
Angeles: An Investigation of Oxidant and Accidents, 1963 and 1965,"
AEH 17 (1968), pp. 334-342.
32. Von Nieding, G., "Studies of the Acute Effects of NO on Lung Function,
Influence of Differsion, Pertusion, and Ventilation"in the Lungs," Int.
Trde. Arbeits Med. 31 (1973), pp. 61-72.
33. Young, W.A., D.B. Shaw and D.V. Rates, "Pulmonary Function in Welders
Exposed to Ozone," AEH 7 (1973), pp. 337-340.
-------
APPENDIX C
The following tables have Che following underlying assumptions.
1. The mean bids within each area are differentiated with respect to
the sequence that the air quality effects are presented, i.e., whether those
effects are introduced in "Aesthetic -> Acute Health -»• Chronic Health" or
"Acute Health ->• Chronic Health -*• Aesthetic" order. In the graphs, "Ae"
denotes the mean bid for aesthetic effects; "Ac" denotes the mean bid for
acute health effects; and "Ch" denotes the mean bid for chronic health
effects.
2. The mean bids within each "A" area are differentiated with respect
to the range of the hypothetical improvement, i.e., whether the improvement
is from A to B or from A to C. Since there is only one range of improvement
for the "B" and the "C" areas, i.e., from B only to C and from C only to C*,
no such differentiation is made for these areas.
Taking into consideration the variations in (1), each A area requires
four different graphs, and each B and C area requires two different graphs.
A denotes poor air quality
B denotes fair air quality
C denotes good air quality
3. The mean bids within each area are differentiated with respect to
the proposed completion date of cleanup; i.e., 2 years versus 10 years.
4. The bids from each respondent are obtained as follows: First, his
maximum bid is elicited following a certain hypothetical improvement in the
aesthetic (acute health) effects of air quality. Second, he is asked how
much he would increase his bid if the acute health (chronic health) effects
are also taken into consideration. Finally, he is asked to revise his bid
for the additional inclusion of the chronic health (aesthetic) effects.
The implicit assumption throughout this procedure is the linear addi-
tivity of bids for each effect.
No differentiation has been made whether a health pamphlet has or has
not been sent to the respondent in advance of the interview.
No differentiation has been made with respect to the different proposed
vehicles for the collection of bids.
No differentiation has been made with respect to the different starting
bids offered by the interviewer.
No differentiation has been made whether a life table has or has not
been shown to the respondent during the interview. A life table depicts
the "stock" counterparts of the elicited monthly bids for various expected
lifespans.
172
-------
Table C.I
Mean Bids by Area by Type*
(Completion Date of Cleanup: 2 yrs.)
Area**
El Monce
(A - B)
El Monte
(A - C)
La Canada
(A •» B)
La Canada
(A - C)
Koncebello
(A -. B)
Moncebello
(A * C)
Canoga Park
(B ~ C)
Mean Aesthetic Bid
($/month)
Typo I
2.00
(0.95)***
(5)****
13.40
(8.15)
(5)
13.20
(9.37)
(3)
22.60
(13.70)
(5)
1.60
(0:93)
(5)
26.43
(20.69)
(7)
10.00
(5.77)
(3)
Tvpe II
1.00
(1.00)
(5)
0.13
(0.13)
(2)
0.00
(0.00)
(2)
0.00
(0.00)
(5)
3.75
(2.39)
(4)
4.67
(2.60)
(3)
1.20
(0.971
C5).
Mean Acute Health Bid
($/~.onch)
Type I 1 Tvnc II
3.00
(1.55)
(5)
1.40
(0.93)
(3)
1.20
(0.97)
(5)
2.40
(1.12)
(5)
0.20
(0.20)
(5)
3.57
(1-71)
(7)
10.17
(7-58)
(3)
9.20
(4.76)
(5)
7.50
(2.50)
(2)
1.50
(1-50)
(2)
a. oo
(3.74)
(5)
Mean Chronic Health Bid
($/iaonth)
TV£C_I
Tvpe II
Mean Total Bid
(S/!so:ith)
Tv.ic I
1
2.00
(1.22)
(5)
3.60
(2.91)
(5)
2.20
(1.96)
(5)
1.00
(1.00)
(51
j
12.50
(4.79)
(•'•)
10.75
(5.45)
(4)
15.40
(4.71)
0.20
(0.20)
(5)
7.86
(7.06)
(7)
1.67
(1.67)
(5) : (3)
0.40
(0.40)
(5)
2.75
(2.25)
(2)
0.00
(0.00)
(2)
21 .00
(10. 30)
(5)
2.50
(2.50)
(4)
0.25
(0.25)
(4)
3.80
(2. 34)
(5)
7 . DO
(3.00)
(5)
IS. 40
01.75)
(3)
16.60
(9.03)
(5)
26.00
(12.69)
(5)
2.00
(0.84)
(5)
37.56
(27.21)
(7)
21.83
(14.14)
(3)
Tv;>o 11
10.60
(5.45)
(5)
10.33
(4.63)
(2)
1.50
(1.50)
<:)
29.00
(9.34)
(5)
18. 75
(6.57)
(4)
11.00
(i.OO)
(3)
20.40
(5.89)
(5)
(continued)
-------
Table C.I
(cone i uued)
Area
Culver City
-------
Table C.1
(continued)
Area
Palos Verdes
(C -* C*)
Redondo Bcarh
(C •* C*)
Mean Aesthetic Bid
($/month)
Ty»c I
/4.31
(2.10)
CO
6.29
(3.39)
(7)
Tyoe II
0.50
(0.50)
(!)
,.29
(4.29)
(7)
Mean Acute Health Bid
($/monthj_
Tvpe I
2.19
(1.29)
•'..86
(2.13)
(7)
Type II
17.75
(10.96)
15.29
(7.85)
(7)
Mean Chr(
Tvpe I
1. 75
O-l-fi)
CO
0. 00
(0.00)
(7)
ronic Heal th Hid
Type II
0.50
(0.50)
CO
', . •'. 3
(2.9-'0
(7)
Mean Total Bid
($/r,onth)
Ty :> e I
8.25
C-.il)
11 .14
(5.07)
(7)
TV we II
1 S . 7 5
(10. '.S)
C-)
(11 .12)
*The implicit assumption in this table has been that of strict addl ::ivity of bids for each ai.r qualify cffoct.. In obc a i.ning the "'.car.
bids, differentiation has been made vith respect to: (1) the completion i!aue oi cleanup; (2) the biddinc, sequence. IP. "Type I" questionnaires,
the air quality effects; are introduced in "Aesthetic -' Acute iiealtii - Chronic Health" order. In "Type li" questionnaires, the air quality
effects; are introduced in "Acute ::eal t'n •+ Ciironlc Health -» Aesthetic" order. In obtaining the mean bids, no differentiation has been !?ace
with rer.pect to: (1) different: proposed vehicles for the collection of bids; (2) whether a health pamphlet has or has not been sent to the
respondent in advance of the i:-.terview; and (3) whether a life table has or has not been shown to the respondent during the interview. A life
table depicts the "stock" counterparts of the elicited monthly bids for various expected lifespans.
**The notation in parentheses represents the change in air quality for which the respondents are bidding. For example, (A -*• 8) denotes
that the respondent is bidding to change air quality fro:r. poor to fair, (B -*• C) denotes that the respondent is bidding to change air quality
from fair to good, and (C -» C*) denotes that the respondent is bidding to change air quality to good across t're entire region.
***Scandard error of the nean bid in all cases.
****Sample size of each case in all cases.
-------
Table C.2
Mean Bids by Area by Type*
(Completion Dace of Cleanup: 10 yrs.)
Area**
El Monte
(A - B)
El Nonce
(A H. C)
La Canada
(A -» B)
La Canada
(A -» C)
Montebello
(A - B)
Montebello
(A -» C)
Kean Aesthetic Bid
($/ir.onth)
Type I
7.50
(7.50)***
(2) ****
5.00
(1.53)
(5)
6.33
(2.54)
(6)
16.83
(8.47).
(6)
4.00
(2.45)
(5)
6.00
(1.97)
(5)
Typo II
0.00
(0.00)
(7)
0.00
(0.00)
(1)
0.00
Co. oo)
(4)
0.00
(0.00)
(1)
1.40
(0.98)
(5)
1.67
(1.67)
(3)
Mean Acute Health Bid
(|/cionch)
Type I
2.50
(2.50)
(2)
6.40
(3.53)
(5)
15.83
(14.84)
(6)
4.17
(2.39)
(6)
12.40
(9.58).
(5)
2.20
(1.11)
(5)
Type II
14. 57
(6.49)
(7)
0.00
(0.00)
(1)
15.25
(11.80)
(4)
50.00
(.0.00)
(1)
5.20
(1.56)
(5)
0.00
(0.00)
(3)
Mean Chronic Health Bid
(S/month)
Type I
0.00
(0.00)
(2)
1.40
(0.98)
(5)
O.S3
(0.83)
(6)
0.50
(0.50)
(6)
1.80
(0.97)
(5)
1.20
(0.80)
(5)
Tvnc II
1.43
(1.43)
(7)
0.00
(0.00)
._(!)
22.50
(22.50)
(4)
0.00
(0.00)
(1)
1.60
(1.03)
(5)
0.00
(0.00)
(3)
Mean Total Bid
($/month)
Type I
10.00
(10.00)
(2)
12.80
(2.54)
(5)
23.00
(15.56)
(6)
21.50
(7.96).
(6)
18.20
(11.39)
(5)
9.40
(2.86)
(5)
Tvpe II
16:00
(6.33)
(7)
0.00
(0.00)
(1)
37.75
(23.81)
(4)
50.00
(0.00)
(1)
8.20
(1.80)
(5)
1.67
(1.67)
(3)
(continued)
-------
Table C.2
(continued)
Area
Canoga Park
(B •* C)
Culver City
(B H. C)
Encino
(B - C)
Huntington Beach
(B -> C)
Irvine
(B -> C)
Newport Beach
(B - C)
Mean Aesthetic Bid
Typo I
5.58
(1.64)
(6)
12.88
(6.30)
(8)
6.00
(3.21)
(6)
16. 38
(5.33)
(8)
13.08
(5. 12)
(9)
•1.10
(0.51)
(5)
Type II
0.95
(0.95)
(5)
7.50
(4.79)
(4)
0.00
(0.00)
(5)
5.75
(3.43)
OD
4.33
(2.96)
(3)
1.20
(1.20)
(5)
Menu Acute Health Bid
(S/mop.th)
Typo I
0.92
(0.58)
(6)
7.25
(3.51)
(8)
0.50
(0.50)
(6)
7.63
(2.05)
(S)
7.72
(1.97)
(9)
0.20
(0.20)
(5)
Tvne II
3.45
(1.86)
(5)
11.13
(3.23)
CO
11. 20
(6.32)
(5)
12.90
(4.30)
(12)
19.00
(15.63)
(3)
3.80
(1.91)
(5)
Menu Chronic Health Bid
(S/month)
Mean Total Bid
($/510Ht!l)
Type I 1 Tvpe II 1 TVPC I
0.88
(0.58)
(6)
4.50
(3.02)
(8.)
1.00
(0.6.3)
.. . .(&)
6.88
(4.20)
(8)
2. 58
(1. 15)
(9)
0.40
(0.40)
(.5)
o.no
(0.00)
(5)
3.25
(2.36)
i-'O
2.00
(1.22)
(5)
6.82
(2.94)
01)
0.00
(0.00)
(3)
6. 50
(3.83)
(5)
7.38
(1-75)
(6)
24.63
(12.07)
(8)
7. 50
(3.10)
(6)
30.88
(S.31)
(8)
23.39
(6.03)
(9)
1.70
(0.54)
(5)
Tvne II
4.40
(1.69)
(5).
21 .88
(9.21)
(4)
13.20
(7.26)
(5)
26.64
(S.52)
23.33
(18.56)
(3)
11.50
(5.04)
(5)
(continued)
-------
Table C.2
(continued)
Area
Pacific Palisades
(C - C*)
Palos Verdes
(C - C*)
Redondo Beach
(C •* C*)
Mean Aesthetic Bid
($/sonth)
Type I
7.40
(4.21)
151
7.29
(1.38)
cn
20. 64
(6.46)
(7)
Tvne II
4.29
(2.30)
(7)
2.00
(1.22)
(4)
1.00
(0.77)
(5)
Htiiin Acute Health Bid
($/mOnth)
Tvnc I
13.60
(5.87)
(5)
7. 71
(5.45)
(?)
9.57
(6. 83)
(7)
Tvse II
15.71
(4.56)
(7)
22.50
(9.46)
(4)
3.30
(1.76)
Moan Chronic Health Bid
($/month)
Tvoe I 1 Tvpo II
139.20
(87.30)
(5)
1.14
(0.77)
(?)
4.14
(2.26)
(7)
2.S6
(1.49)
(7)
5.50
(4.S6)
(•'•)
4.80
(3.01)
(5)
Mean Total Bid
($/r,onth)
Type I
160.20
(92.16)
(5)
16.14
(5.91)
(7)
34.36
(14.35)
(7)
Tv»c II
22.36
(4.21)
(7)
30.00
(20.21)
CO
9.10
13.85)
(5)
. . *The implicit assumption in this table h.'.s been that of strict additivity of bids for each air quality effect. In obtaining the mean
-j bids, differentiation has been made with respect to: (1) the completion date of cleanup; (2) the biasing sequence. In "Type I" questionnaires,
W the air quality effects are introduced in "Aesthetic •» Acute Health -» Chronic Health" order. In "Type II" questionnaires, the air quality
effects are introduced in "Acuta ilealth •* Chronic Health •» Aesthetic" order. In obtaining the mean bids, no ailferenciation has been zade with
respect to: (1) different proposed vehicles for the collection of bids; (2) whether a health pamphlet lias or has not been sent: to the
respondent in advance of the interview; and (3) whether a life table has or has not been shown to the respondent during the interview. A life
table depicts the "stock" counterparts of the elicited monthly bids for various expected lifespans.
**The notation in parentheses represents the change in air quality for which the respondents are bidding. For example, (A •+ B) cenotes
that the respondent is bidding to change air quality from poor to fair, (B •* C) denotes that Che respondent is bidding to change air quality
froa fair to good, and (C * C*) denotes that the respondent is bidding to change air quality to good across che entire region.
***Standard error of the mean bid in all cases.
****Samplc size of each case In all cases.
-------
APPENDIX D
Preliminary Regression Relationships for Selected Variables
on the South Coast Experiment
Introduction
The following Tables present a very preliminary set of regression
results on examining the raw data for bid relationships in the South Coast
Air Basin. These data sets must be viewed as a preliminary set to give the
researchers further guidelines on how to statistically analyze the data set.
They are not meant to be viewed as definitive in either a computational or
final set sense. However, they should indicate to other researchers the
degree of variation in both the estimates and the sets of relationships
hypothesized for computation. It is anticipated that it will take a least
four to five months for all such relationships to be adequately statisti-
cally analyzed.
Table D-l contains a preliminary set of regression equations across all
areas and bid types. Aesthetic, acute and chronic health bids, along with a
total bid, were regressed against various variables of possible interest.
One of these variables was the interviewer, to find out whether a detectible
bias might exist in terms of the interview selected. In most instances, no
interviewer bias was discovered; however, for the acute health bid, chronic
health bid, total bid, when related to a small number of variables, there
was an indication of a detectible interviewer bias. The researchers will
continue to explore this possibility to discover whether, in fact, such a
bias is present and how it might be removed from further statistical com-
putations. A further test was to examine whether years of education in some
significant way influenced the amount of the bid. In no circumstances was
a significant relationship (at the 95% level of confidence) discovered. A
third possible premise was that the duration of years lived in Los Angeles
would influence the bid. The results here are mixed, although in almost all
circumstances, statistically nonsignificant. Both positive and negative
effects of years living in Los Angeles was discovered. Finally, as a general
variable to examine, individuals who had read the health pamphlet and those
who had not were examined. Again, the results were mixed. However, in each
circumstance, those who had read the health pamphlet tended to bid signifi-
cantly higher (at the 95% level) than those who had not. Alternatively, the
bids on aesthetic and chronic health effects appeared to not be related in
any reasonable way to whether the individuals had, in fact, had access to
additional information on health effects.
Dummy variables were inserted for each of the locational sites of the
experiment. In almost all circumstances, with a few exceptions, the site-
specific dummy variables were nonsignificant, indicating at least in a
preliminary way that site-specificity would not significantly influence the
bid. The pollution variable in every circumstance but one was insignificant
at the 95% level of confidence. This would be anticipated on the basis of
the conceptual research reported in Chapter 2. That is, when one nets out
all the effects on the various bids with the exception of pollution, inclusive
of income, then the pollution variable itself may or may not be statistically
significant. For example, if preferences are nonhomogeneous and nonidentical,
we could presume that those willing to pay a higher price for clean air
179
-------
H'muY trcmsiojt WAIIOXS lea »
CO
O
. . —
,
• i.iU;;+ 1 c**wu*t tc Tt LJI
j
j ; lrl»l . (O.*tl) { .10))
1 &ml«».«!lfcSU -It-lli j on* ' l!J . 711
| * f).6-V») . I 5il) ( . y,tj
' T.ut MJ -W.JJI ! 0.1*0 : .M» - .K'
10.111) i itt) ( .*•»*>
«Mf,*ilc lit 1. tit -O.JJ3 • .!}) - .fit-;
1:11 ' ' CO **> C i**t ( ' Jill
«e»t. B..-,t, in -> oil a 1M -«» • 64[ '
(0,1)0) • ( .JC4) (O.J)»
(o.nn ( .011) .o:i
(t.ilSl
:i
I .i*t-*»
{ .nn>
.»J5-J
( .««-»)
II
( .i«t- J
( .3*1- )
- .i*i-
( . Jfi- f
'„
( .ut- j
t .111- )
< .^'*- t
;!:;;!::,
-l.B t-*
O.) £-*)
J.. /.-J
•J.I I-*
(l.J f-JJ
d f
»
J.6U-
(t.)iE. >
U.l^t* )
J.Wt-*
(J.t«-4)
**
J.OH-
(i.e:t- )
-*.»«-
<*.*'!- )
•*.m-
l».tu. )
:, 1 ' ,.
.; i- i .jjt-*
(.it-)] < . ilt-H
{.»(-> ( .WI-4J
. 0 t- ' .lit-}
( .1 *- ) j 1 .*it-4)
1
j
'
"
-Jt(-*
( .Ut-»)
{ ,!«'*>
.».-.
C .!"-•!
:
-Mil-
(t.)lt- }
-LltE-*
t
: = , !
( .m-*i
,:oi-*
{ .»»i-»)
I . !>l^*)
.04(_)
-
'
d.i»
:.)i
:. js
10. >9')
- . IM
• , 1M
, ( .Ui>
-.,,.
— .
-7"1
O.'V
(!.*" •"
0 I*
3.J.*
-
"
I.1W
: i )£•!
i.ijoo
.
r)
i.n*
:. 1:3
*;.)«
—
'
(>.t*r->
:.)Ei-«
<'. \$t- i
I.JJi-*
1
* '
no '
,,0
no
,,«,
110
1-.6
'"
1
1110
!
mo
1
< !
^— — i
o :* | i; i»
1
„.:> | .: .»
o. ;> .> MJ j
e. )i >c **j 1
o. 10
,.,» .,», j
i
!
o.ii u.ft*f
i
1 I
»U. » »«M«J» ****. «) »**i
-------
TABU. 0-2
PRELIMINARY REGRESSION EQUATIONS KOR THE AGGREGATED "A" AND "B" AREAS
00
Dependent
Variable
Aesthetic Bid
Acute Health
Bid
Chronic Health
Bid
Total Bid
AeftthctU Bid
! Acute Health
Bid
Chronic Health
Bid
Tot.il Bid
log. (Aesthetic
Bid)
log (Acute
Health Bid)
log (Chronic
H.alth Bid)
log (Total Bid)
Constant
8.682
1.327
-0.896
11.112
5.153
11.94;)
0.205
17.301
j.«;s
0.387
1.597
3.447
Independent Variables
LA
-5.66E-2
(B.769E-2)
-2.47E-2
(7.59E-2)
-8.12E-2
(0.111)
-0.162
(0.166)
HP
0.662
(1.161)
1.166
(1.005)
2.745
(1.466)
4.573
(2.203)
1 tf
-5.53E-5
(6.79E-5)
1.49E-4
(5.87E-S)
1.34E-4
(8.57E-5)
2.27E-4
(1.29E-4)
0.145
(0.600)
-0.442
(0.518)
1.231
(0.757)
0.934
(1.136)
5.953
(P1.245)
-21.849
(19.020)
13.246
(26.681)
-2.650
(41.657)
6P2
-0.562
(2.110)
2. OSS
(1.889)
-1.404
(2.650)
0.123
(4.138)
YAP
-9.53E-6
(2.31E-5)
2.53E-8
(2.07E-5)
7.06E-5
(2.90E-5)
6.11E-5
(4.52K-5)
LLA
-0.263
(0.386)
-0.248
(0.179)
-0.343
(0.168)
-n.283
(0. 186)
LHP
-4.72E-?
(0.390)
0.191
(0.377)
0.387
(0.354)
0.223
(0.391)
LY
-0.159
(0.110)
0.152
(0.107)
-5.48L'-:
(0.100)
-4. HE-:
(0.111)
LAP
7.70EI-2
(0.155)
-2.-i3r.-2i
(0.150)
8.82E-2
(0.141)
S.25E-2
'0. 156)
N
75
?5
15
75
75
75
75
75
T,
JS
75
75
R2
0.02
0.12
0.10
0.11
0.00
O.OS
0.09
0.02
0.06
O.OJ
0.03
0.04
S£
10.385
8.986
13.112
19.698
10.400
9.311
13.061
20.392
1.271
1.227
1.153
1.275
Independent variables:
LA - Years lived In L.A.
HP - Amount of health paoplet read
/.P, - Ch.lnjc In pollution level. I.e. ISO
change in the sqoared
Ion Itvela, i.e. 1/2(P^-P
hafige in pollution level
1th pnrr.phlet read)
£P~ - On.' hoi/ times th
values of polio
LLA - loj (Vo.irs lived
LHP • !(-£ (A.-OU.U of he
L.'-P - log (Chan3f In pollution level)
-0-
S, - N-_iber of cases
R" - CooJ.ipss of fit
SE - Standard error of regression
(E-n •* 10""; I.e. E-2 •> 10"
Observations are aggregated without any differentiation with
retpect to 1) Bidding sequence, 2) Starting bid. 3) Vehicle
used, 4) Health pamphlet vs. no health pamphlet, 5) Ufa
table vu. no life Cable.
Bids for each eir quality effect are assumed to be strictly separable.
-------
TABLE 0-3
PRELIMINARY REGRESSION EQUATIONS TOR THE PAIRED AREAS*
(S'EWPOST BEACH-PACIFIC PALISADES)
CO
Ni
Dependent
Variable
Aesthetic Sid
Acute Health Bid
Chronic Health Btd
Total Bid
log (Aesthetic
31d)
log (Acute Health
Bid)
log (Chronic
Health Bid)
log (Total Bid)
Constant
-49.578
-30.484
179.120
99.057
-21.328
-5.907
0.184
-3.303
1 -
Iniiirpctuient Variable
LA
0.728
(0.735)
0.451
(0.812)
-5.765
(5.386)
-4.586
(5.837)
HP
13.773
(4.634)
13.777
(5.397)
-26.756
(35.810)
0.794
(38.808)
V
7.89E-4
(2.94C-4)
5.713
(3.25E-4)
-1.12E-4
(2.15E-3)
1.25E-3
(2.33E-3)
j
-M'
-41.709
(26.624)
-44.660
(29.421)
24.615
(195.197)
-61.754
(211.543)
LLA
0.155
(0.861)
-0.118
(.-6.06.1:-?'
-1.657
(1.263)
-0.956
(0.902)
LH? w
0.795
(0.856)
1.144
(0.632)
-O.S50
(1.256)
0.263
(0.69(0
2.067
(1.146)
0.793
(0.847)
0.624
(1.687)
0.346
(J.;OD
UP
3.965
(2.626)
5.468
(l.'J40>
-0.164
(3.854)
3.376
(2.750)
N
13
13
13
1
13
! i3
!
;
R2
0.65
0.57
0.2?
0.27
0.54
13 I 0.66
j
1
13 ! 0.43
,
1
13 j 0.57
1
i_
SE
19.466
21.512 .
142.720
154.672
l'.235
0.912
1.812
1.293
Independent vari^bleo:
LA - Years lived in L.A.
HP - Anount of health pamphlet read
Y - Incoae
-P - Chnnpe In pollution level, i.e. AttOj
LLA • log (Yenrs lived in L.A.)
LHP - lop (Arount of licalch ponphJet resd)
LY - log (Incooc)
LiP -. log (Cltun&tt tn pollution level)
-0-
K « Xunber of cases
R " Goodness of fit
SE - Standard of error regression
Values In paroncheiiCS arc coi*f f iclenc standard errora
(E-n •> 10~n; I.e. E-2 •* 10" )
Observations are .ip,j;r«:K)ittd without any differentiation with
respect to 1) Bidding sequence, 2) Scartinfc bid, 3) Vehicle
uufcd, A) Health p.itnphiet vs. no health paephleC, 3) Life cab
ve. no life tabli.'}
Bids for each air quality effect are assumed to be strictly Be
-------
TABLE D-*
PRELIMINARY REGRESSION EQUATIONS TOR THE PAIRED AREAS
(IRVINE-PALOS VERDES)
«, >>. c, d
GO
LO
Dependent
Variable
Aesrhe'ic Bid
Acute Health Bid
Chronic Health Bid
Total Sid
Log (Aesthetic Bid)
Log (Chronic Health
Bid)
Log (Total Bid)
Constant
1.086
6.420
0.332
7.838
-8.095
-1.046
-4.534
-6.527
Independent Variables
LA
-8.48E-2
(6.58E-2)
0.187
(0.191)
-1.29E-2
(5.67E-2)
8.90E-2
(0.201)
HP
0.457
(1.014)
3.055
(2.942)
0.197
(0.873)
3.709
(3.102)
Y
1.01E-4
(4.65F.-5)
-2.48E-5
(2.80E-4)
3.13E-5
(8.31E-5)
1.071
(2.95E-4)
OP
2.661
(3.231)
-3.591
(9.375)
2.252
(2.782)
1.322
(9.8861
I.LA
-0.340
(0.205)
-4.48E-2
(0.245)
-0.132
(0.161)
-4.99E-2
(0.183)
LHP
0.145
(0.688)
0.516
(0.823)
0.448
(0.541)
0.468
(0.613)
LY
0.940
(0.602)
0.265
(0.720)i
0.505 i
(0.473):
LAP
-0.517
(1.186)
0.330
(1.418)
-0.506
(0.93J)
0.862 -0.205
(0.536)1 (1.057)
N
26
26
26
26
26
26
26
26
R2
0.15
0.08
0.04
0.07
0:22
0.04
0.13
0.16
SE
4.446
12.900
3.828
13.603
1.083
1.295
0.852
0.965
Independent variables:
LA - Years lived In L.A.
HP - Aaount of health panphlet read
Y - Income
£P - Change in pollution level, i.e. ANO
LLA - log {Years lived In L.A.)
LHP - log (Acount of health pamphlet read)
LY - log (Incoff.e)
L.1P - log (Change in pollution level)
-0-
N - Suaber of cases
R - Goodness of fit
SE - Standard of error regression
Values in parentheses are coefficient stAndard errors
(E-n * Itf"; I.e. t-2 •* 10 )
C Observations arc aggregated without any d I f f erenc lat Ion uUh
respect to 1) Bidding sequence. 2) Starting bid, 3) Vehicle
used. 4) Health pamphlet vs. no health pamphlet, 5) Life table
vs. no life table)
Bids for each air quality effect are assumed to be strictly tep*rjbl«
-------
TABLE D-5
PRELIMINARY REGRESSION EQUATIONS FOR THE PAIRED AREAS8' b' c> d
(LA CANAilA-ENCINO)
Dependent
Variable
Aesthetic Bid
Acute Health
Bid
Bid
Tc>tal Bid
Aesthetic Bid
Acute Health
Bid
Chronic Health
Bid
Total Bid
log (Aesthetic
Bid)
log (AcuLe
Health Bid)
IOR (Chronic
Health Bid)
log (Total Bid
Constant
19.224
2.220
-4.03E-2
21.384
12.099
6.569
-5.559
13.101
5.589
0.770
4.477
6.322
LA
-0.225
(0.229)
-3.70--2
(0.165)
-0.280
(0.372)
-0.542
(0.420)
Independent Variables
HP
-3.122
(2.838)
3.066
(2.049)
6.971
(4.611)
6.915
(5.207)
t
-1.72E-4
(1.14E-4)
2.37E-4
(8.26E-5)
1.30E-4
(1.86E-4)
1.95E-4
(2.10E-4)
AP
0.977
(1.298)
-1.581
(0.937)
3.005
(2.109)
2.4111
(2.382)
-30.4B9
(48.614)
24.025
(40.736)
29.340
(80.346)
22.877
(94.111)
AP2
3.257
(4.872)
-2.589
(4.083)
-2.771
(8.053)
-2.103
(9.433)
YiP
-5.62E-5
(4.27E-5)
1.94E-5
O.53E-5)
5.30E-5
O.05E-5)
1.62E-5
(8.26E-5)
LLA
-0.640
(0.582)
-0.408
(0.572)
-0.732
(0.632)
-0.910
(0.681)
LHP
-1.249
(0.759)
0.717
(0.745)
0.413
(0.82(.)
0.335
(0.887)
LY
-0.246
(0.126)
0.171
(0.124)
-0.160
(0.137)
-0.113
(0.147)
LAP
0.246
(0.287)
r0.247
(0.282)
0.486
(0.312)
0.184
(0.336)
N
22
22
22
22
22
22
22
2:
22
22
22
22
R2
0.19
0.42
0.24
0.23
0.10
0.12
0.13
0.04
0.29
0.20
0.20
0.13
SE
13.241
9.559
21.514
24.292
13.574
11.375
22.435
26.279
1.323
1.300
1.437
1.543
Independent variables:
LA
HP
Y
t?
LT
YSP
LLA
LHP
LY
LiP
R
SC
- I
- Ch
- 0"
rs lived In L.A.
nnt of health paciplet read
e In pollution level, i.e. AUG.
olf c tuics the cbonp.e in the squared. .
ucs of pollution levels, i.-e. l/2(P^-Pp
c cir.es the change in pollution level
- log (Years lived in L.A.)
• log (Anount of health pamphlet read)
• IJR (Incoce)
- log (Change in pollution level)
-0-
•• S'u:i!»cr of cases
» Coodneii of fit
• Standard error of regrtsalon
Values In parenthu^es are coeff ieicnt standard errors
(E-n * l(f °; I.e. E-2 •*• 10" )
Observations are aggregated without any differentiation vith
respect to 1} Bidding sequence, 2) Starting bid, 3) Vehicle
used/ 4) Health pamphlet vs. no health pamphlet, 5} Life
table vs. no life table.
Bids for each air quality effect are assumed to b« atrictly »
-------
TABLE n-h
PRELIMINARY REGRESSION EQUATIONS TOR THF. PAIRED AREAS*' b> C' *
OS BEACH-REDONDO BEACH)
Dependent
Variable
• Aesthetic Sid
' Acute Health Bid
Chronic Health Sid
Total Bid
log (Aesthetic Bid)
log (Acute Health Bid)
log (Chronic Health Bid)
log (Total Bid)
Constant
-11.181
-8.864
-2.539
-22.58'.
-8.499
-8.424
-2.636
-8.980
Independent Variables
LA
0.201
(0.367)
0.171
(0.305)
0.120
(0.237)
0.491
(0.70-4)
HP
1.577
(2.992)
-1.180
(2.491)
0.284
(1.934)
0.480
(5.741)
V
5.34E-4
(3.79E-4)
5.47E-4
(3.15E-4)
1. 30E-4
(2.45E-4)
1.21E-3
(7.27E-4)
1
I
1
j
ar
7.166
(11.006)
0.764
(9.212)
9.297
(7.155)
15.700
(21.233)
LLA
0.330
(0.685)
0.793
(0.592)
0.285
(0.645)
0.588
LHP
-3.MF.-2
(0.749)
-0.671
(0.647)
0.332
(0.705)
-0.30B
(0.694) (0.758)
I
LY
O.S'A
(0.618)
0.774
(0.534)
0.240
(0.532)
0.932
(0.626)
LAP
-1.678
(1.578)
-1.096
(1.364)
-1.077
(1.485)
-1.474
(1.593)
N
73
23
23
23
23
23
23
23
7
R
0.24
0.17
0.14
SE
16.054
13.365
10. J80
0.25 1 30.804
1
0.20 ! 1.522
0.18
0.09
0.19
i
1.316
1.433
1.5U
ndependenc variables:
LA - Years lived In L.A.
HP - Amount of health paapMet read
A? - Change in pollution level, I.e. &KO
LLA - log (Years lived In L.A.)
LH? - log (Amount of health pamphlet rend)
LY - log (Income)
LAP - Ion (Clisr.Ji! In pollution level)
-0-
N - Kus:ber of cases
R " Coodneus of fit
St. " Standard oi error regression
Vflluca In parL-nthcscs ore coefflcitnt standard errors
(E-n -> 10"°; i.e. £-2 •» 10" )
Ob«erv.ir lona arc *ij;r>reSated without any dif ferentlation uith
respect to 1) Bidding sequence, 2} Starting bid, 3) Vehicle
used, 4) Health pamphlet vs. no health paophlet, 5) Life table
VB. no life table)
bldtj for each sir quality effect are assumed to be strictly sepArable.
-------
TASU
PRELIMINARY REGRESSION EQUATIONS FOR HIE PAIRED AREAS
(WNTEBELtO-rULVER CITY)
KAS'- "' C'
OO
Dependent
Variable
Aesthetic Bid
Acute Health Bid
Chronic Uraleh
Bid
Total Bid
Aesthetic Bid
Actute Health
Bid
Chronic Health
Bid
Total Bid
log (Aesthetic
Bid)
log (Acute
Health Bid)
log (Chronic
Health Bid)
log (Total Bid)
Constant
14.371
13.877
13.164
41.611
2.018
17.139
6.149
25.306
22.984
8.958
19.392
25.998
Independent
' LA
-3.40E-2
(0,103)
-8.32E-2
(0.141)
-1.27E-2
(4.5E-2)
-0.130
(0.190)
HP
-1.277
(2.042)
-2.853
(2.800)
-1.229
(0.892)
-5.359
(3.760)
Y
-2.77E-4
U.52E-4)
-1.55E-4
(2.09K-4)
-2.33E-4
(6.66E-5)
-6.65E-4
(2.B1E-4)
4P
-0.25*.
(1.042)
-0.313
(1.428)
-1.150
(0.455)
-1.719
(1.91H)
17.058
(19.864)
-55.831
(20.201)
-8.323
(12.024)
-47.096
(39.363)
4P2
-1.532
(1.970)
5.574
(2.003)
0.815
(1.192)
4.857
(3.963)
Variables
YAP
-7.31E-5
(5.92E-5)
-3.41F.-S
(6.02E-5)
-2.48E-5
(3.59E-5)
-1.32E-4
U.19E-4)
1XA
6.11E-3
(0.639)
-1.189
(0.664)
-8.92E-2
(0.423)
-0.518
(0.683)
LUt
-0.701
(1.185)
-2.123
(1.232)
-1.349
(0.734)
-1.718
(1.266)
LY
-2.133
(1.008)
-0.423
(1.048)
-1.773
(0.667)
-2.183
(1.077)
LAP
-0.294
(0.533)
0.541
(0.554)
-0.540
(0.353)
-0.159
(0.570)
N
14
14
14
14
14
14
14
14
14
14
14
14
IT2
0.30
0.17
0.66
0.46
0.18
0,47
0.25
0.25
0.36
0.36
0.53
0.40
SE
S.139
7.045
2.244
9.461
5.256
5.345
3.181
10.573
1.078
1'.120
0.713
1.152
Independent var 1 ahles :
LA
HP
Y
YAP
LLA
LHP
LY
L.iP
R
SE
• Y«Mrs lived In L.A.
- Amount of health pamplet read
• Income
• Chang* In pollution level. i.e, QKQ
- One half timer, the chance In the squared
values of pollution levels, i.e. 1/2(P -P )
- ncome times the change* In pollution level ^
• og (Years lived In L.A.)
- og (Anount of health pamphlet read)
- og ( Incorio)
- og (Change in pollution level)
-0-
• Nuober of cases
- Coodneaa of fit
- Standard error of regression
Values In parentheses aro coefficient standard* erron
(E-n * 10" ; I.e. E-2 * It)" )
Observations are aggregated without any differentiation with
respect to 1) Bidding sequence, 2) Starting bid. 3) Vehicle
used, A) Health pamphlet vs. no health pamphlet, 5) Life
table vs. no life table.
Bids for each air quality effect are uat>u
-------
REGRESSION' EIJUATIOSS VOR Tin; VAIRKD AREAS'
(Kl. MONTt'-CASOCA PARK)
.1. b. c. d
CO
Variable
Aesthetic Bid
Acute Health
BU
Chronic Health
KM
Total Bid
Aesthetic Bid
1 Acute Health
B!
-------
years lived in L.A., amount of health pamphlet read, income and change in
pollution indicating a reasonable relationship except for, perhaps, the
sign on the number of pages read of the health pamphlet. The income var-
iahle is highly significant as is the years of residence in Los Angeles.
However, only after substantial further experimentation on these pairs can
we anticipate that reasonably defensible estimates of coefficients or
elasticities will be forthcoming.
138
-------
Conclusion
In this Appendix, we have attempted to indicate the. rough, orders of
magnitude of variability of relationships- between ob'served Kids and some
variables of interest. As yet, the regression results- Rave only roughly
illuminated possible further zones of research. Both signs and magnitudes
seem to be highly insignificant when the data set is regressed totally.
Thus, it is anticipated that substantial additional research from a statis-
tical perspective and also incorporating x^ell-defined theoretical hypotheses
will need to be developed for this data set to be adequately exploited. Of
particular importance is the examination of bias effects and disaggregation
down to the paired comparisons. For our first estimate of the magnitude of
bid in Los Angeles reported in Chapter 6, we selected the last equation
total bid in the preliminary regression equations in Table D.I. From the
coefficient for pollution and adjusting for the effect of the health pam-
phlet on bids along with adjustments for capital recovery facto.rs and the
length of time to achieve clean air, the numbers reported in Table D.I of
Chapter 6 were obtained. The researchers believe this is only a preliminary
estimate of the value of the average Bid for Los Angeles. It is anticipated
that further research will have a highly significant impact on ultimate
calculation of a reasonable, accurate value for citizens' preferences
associated with improved air quality.
189
-------
APPENDIX E.
Thia appendix presents the variable list for the non-market valuation
experiment in the South. Coast Air Basin.
190
-------
Variable
Name
MVFA
APLV
APWK
AZLV1
AQLV2
AQWK1
AQWK2
ZHAZ
RESCOP
RESEV
RESEN
RESSP
RESUN
RESGM
RESVAL
PCTIN
INTTM
DTINT.
PSI
EQBL
Description
Would you move if like A everywhere (l=yes)
Air pollution influenced where you live (l=yes)
Air pollution influenced where you work (l=yes)
Air quality where live (Good=00; Fair=01;
Poor=10)
Air quality where work (Good=00; Fair=01;
Poor=10)
Are you aware of any health hazards of air
pollution (l=yes)
Respondent cooperative (l=yes)
Respondent evasive (l=yes)
Respondent enthusiastic (l=yes)
Respondent suspicious (l=yes)
Respondent understanding (l=yes)
Respondent playing games (l=yes)
Respondent giving true value (l=yes)
Percent work time indoors
Minutes taken for formal interview
Date of interview (1=365)
Pollution index by location and date
Cost/person/month for air cleanup if all
Column (s)
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55-57
58-60
61-63
64-66
Format
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
F3.0
F3.0
F3.0
F3.0
billed equally
APCLUP Total figure for cleanup of pollution
(100,000's of dollars)
67-72
73-E
F6.0
F8.0
1.91
-------
Card //I: Socioeconomic Information and Enumerator Evaluation
Variable
Name
QNM
CRONM
QTYP
QCC
INTCD
TIKE
WORK
AGE
NPER
YRED
LIVLA
PLLVLA
LOCEMX
LOCEMY
LOCLVX
LOCLVY
ADDCON
SEX
MARST
DEC
Description
Questionnaire number
Card number
Questionnaire type: 1 = aesthetic first; no
health; 2 = aesthetic first, health;
3 = acute, no health; 4 = acute, health
City code
Interviewer code
Time spent at leisure (hours per week)
Time spent at work (hours per week)
Age of respondent
Number of persons in household -
Years of education.
Years lived in LA area
Years plan to live in LA area
X-coordlnate, location of employment
Y-coordinate, location of employment
X-coordinate, location of home
Y-coordinate, location of home
Additional conversation time (minutes)
Sex of respondent (l=male)
Marital status of respondent (l=married)
Highest degree obtained (1=K.S.; 2=Coll.;
Column(s)
1-3
4-5
6
7-8
9-10
11-12
13-14
15-16
17-18
19-20
21-22
23-24
25-26
27-28
29-30
31-32
33-34
35
36
Format
F3.0
F2.0
Fl.O
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
Fl.O
Fl.O
r* 1 r\
ENHAZ
PAMP
3=Voc.; 4=Adv.; 0=no degree)
Environmental hazards associated with job
(l=yes)
37
38
How much of pamphlet did you read? (1=0-5'pages;
2=5-10 pages; 3=10 + pages; 0 = did not receive) 39
Fl.O
Fl.O
Fl.O
192
-------
Card 112: Bidding Game and Secret Ballot
Variable
Name
QNM
QTYP
CRONM
STBID
CPDT
ZMV1X
ZMV1Y
ZMV2X
ZMV2Y
ZMV3X
ZMV3Y
MAXBD1
MAXBD2
MAXBD'3
VEH
LFCK
MV1
MV2
MV3
ZOTVEH
SB11
SB16
SB21
SB29
Description
Questionnaire number
Questionnaire type (l=aes. + no health;
2=aes. + health; 3=acute + no health;
4=acute + health)
Card number
Starting bid
Completion date of cleanup
Where you would move, 1st stage, X-coordinate
Where you would move, 1st stage, Y-coordinate
Where you would move, 2nd stage, X-coordinate
Where you would move, 2nd stage, Y-coordinate
Where you would move, 3rd stage, X-coordinate
Where you would move, 3rd stage, Y-coordinate
Maximum bid, 1st stage
Maximum bid, 2nd stage
Maximum bid, 3rd stage
Vehicle used (l=monthly charge; 2=utility bill)
Life table checked (1-yes)
Would you move, 1st stage (1-yes)
Would you move, 2nd stage (1-yes)
Would you move, 3rd stage (1-yes)
Is there another vehicle? (1-yes)
Secret ballot; question 1, bracket 1
Secret ballot; question 1, bracket 6
(l=checked ; 0=unchecked)
Column (s)
1-3
4
5-6
7-8
9-10
11-12
13-14
15-16
17-18
19-20
21-22
23-27
28-32
33-37
38
39
40
41
42
43
44
49
50
58
Format
F3.0
Fl.O
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F2.0
F5.0
F5.0
F5 . 0
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
193
-------
Card #2 (continued)
Variable
Name Description Coluinn(s) Format
SB31 59 Fl.O
SB37 Secret Ballot; question 3, bracket 7 65 Fl.O
SB41 (l=checked; 0=unchecked) 66 Fl.O
SB46 71 Fl.O
RVBD1 Reverse bid to B from C 72-75 FA. 0
RVBD2 Reverse bid to A from B or C 76-80 F5.0
-------
CARDS 3, 5, 7, 9, . . ., 67
ACTIVITIES: OUTDOOR THEN INDOOR
Variable .
Name
QNM
QTYP
CRDNM
10PT
10IMP
10TMI
10SBB
10SBC
10SBD
10TMSB
10FQSB
10LCSB
100TSB
10PD
10EQX
10HRA
10FQA
10LCXA
10LCYA
10MIA
10DCA
Description Column(s)
Questionnaire number
Questionnaire type
Card number
Activity participation (l=yes)
Importance column checked (l=yes)
Was the activity only time important (l=yes)
Did a substitution occur at 1st stage (l=yes)
Did a substitution occur at 2nd stage (l=yes)
Did a substitution occur at 3rd stage (l=yes)
Was there a time substitution (l=yes)
Was there a frequency substitution (l=yes)
Was there a locational substitution (l=yes)
Was there some other kind of substitution (l=yes)
Percent of activity done during day
Equipment replacement expenditures
Hours per week in activity
Frequency per week in activity
X-coordinate location of activity
Y-coordinate location of activity
Miles travelled
Direct costs
1-3
4
5-6
7
8
9
10
11
12
13
14
15
16
17-19
20-24
25-28
29-32
33-34
35-36
37-40
41-44
Format
F3.0
Fl.O
F2.0
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
F3.0
F5.0
F4.0
F4.0
F2.0
F2.0
F4.0
F4.0
195
-------
CARDS 4, 6, 8, 10, . . ., 68
Variable Name Descripti-on
QNM
QTYP
CRDNM
10HRB
10FQB
10LCXB
10LCYB
10MIB
10DCB
10HRC
10FQC
10LCXC
10LCYC
10MIC
10DCC
10HRD
10FQD
10LCXD
10LCYD
10MID
10DCD
Questionnaire number
Questionnaire type
Card number
Hours
Frequency
X-Location
Y-Location
Miles travelled
Direct costs
Hours
Frequency
X-Location
Y-Location
Miles travelled
Direct costs
Hours
Frequency
X-Location
Y-Location
Miles travelled
Direct costs
Coluran(s)
1-3
4
5-6
7-10
11-14
15-16
17-18
19-22
23-26
27-30
31-34
35-36
37-38
39-42
43-46
47-50
51-54
55-56
57-58
59-62
63-66
Format
F3.0
Fl.O
F2.0
F4.0
F4.0
F2.0
F2.0
F4.0
F4.0
F4.0
F4.0
F2.0
F2.0
F4.0
F4.0
F4.0
F4.0
F2.0
F2.0
F4.0
F4.0
196
-------
Card 69: Home Characteristics
Variable
Name
QNM
QTYP
CRDNM
LVAR
NRM
NBDRM
NBTRM
DEN
FAM
DIN
PCH
ATTIC
BASE
UTRM
OTRM
SCVW
STOR
REMD
DISH
DISP
CEAIR
TRASH
CEHT
POOL
FRPL
Description
Questionnaire number
Questionnaire type
Card number
Living area (sq. ft.)
Number of rooms
Number of bedrooms
Number of bathrooms
Den (l=yes)
Family room (l=yes)
Dining room (l=yes)
Enclosed porch (l=yes)
Attic (l=yes)
Basement (l=yes)
Utility room (l=yes)
Other room (l=yes)
Scenic view (l=yes)
Number of stories (include basement)
Remodeled (l=yes)
Dishwasher (l=yes)
Disposal (l=yes)
Central air conditioning (l=yes)
Trash compactor (l=yes)
Central heating (l=yes)
Swimming pool (l=yes)
Fireplace (l=yes)
Column(s)
1-3
4
5-6
7-11
12-13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
Format
F3.0
Fl.O
F2.0
F5.0
F2.0
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
197
-------
Card 69 (continued)
Variable
Name
AGERM
YRPC
LVHM
PCPR
MTPY
TDYVI
PTYTX
LVAPT
MTRT
INSPY
UPKP
PCTBST
STD
Description
Age of home (years since construction)
Year of purchase (last two digits)
Length of time (years) living in home
Purchase price of home
Monthly payments (rounded to nearest dollar)
Value of home in today's market
Property tax payments per year
Length of time (years) in apartment
Monthly rent
Insurance payments/year
Monthly upkeep around home
Percent of basement completed
Automobile standards (l=increase; 2=decrease;
Column(s)
34-35
36-37
38-39
40-45
46-48
49-54
55-58
59-60
61-64
65-68
69-72
73-75
Format
F2
F2
F2
F6
F3
F6
F4
F2
F4
F4
F4
F3
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
3=same)
76
Fl.O
198
-------
Card 70: Home Characteristics and Transportation
Variable
Name
QNM
QTYP
CRDNM
HC71
HC72
HC713
AVGXFD
AVGXCL
INCOME
PYFC
VLUOTL
NMVEH
LICDR
US TMPC
LWSTMPG
MLTV1.D
HRCMWK
HRCHSP
HRCMREC
CRPOOL
GASCST
MTCST
RTD
AUTOINS
ZVAC
VACX
Description
Questionnaire number
Questionnaire type
Card number
Why have you chosen tc live in chis area
(0=not ranked; other ranked scale of 1 to 5
with 1 the best) 6 if only checked
Average monthly expenditures for food
Average monthly expenditures for clothing
Annual household income (midpoint of groups)
How much would you pay for house if in area
like C
How much of value of your home is due to no
air pollution
Number of vehicles in family
Licensed drivers in family
Highest average miles per gallon
Lowest miles per gallon
Miles travelled per week
Hours per week spent commuting for work or
school
Hours/week spent commuting for shopping
Hours/week spent commuting for recreation
Are you in a car pool (l=yes)
Gasoline costs/month
Maintenance costs/month
Public transportation fares/month
Auto insurance/month
Vacation within last year (l=yes)
Vacation expenditures
Column(s)
1-3
4
5-6
7
19
20-23
24-27
28-33
34-39
40-44
45
46
47-48
49-50
51-54
55-56
57-58
59-60
61
62-64
65-67
68-70
71-73
74
75-80
Format
F3.0
Fl.O
F2.0
Fl.O
Fl.O
F4.0
F4.0
F6.0
F6.0
F5.0
Fl.O
Fl.O
F2.0
F2.0
F4.0
F2.0
F2.0
F2.0
Fl.O
F3.0
F3.0
F3.0
F3.0
Fl.O
F6.0
199
-------
Card 71: Medical and Attitudes
Variable
Name
QNM
QTYP
CRDNM
MD1A
MD1H
MD2I
MD20
AGGAP
DSAQ
PHYDIS
LFPL1
LFPL2
DRG1
EASWK1
NGTRDY
SMOKE
PACKS
MEDCTN
HEDX
MEDXAP
DR
MEDINS
Description
Questionnaire number
Questionnaire type
Card number
Medical, question 1, part a
Medical, question 1, part h
Medical, question 2, part I
Medical, question 2, part 0
Conditions aggravated by air pollution (l=yes)
Diseases which could be made worse by air
pollution (l=yes)
Physical disabilities (l=yes)
Life more pleasant (not at all=00; to some
degree=01; greatly=10)
Spend more on drug items (not at all=00;
to some degree=01; greatly=10)
Make it easier to do your work (not at all=00;
to some degree=01; greatly=10)
Prefer night or day (l=day; 2=night ; 3=no
difference)
Do you smoke (l=yes)
How many packs
Do you take medication regularly (l=yes)
Medication expenditures/month
Medical expenditures associated with air
pollution
Yearly doctor's fees
Yearly payments on medical and life insurance
Column(s)
1-3
4
5-6
7
14
15
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35-37
38-40
41-44
45-47
Format
F3.0
Fl.O
F2.0
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
Fl.O
F3.0
F3.0
F4.0
F3.0
200
-------
CARD 71 (CONTINUED)
Variable
Name
DEFX
ATT 11
ATT 12
ATT 13
ATT 21
ATT 2 2
ATT23
ATT31
ATT32
ATT 41
ATT A 2
ATT A 3
ATT A 4
ATT45
ATT46
ATT47
ATT48
ATT49
ATT410
ATT411
ATT412
Column (s) Format
Have you ever purchased any item to reduce your exposure
to air pollution? (Such as filter) (1 = yes)
Since living in Los Angeles:
I have not been bothered by air pollution. (1 = yes)
I have been somewhat bothered by air pollution. (1 = yes)
I have been bothered by air pollution. (1 = ves)
Do you believe that air pollution in Los Angeles: (check one)
Has become worse since you have lived here.
Has stayed about the same since you have lived here.
Has gotten better since you have lived here.
What do you think should be done about air pollution?
(Check one)
Ignored.
Reduced.
Rank the following problems in terms of importance
(most to least) as the major issues facing the community.
(Choose top five.)
Juvenile delinquency.
Communicable disease.
Unemployment.
Air pollution.
Car accidents.
Crime.
Nuclear Energy.
Alcoholism
Water pollution
Energy
Congestion
Other
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
Fl
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
.0
201
-------
CARD 71 (CONTINUED)
Variable
Name
Column(s) Format
Do you believe air pollution in the Los Angeles area:
ATT51 Cannot be reduced below the present level. 69 Fl.0
ATT52 Can be reduced below the present•level. 70 Fl.0
ATT53 Can be almost completely eliminated. 71 Fl.O
What do you think the words air pollution mean to most
pepple in the Los Angeles area? Do they mean:
ATT61 Frequent bad smells in the air. (1 = yes; 0 = no) 72 Fl.O
ATT62 Too much dirt and dust in the air. (1 = yes; 0 = no) 73 Fl.O
ATT63 Frequent haze or fog in the air. (1 = yes; 0 = no) 74 Fl.O
ATT64 Frequent irritation of the eyes. (1 = yes; 0 = no) 75 Fl.O
ATT65 Frequent nose or throat irritations. (1 = yes; 0 = no) 76 Fl.O
ATT66 Other. 77 Fl.O
ATT7 Have you read or seen anything in the newspaper recently
about air pollution? (1 = yes; 0 = no) 78 Fl.O
ATT8 When you read the newspaper, do you generally choose to
read articles on air pollution? (1 = yes; 0 = no) 79 Fl.O
ATT9 Do you consult the daily air pollution index before
engaging in any activities? (1 = yes; 0 = no) 80 Fl.O
202
-------
APPENDIX F
This appendix details the actual streets in the paired areas from which
the respondent sample was drawn.
203
-------
PARK
u
sr
I
I
w
COAV
\JFM.
I
JS/Y
1,1
¥
sr
\"
204
-------
CUU/ER
^
^
•+ K^v.w/wvl
205
-------
ELL
206
-------
N
207
-------
Huf/T[NfiTDM
268-
-------
IRVINE
STJ\£JTH1STL£
209-
-------
LA CANADA
210
-------
UWCOlN/
TE
211
-------
AVEL
212
-------
RCIFSC PAM SAFES
KJ
213
-------
214
-------
H
REDOMDO BEACH
-------
REFERENCES
Ajzen, I. and Fishbein, M. , "Attitude-Behavior Relations: A Theoretical
Perspective and Review of Empirical Research," Psychological Bulletin
84_ (1977) 888-918.
Anderson, R.J., Jr. and Crocker, T.D. (1971), "Air Pollution and Residential
Property Values," Urban Studies 8, 1971-180.
Babb, E.M., and Scherr, B.A. (1975), "Pricing Public Goods: An Experiment
with Two Proposed Pricing Systems," Public Choice 23, 35-48.
Bayes, T., "An Essay Towards Solving a Problem in the Doctrine of Chances,"
in The Royal Society, Philosophical Transactions, Giving Some Account
of the Present Undertakings, Studies, and Labours of the Ingenious in
Many Considerable Parts of the World for the Year 1763, London: The
Royal Society (1765).
Blank, F., et. al. , "Valuation of Aesthetic Preferences: A Case Study of
the Economic Value of Visibility," Research Report to the Electric
Power Research Institute, Resource and Environmental Economics Labor-
atory, University of Wyoming (1977).
Bohm, P. (1971), "An Approach to the Problem of Estimating Demand for Public
Goods," Swedish Journal of Economics 73.
Bohm, P. (1972), "Estimating Demand for Public Goods: An Experiment,"
European Economic Review 3; 111-130.
Bradford, D.F., "Benefit-Cost Analysis and Demand Curves for Public Goods,"
Kyklos, Vol. 23 (1970).
Brookshire, D., et. al. , "The Valuation of Aesthetic Preferences," Journal
of Environmental Economics and Management, Vol. 3 (1976).
Brookshire, D.S. and Crocker, T.D., "The Use of Survey Instruments, in
Economic Valuations of Environmental Goods," Paper #2, Resource and
Environmental Economics Laboratory, University of Wyoming (1978).
Brookshire, D.S., and Eubanks, L.S., "Contingent Valuation and Revealing
the Actual Demand for Public Environmental Commodities," Paper Pre-
sented at Public Choice Society, New Orleans, La., 1978. Forthcoming
in Proceedings.
Brookshire, D.S., and Randall, A.J., "Public Policy Alternatives, Public
Goods and Contingent Valuation Mechanisms," Paper presented at the
Western Economic Association Meeting, Honolulu, Hawaii, 1978.
Brookshire, D.S., and Randall, A., et. al., "Economic Valuation of Wildlife,"
Final report for U.S. Fish and Wildlife Service, Phase I, Resource
and Environmental Economics Laboratory, University of Wyoming, (Nov-
ember 8, 1977).
216
-------
Clark, E.H.. (1971), "Multipart Pricing of Public Goods.," Public Choice 11,
17-33.
Clawson, M. and Knetsch, J.L. (19.66), Economics: of Outdoor Recreation, Johns
Hopkins Press: Baltimore.
Crocker, T., "National and Regional Models- for the Estimation of Recreational
Benefits Accruing from PL 92-500,"' Research Paper No. 126, Department of
Economics, University of Wyoming (1975).
Davis, R. (1963), "Recreation Planning as- an Economic Problem," Natural
Resources Journal 3.
Davis, R.K. and Knetsch, J.L. (1966), ""Comparisons of Methods for Recreation
Evaluation" in Water•Research, A. V. Kneese and S.C. Smith (eds.), Johns
Hopkins Press: Baltimore.
Deyak, T.A. and V.K. Smith, "Residential -Property Values and Air Pollution:
Some New Evidence," Quarterly Review of Economics and Business 14(4),
(1978), 93-100.
Diewert, W.E., "Applications of Duality Theory," in M.D. Intriligator and D.A.
Dendrick (eds.)» Frontiers-of Quantitative Economics, Vol. II, Amsterdam:
North Holland Publishing Co. (1974), 106-160.
Freeman, A.M., Ill (1974), "On Estimating Air Pollution Control Benefits from
Land Value Studies," Journal of Environmental Economics and Management 1,
74-83.
Fromm, G., "Comment," in S.B. Chase, Jr. (ed.) Problems in Public Expenditure
Analysis, Blockings Institution, Washington, D.C., 166-176 (1968).
Goldberger, A.S., A Review of Consumer Demand Theory, Social Systems Research
Institute, University of Wisconsin-Madison, Systems Formulation, Method-
ology, and Policy Workshop Paper 6703 (Oct. 1967).
Hall, R., "The Specification of Technology with Several Kinds of Output,"
Journal of Political Economy 81 (1973), 387-398.
Hammack, J. and Brown, G.M., Waterfowl and Wetlands: Toward Bioeconomic
Analysis, Baltimore: Johns Hopkins University Press (1974).
Harrison, D., Jr. and D.L. Rubinfeld, "Hedonic Housing Prices and the Demand
for Clean Air," Journal of Environmental Economics and Management 5
(1978), 81-102.
Hicks, J.R., "A Reconsideration of the Theory of Value," Part 1, Economica 1
(Feb.. 1934), 52-76.
Hori, H., "Revealed Preferences, for Public Goods," American Economic Review,-
Vol. 65 (1975).
217
-------
Horst, R., Jr. (L978) , Es^jmotion of the ____
Application of the Expenditure Function, unpliblTslTefl Ph.D. dissertation,
University of Wyoming.
Hurwicz, L., "On the Problem of Integrability of Demand Functions," in J.
Chipman, et. al. (eds.), Preferences, Utility and Demand, New York:
Harcourt Brace lovanovich, Inc. (1971)" 174-214.
Johansen, L. (1977), "The Theory of Public Goods: Misplaced Emphasis?"
Journal of Public Economics 7, 147-152.
Knetsch, J.L. (1963), "Outdoor Recreation Demands and Benefits," Land
Economics. ~~
Kurz, M. (1974), "Experimental Approach to the Determination of the Demand
for Public Goods," Journal of Public Economics 3, 329-348.
Lancaster, K. J. , "A New Approach to Consumer Theory," Journal of Political
Economy 74 (1966), 132-157. " ~ '
Lave and Nagin, "Automobile Related Air Pollutants and Health, "REF
Working Paper No. 52-73-74.
Lucas, R.C., "Wilderness Perception and Use: The Example of Boundry Waters
Canoe Area," Natural Resources Journal 3 (3), 1964, 394.
Lucas, R.E.B., "Hedonic Price Functions," Economic Inquiry 13(2), 1975,
157-178.
MMler, K. (1974), Environmental Economics: A Theoretical Inquiry, Johns
Hopkins Press: Baltimore.
Mishan, E.J. (1971), Cost-Benefit Analysis: An Introduction, Praeger
Publishers: New York.
Muellbauer, J., "Household Production Theory, Quality, and the Hedonic
Technique," American Economic Review, Vol. 64 (1974).
Pearse, P.H., "A New Approach to the Evaluations of Non-Priced
Recreational Resources," Land Economics, 44 (February, 1968).
Pollak, R., and Wachter, M., "The Relevence of the Household Production
Functions and Its Implications for the Allocation of Time," Journal of
Political Economy 83 (1975), 255-277.
Raiffa, H., Decision Analysis, Reading, Mass.: Addison-Wesley (1970).
Randall, A., et. al., "Bidding Games for Valuation of Aesthetic Environmental
Improvements," Journal of Environmental Economics and Management, Vol 1
(1974).
218
-------
Randall, A., et. al_._, Estimating Environmental Damages from Surface Mining of
Coal in Appalachia: A Case Study, U.S. Environmental Protection Agency,
EPA-600/2-78-003, Jan. 1978.
Randall, A., "Evaluating Non-Market Goods, and Services.: Some Conceptual
Considerations," presented to the Symposium on Evaluation o£ Non-
Market Goods at the Annual Meetings- of the American Agricultural
Economics Association, San Diego, California (July 31-August 3, 1977).
Randall, A.: Ives, B.: and Eastman, E., "Bidding Games for Valuation of
Aesthetic Environmental Improvements-," Journal of Environmental Econo-
mics and Management 1 (1974a) 132-149.
, "Benefits of Abating Aesthetic Environmental Damage from
the Four Corners Power Plant, Fruitland, New Mexico," Bulletin 618, New
Mexico Agricultural Experiment Station, Las Cruces (1974b).
Randall, A. and Stoll, J., "Consumer's Surplus in Commodity Space," Working
paper, Department of Agricultural Economics, University of Kentucky
(1978).
Ridker, R.B. and Henning, JcA. (1967), "The Determinants of Residential
Property Values with Special Reference to Air Pollution," Review of
Economics and Statistics 49, 246-257.
Roberts, E.M., et. al., "Visibility Measurement in the Painted Desert
through Photographic Photometry," Domes and Moore Engineering Bulletin
No. 47 (1974).
Rosen, S., "Hedonic Prices and Implicit Markets: Product Differentiation in
Pure Competition," Journal of Political Economy, Vol. 82 (1974).
Samuelson, P.A., "The Problem of Integrability iri Utility Theory," Economica
17 (1950), 355-385.
Samuelson, P.A., "The Pure Theory of Public Expenditures," Review of
Economics and Statistics, 36 (November, 1954).
Samuelson, P., "Diagrammatic Exposition of a Theory of Public Expenditure,"
Review of Economics and Statistics 37 (Nov. 1955), 350-356.
Schulze, W.D. and d'Arge, R.C., "On the Valuation of Recreation Damages,"
prepared for the Joint Session between the American Economic Association
and the Association of Environmental and Resource Economists, New York
(December 28, 1977) „
Steele W J "The Effect of Air Pollution on Value of Single Family Owner-
Occupied Redisential Property in Charleston, South Carolina, Masters
Thesis, Clemson University (1972)
219
-------
Thayer, M. and Schulze, W., "Valuing Environmental Quality: Th.e
Contingent Substitution and Expenditure Approach," unpublished
manuscript. (.1977).
U.W: Bureau of the Census, Census of Population and Housing: 1970,
Census Tract Final Report.
Wieand, K.F., "Air Pollution and Property Values; A Study Qf the St» Louis
Area," Journal of Regional Science, Vol. 13, April 1973, 91-95.
220
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/6-79-001b
4.TITLEANDSUBTITLE Methods Development for Assessing Air"
Pollution Control Benefits: Volume II, Alternative
Benefit Measures of Air Pollution Control in the
South Coast Air Basin of Southern California
3. RECIPIENT'S ACCESSION NO.
5. REPORT DATE
February 19>9
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
David S. Brookshire, Ralph C. d'Arge, William D.
Schulze, and Mark D. Thayer
8. PERFORMING ORGANIZATION REPORT NO
10. PROGRAM CLEMENT NO.
1HA616 and 630
. PERFORMING ORGANIZATION NAME AND ADDRESS
Universi'ty of^y
Laramie, Wyoming
82071
11. CONTRACT/GRANT NO.
R805059-01
12. SPONSORING AGENCY NAME AND ADDRESS
Office of Health and Ecological Effects
Office of Research and Development
U.S. Environmental Protection Agency
Washington, DC 20460
13. TYPE OF REPORT AND PERIOD COVERED
Interim Final, 10/76-10/78
14. SPONSORING AGENCY CODE
EPA-600/18
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This volume of a five volume study on the economic benefits of air pollution
control includes the empirical results obtained from two experiments to measure
the health and aesthetic benefits of air pollution control in the South Coast Air
Basin of southern California. Each experiment involved the same six neighborhood
pairs, where the pairings were made on the basis of similarities in housing charac-
teristics, socio-economic factors, distances to beaches and services, average
temperatures, and subjective indicators of housing quality. The elements of
each pair differed substantially only in terms of air quality. Data on actual
market transactions, as registered in single-family residential property trans-
actions, and on stated preferences for air quality, as revealed
surveys, were collected.
Given various assumptions on income, location, aggregation
housing characteristics, and knowledge of the health effects of air pollution,
both the survey and the property value experiments yielded estimates of willingness-
to-pay in early 1978 dollars for an improvement from "poor" to "fair" air quality
of from $20 to $150 per month per household. The results, therefore, indicate that
air quality deterioration in the Los Angeles area has had substantial negative
effects on housing prices and that these effects are comparable in magnitude to
what people say they are willing to pav for improved air Quality.
in neighborhood
by areas, specific
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Economic analysis
Air pollution
Environmental survays
Real property
Economic benefits of
pollution control
13B
18. DISTRIBUTION STATEMENT
Release unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
230
20. SECURITY CLASS {Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
o U S GOVtRNMWI PRfflTINC OFflCC: 1979 -281-147/21
-------
EPA Library
•f States
nomental Protection
Agei
$300
T 20539
Washington DC 20460
c
-------