Forgot your password?
Document Actions
Contingent Valuation of an urban salt marsh restoration
Contingent Valuation of an Urban Salt Marsh Restoration
Matthew K. Udziela
Yale School of Forestry and Environmental Studies
Lynne L. Bennett
Yale School of Forestry and Environmental Studies
ABSTRACT
This paper discusses the validity of the contingent valuation method (CVM) and presents the results of a CVM survey
we conducted to estimate the nonmarket value of restoring a salt marsh in New Haven, Connecticut. The literature
suggests that CVM is valid, but results are often biased by poor survey design. Using a survey that was extensively
pretested, we addressed potentially problematic issues such as payment vehicle bias, information bias, and hypothetical
bias. We also combined techniques from conjoint analysis. Fifty-one households near the West River in New Haven
were surveyed on Saturdays in March and April 1996. The households represented a wide range of socio-economic
characteristics. Survey respondents were provided with background information including the current condition of the river
and possible costs and benefits of restoration. Respondents were asked if they supported or opposed the restoration.
Their maximum willingness to pay either for the restoration, or to keep the river in its current condition was solicited.
The contingent valuation survey performed well and was effective in educating participants about the current
condition of the West River, the restoration measures proposed, and the possible impacts of the restoration; it was also
effective in obtaining willingness to pay responses from people interviewed. Over three-quarters of people interviewed
supported the proposed opening of the tidegates. Moreover, this support extended to a willingness to pay considerable
amounts of money for the restoration. People clearly have high values for nonmarket and even nonuse environmental
resources. Without a means to estimate their economic value, these significant amenities will be ignored and thus not
protected. Contingent valuation can be an effective method of obtaining price estimates for non-market values that
would be excluded from consideration in traditional economic decision-making because they have no associated price.
Consequently, economic valuation has a crucial role to play in preserving and restoring priceless environmental resources.
We conclude that a carefully designed contingent valuation survey is a promising method for bringing nonmarket
economic values into the cost-benefit analysis of restoring the West River salt marsh.
The West River and its flood plains occupy approximately 35 square
miles in New Haven and West Haven, Connecticut. Historically, 800
acres of this area were salt marsh (Casagrande, pp. 13-40, this volume).
It was a highly productive estuarine ecosystem with an associated bio-
logical community.
In 1919 tidal gates were installed where Route 1 crosses the river
in New Haven for the purpose of marsh reclamation. The gates
stopped Long Island Sound’s saltwater from traveling upstream,
thus reducing upstream salinity and tidal inundation. As a result,
Phragmites australis have aggressively colonized the area north of the
tide gates. Such large, man-made stands of Phragmites provide poor
wildlife habitat.
There is currently significant interest and some controversy with
regard to restoring about 70 acres of the historic salt marsh located
just upstream of the gates within West River Memorial Park. Both
the interest and the controversy revolve around the effects of opening
some of the 12 tide gates to increase tidal flushing. Other restoration
projects in Connecticut have shown that restored tidal flushing
reduces Phragmites and allows a salt-marsh community to return
(Steinke 1986, Rozsa and Orson 1993). However, opening the tidal
gates could also have negative impacts such as flooding.
For a policy decision regarding the restoration to be balanced
and objective, it is helpful to aggregate economic benefits and costs.
Economic valuation is complicated, however, by the fact that resto-
ration benefits include salt-marsh resources that do not have a mar-
ket price (e.g., scenic views and wildlife). In other words, value is not
For a policy decision regarding the
determined or set by the market. Although scenic views and wildlife
restoration to be balanced and objective,
have a significant value to the public, they risk being excluded from
it is helpful to aggregate economic benefits
economic analysis because they appear to have no economic value. and costs. Although scenic views and
The historic eradication of salt marshes provides evidence that the wildlife have a significant value to the
exclusion of nonmarket values can result in environmentally destruc- public, they risk being excluded from
tive policies (Casagrande, pp. 13-40, this volume). For this reason, we economic analysis because they appear
have focused on the nonmarket benefits and costs of restoration. to have no economic value.
The most widely recognized method for the valuation of
1
See Portney 1994, Cummings et al. 1986, and
nonmarket resources is the contingent valuation method (CVM).
Mitchell and Carson 1989 among others for
CVM involves presenting survey respondents with a hypothetical details of methodology.
market for the resource and eliciting a willingness to pay for that 2
See for example Samples et al. 1986, Boyle
resource.1 Using CVM to value natural resources is not new. CVM and Bishop 1987, and Stevens et al. 1991.
has been used in several studies of endangered species.2 Other stud-
ies have focused more generally on wildlife preservation (Stevens et
al. 1994, Desvousges et al. 1993).
CVM has been the subject of debate among economists and
others for the last decade. Even though a panel of experts convened
by the National Oceanic and Atmospheric Administration (NOAA)
in 1993 concluded that contingent valuation could produce reliable
estimates of some types of values, critics have stated that CVM has
failed to produce reliable and accurate measures of value (Diamond
and Hausman 1994, McFadden and Leonard 1993, Diamond et al.
1993). Others emphasize that most (or even all) of CVM’s perceived
failures have been due to surveys that were inadequately planned or
poorly structured, and that CVM works when properly executed
(Portney 1994, Hanemann 1994). This study helps to alleviate some
concern. We argue that a careful survey design can eliminate many
of the biases that have caused the controversy.
We used the CVM method to assess the nonmarket economic
value of restoring the salt marsh in West River Memorial Park.
Our goal was to produce a reliable estimate by adhering to survey
methods that recent studies suggest can reduce bias.
SURVEY DESIGN
The survey instrument was carefully designed and pre-tested
using focus groups and practice interviews. The survey is presented
in the Appendix. The survey was conducted as personal interviews
and consisted of four key components.
A DESCRIPTION OF THE CONTEXT AND THE COSTS AND
BENEFITS TO BE VALUED
An important factor in the effectiveness of a contingent valuation
survey is that the survey respondent must have a clear and accurate
idea of the specific good that is being valued. In the case of natural
resources, this is often a difficult concept to convey, because there
are countless ways of viewing the resource. As a result, respondents
could be considering their willingness to pay for a good that is dif-
ferent from the one being asked about in the survey. Careful survey
design can help to avoid this information bias.
For example, the West River provides foraging habitat for the least
tern, a threatened species in Connecticut that dives for fish (Lewis
and Casagrande, this volume). The degraded condition of the marsh, Least tern
however, reduces the habitat value for the tern. Least terns are common
visitors to salt marshes in the northeast (Post 1970), and have been
shown to benefit from salt-marsh restoration (Bontje 1987). Open-
3
The CV study of migratory waterfowl by
ing the tide gates and restoring the salt marsh may benefit the least
Desvousges, et al. (l993) is commonly cited
tern by increasing fish abundance and improving water quality. as an example of the severity of the
Nevertheless, a survey question that asked for the value of the least embedding effect, because respondents
failed to distinguish between 2,000, 20,000
tern only could have been ambiguous. Could the question be referring and 200,000 birds. However, Hanemann
to the bird itself, reduced chance of extinction for the bird, probabil- (1994) has noted that when the data were
trimmed to remove the extreme 10% of
ity of viewing a least tern during a visit to the West River, or the
the values, the willingness to pay for each
habitat it lives in? of the three quantities of birds was
different.
Related to this is an issue known as “part-whole bias” or the
“embedding effect,” in which respondents produce equal values of
willingness to pay for two goods, one of which is a subcomponent of
the other (Brown et al., 1995). A good of interest, known as α, is
valued by one set of subjects. Then, a larger good (Σ), which includes
α, is valued by another set of subjects. Since studies of this type have
obtained results that α ≈ Σ,3 the conclusion is that either Σ is not
valued at all (which is unlikely), or that respondents value α as a
component of Σ. In our case, people may express a desire to preserve
the least tern because they value the habitat it lives in. In addition,
respondents may be valuing other benefits of the restoration, such
as enhanced aesthetics. Again, careful survey design can reduce the
magnitude of this effect or eliminate it entirely.
This research addressed the embedding effect by phrasing the
WTP question so that the good offered was a restored salt marsh
that contained several benefits, including a specific number of
acres (70) of habitat for the least tern, habitat for other wildlife,
“naturalness” of salt marshes, and reduced fire hazard. Likewise,
costs such as increased risk of flooding were included. Thus, respon-
dents were considering an entire bundle of costs and benefits, all of
which derive from the salt-marsh restoration. Pretesting of the
survey had indicated that it would be beneficial to present respon-
dents with all of the advantages and disadvantages of both condi-
A panel of experts convened by the
tions. They could then decide which “bundle” of benefits and
National Oceanic and Atmospheric
drawbacks they preferred. This technique is called conjoint analysis
Administration (NOAA) in 1993
and has been used extensively in marketing and transportation concluded that contingent valuation
economics and more recently in natural resource valuation (Roe et could produce reliable estimates of
al. 1996). Thus our survey not only incorporated CVM questions, values.
but included conjoint analysis questions as well.
If information is comprehensive and properly presented, bias
can be reduced. Therefore, we used visual aids extensively including
photographs and maps to present a detailed description of current
conditions and the effects of restoring tidal flushing.
Interviews began with questions about environmental attitudes
and behavior. Respondents were asked how often they used the
West River, where along the river they went, and what activities they
did there, or why did they not use the river. Respondents were also
asked about their perceptions of the West River’s water quality,
recreation potential, and habitat value. This allowed people to express
their opinions and become comfortable with the survey.
Respondents were then asked whether they preferred the West
River in its current state or as a restored salt marsh. This question
was based on two photographs, one of each condition. Therefore,
initial preference was based largely on aesthetics. A series of questions
then gave successive information about the two river types. After each
new fact was presented, respondents were asked which river they
preferred. Finally, respondents were handed a list of all the advantages
and disadvantages of each river type and asked to state a final prefer-
ence based on all the information.
Interviewers next described the restoration proposal. The restored
salt marsh would have both benefits (e.g., habitat for the least tern)
and costs (e.g., increased flooding). Respondents were then asked if
they supported or opposed the restoration.
THE WILLINGNESS TO PAY QUESTION
Economic value can be elicited with a CVM survey either using a
question of willingness to pay (WTP) for a resource or willingness to
accept (WTA) the loss of a resource. WTP is a measure of the maxi-
mum amount that a person would pay for a particular good or
service, such as environmental quality or species existence, or for an
incremental change in the amount of an environmental service
provided. WTA refers to the minimum level of compensation that
person would accept for losing those resources. Traditional economic
theory suggests that the two should be equivalent. But studies have
often found WTA to be larger than WTP (in some cases orders
of magnitude larger). The NOAA panel (1993), among others,
recommend using WTP since it elicits more conservative responses
of value.
WTP is a measure of the maximum
Furthermore, a WTP response implies the respondent has no
amount that a person would pay for
property right to the resource being valued. Alternatively, willing- a particular good or service, such as
ness to accept compensation for the loss of a resource would suggest environmental quality or species
that the respondent has a property right to the resource. We did not existence, or for an incremental
change in the amount of an
feel WTA applied to the West River situation and we elicited economic
environmental service provided.
value using WTP.
The WTP question can be open-ended or presented as a dichoto-
mous choice. In an open-ended question, respondents are asked for
the maximum amount they would be willing to pay to obtain a
particular environmental amenity. With dichotomous choice,
respondents are presented with a price and asked to indicate
whether or not they would be willing to pay the stated price. The
NOAA panel recommended dichotomous choice in CVM surveys,
because respondents are asked to make a familiar decision similar to
voting or deciding whether or not to buy something. An open-ended
question, the panel argued, places respondents in the unfamiliar
situation of determining a value. However, recent comparative
studies have revealed dichotomous choice WTP bids that are much
higher than those from open-ended questions (see for example,
McFadden and Leonard 1993). We used the open-ended format in
order to obtain a more conservative estimate of New Haven’s resi-
dents’ values of salt-marsh restoration. We also chose open-ended
questions in order to eliminate starting point bias, which is typically
associated with dichotomous choice methods.
If respondents in our survey expressed support for the restoration,
they were asked the maximum amount they would pay to ensure
that it was implemented. If respondents opposed the restoration, the
next question asked how much they would be willing to pay to
ensure that the restoration was not implemented. Thus, the WTP
question measured the economic value both for and against the
restoration. If a respondent gave a range of values, the lowest value
was used for analysis.
WTP can be biased if respondents do not think in the framework
of economic choices that they make with their own money. There
may be an overestimation of price if they ignore what they will
sacrifice by paying for the environmental amenity. When interview-
ing people, we kept the realism level high by making the scenario
very specific: payment would only occur once; it would occur within
the next year; and subjects were required to make their decision on
the basis of their present income, not their expected future income.
Respondents were also reminded that any amount they were willing
to pay would reduce their ability to purchase other goods and services.
In order not to bias WTP estimates, protest bids must be identi-
fied and removed. A protest bid is a stated response of $0 to a will-
ingness to pay question even though actual value might be greater
than zero. Protest bids occur when respondents are protesting some-
thing other than the good in question. Since these responses are not
valid “zeros,” they must be eliminated from the sample. A common
example of a protest bid is the objection to the type of payment
mechanism, such as a tax. This type of protest bid creates what is
called vehicle bias. To test for vehicle bias, half of our surveys were
conducted with a one-time tax as the payment mechanism. The other
half included a one-time private donation.
Protest bids can also result from a moral objection to placing a
price on the environment. At the end of interviews, we asked re-
spondents to give the reasons for their WTP value. This question
was included to identify protest bids, though it also provided other
useful information, such as confirmation that respondents under-
stood the restoration scenario.
SOCIOECONOMIC QUESTIONS
Socioeconomic data were solicited confidentially. Respondents
were asked to fill out a form about their age, education, income, and
race. They folded the form and sealed it in an envelope, which was
attached to the survey by the interviewer. This information was used
to determine the range of socioeconomic characteristics within our
sample, and to test for influence of income on WTP.
FOLLOW-UP QUESTIONS
We asked follow-up questions to ensure that respondents under-
stood the scenario and the payment format and believed the infor-
mation presented. Respondents were also asked to give reasons for
answers of “$0” WTP, so that we could identify protest bids.
SAMPLING DESIGN
Three neighborhoods were selected to represent the greatest
range of socio-economic characteristics adjacent to the West River.
The Hill is located in southern New Haven and borders the lowest
stretch of the West River downstream from the tide gates. Households
in the Hill are mostly Hispanic or African American and represented
the lowest income. Edgewood is located to the north of the Hill, and is
upstream from the tide gates. Edgewood is racially mixed, and incomes
are higher than the Hill. Westville is in the north of New Haven;
residents are mostly White and have high incomes.
Land use adjacent to the river differs for the three neighborhoods.
Land use along the river in the Hill is mostly commercial, landfill, or
junkyards. Adjacent to the Edgewood neighborhood, the West River
flows through Edgewood Park. Westville is bisected by the river, and
some homes are located close to the river. Westville suffers major
flooding about every 30 years. The most recent flood in 1982 caused
substantial loss of property.
Interviews were conducted with teams of two. This was initially
done for safety, but it was also found that one interviewer could ask
questions while the other handled the numerous visual aids. As soon
as the interview was completed, interviewers filled out a short evalu-
ation form to record time, date, address, payment mechanism, and
their impression of the interview.
Interviewers went door-to-door to survey people in their homes
during four Saturday afternoons in March and April. Each interview
team was assigned a block and called at every house. People who
were home and willing to participate were interviewed. In order to
reduce voluntary response bias, subjects were offered $5 in cash to
be interviewed. A total of 51 households were surveyed.
Interviewers who were native speakers of Spanish conducted
impromptu translations of the survey for Spanish-speaking respon-
dents. This enabled a wider cross-section of the population to be
interviewed.
Table 1. Number of respondents in income groups by neighborhood.
Income Hill Edgewood Westville Total % Total
$10,000-$19,999 4 2 0 6 14
$20,000-$29,999 4 3 1 8 19
$30,000-$39,999 2 2 1 5 12
$40,000-$49,999 1 2 2 5 12
$50,000-$59,999 0 2 0 2 5
> $60,000 0 6 7 13 30
not indicated 2 0 2 4 9
RESULTS
Of the 51 surveys, 19 were conducted in the Hill, 19 in Edgewood,
and 13 in Westville. Household income of respondents was diverse,
with an increasing gradient from the Hill (lowest), to Edgewood and
Westville (highest, Table 1). Respondents were also racially diverse –
53% were Hispanic or African American (Table 2).
We removed protest bids in order to compute willingness to pay
(WTP). A protest bid was considered a bid of $0 that was due to a
protest against the payment mechanism (tax or donation) or moral
objection to placing a dollar value on an environmental amenity.
Protest bids were identified using follow-up questions. Nine protests
to the tax mechanism were identified and removed. We did not
detect any moral protests.
Table 2. Race of respondents.
There were three respondents who were unable to determine
their WTP. These were not protest bids, but were removed in order
Number Percent
to calculate mean WTP. Respondents who wanted to pay, but were
Hispanic 14 25
not able to pay anything were considered to have given valid bids of
Black 13 28
$0, and were included in mean WTP.
White 22 43
Of the seven responses that opposed opening the tide gates
Asian 1 2
(Table 3), five were protest bids. This left two responses: a $20 tax
not indicated 1 2
bid and a $50 donation bid. The two bids against opening the tide
gates were included in mean WTP for salt-marsh restoration as
negative values. Of the 39 responses that favored opening the tide
gates (Table 3), 4 of the tax responses were protest bids. There were
no protest bids by respondents who favored opening the gates who
took the donation survey. Three others were unable to determine
WTP. Thus, there were 32 responses with valid WTP bids for open-
ing the tide gates; 17 by donation and 15 by tax.
Table 3. Responses to the proposal to open tide gates.
For the TAX survey, with 26 people surveyed:
3 people were against opening the tidegates (11%)
22 people were for opening the tidegates (85%)
1 person had no opinion (4%)
For the DONATION survey, with 25 people surveyed:
4 people were against opening the tidegates (16%)
17 people were for opening the tidegates (68%)
4 people had no opinion (16%)
For the TOTAL:
7 people were against opening the tidegates (14%)
39 people were for opening the tidegates (76%)
5 people had no opinion (10%)
Table 4. The contingency table for the χ2 test for the relationship between income and WTP.
WTP The expected values are:
Income ≤ Med > Med Total WTP
Income ≤ Med > Med
Low 14 5 19
High 5 8 13 Low 11.28 7.72
Total 19 13 32 High 7.72 5.28
There were many $100 bids (and a few $200 bids) and many bids
of $25 or less. Only one value occurred between $25 and $100. As a
result, there is a wide difference between mean WTP ($61.41) and
the median value ($25.00), and the standard deviation is large
($100.22). We used chi-square (χ2) tests of two-way contingency
tables to determine if WTP was influenced by income or payment
mechanism (tax vs. donation). A χ2 test was used because the WTP
values had a discrete probability distribution rather than continuous
(people tended to bid $25 or $100, but not values in between). Tests
were conducted at the 5% significance level.
WTP bids were divided into two categories: bids less than the
median, and bids greater than the median. Respondents were
grouped as those having household income below the median for
the survey, and those above (Table 4). The null hypothesis (Ho) was
that income and WTP are independent. There was one degree of
freedom because we used two-by-two contingency tables. Values for
χ2 were obtained from a reference table. WTP was dependent on
income (χ2 = 3.97).
Respondents’ bids were also grouped by payment mechanism
(Table 5). WTP bids were grouped as above. The null hypothesis
(Ho) was that payment mechanism and WTP were independent.
Final WTP values appeared to be independent of whether the
purchase mechanism was by tax or donation (χ2 = 0.43).
Table 5. The contingency table for the χ2 test for the relationship between payment mechanism and WTP.
The expected values are:
Payment WTP
mechanism ≤ Med > Med Payment WTP
Total
≤ Med > Med
mechanism
Tax 8 7 15
Tax 8.91 6.09
Donation 11 6 17
Donation 10.09 6.91
Total 19 13 32
DISCUSSION
Given that our sample size was relatively small, we view this
as a “pilot project” for what will ultimately be a survey of a larger
number of households. Testing our survey design for a small
sample allows us to determine whether or not to proceed with a
larger group.
The survey’s overall performance was satisfactory. People
generally understood the content, purpose, and logic, and
respondents tended to make WTP bids in terms of real economic
decisions. One respondent mentioned goods he would sacrifice
in his determination of WTP by deciding to “give up a couple of
twelve-packs.” Other comments included: “I know my budget for
April, and this is the amount I have free.” “I’m bringing up three
children; this is all I can spare.” “I can’t pay anything because I’m
not working right now.”
Respondents often anticipated questions. For example, when
asked, “Do you think the restoration project is a good idea?”
respondents often asked how much it would cost. Respondents
were also curious as to why the initial photo with Phragmites was
unnatural, when it looked natural. They appeared satisfied to
learn a few questions later how the landscape had been altered by
humans. Such synchronization within the survey provides evidence
that its structure was cognitively sound; it made sense; it flowed
well. Nevertheless, some considerations need to be discussed.
PAYMENT MECHANISM
The χ2 test indicated that the amount respondents were willing
to pay for the restoration was independent of payment mechanism.
Protest bids, however, appeared to be more prevalent among tax
payment mechanism surveys (almost one-fifth of all tax surveys).
One respondent who was willing to make a donation of $500-$1000
told interviewers that had the question been in the form of a tax
increase, his bid would have changed to $0. The statement, “I’m
sorry, but if its dealing with taxes, then I can’t support it,” occurred
several times.
Nevertheless, the proportion of people who were willing to pay a
positive amount for the restoration with taxes (73%) was similar to
those willing to pay with donations (71%). Therefore, it does not
appear as though the subset of people who were willing to pay for
the restoration was biased by payment mechanism, and our mean
WTP value was not biased in this way.
GENDER BIAS
More males (53%) than females (37%) were interviewed. (Some
respondents left this portion of the follow-up written questionnaire
blank, explaining the discrepancy from 100%). More males than
females answered the door. Also, men tended to participate when
both men and women were present in the house. But the decision
whether or not to make a donation or support a new tax is often a
household decision, and our data analysis focused on households.
Often, surveys with individuals were quickly transformed into
household discussions. For example, a couple debated how to rate
the current condition of the river for wildlife habitat. In another
home, a wife convinced her husband to change his preference to the
salt-marsh photo because she was concerned about the least tern.
These decision-making processes are common, and thus represent
appropriate responses to this survey.
UNCERTAINTY
Most economic surveys, CVM and otherwise, focus on demand
for goods that are certain with regard to quantity, quality, and timing
of availability. For environmental goods such as wildlife, however,
these are not always certain. Threat of extinction may decrease for a
rare bird, but the extent of decrease may not be defined. Respondents
in a contingent valuation survey must determine willingness to pay
even though they may not know if the species will survive long
enough for them to experience a benefit.
Whitehead (1993) studied the effects of uncertainty on existence
values of marine and coastal wildlife in North Carolina. His two
main findings were that (1) there is evidence that “total economic
values under uncertainty are theoretically valid”, and (2) omitting
consideration of uncertainty from CVM studies will produce a
distorted value for willingness to pay. The first conclusion lends
support for the use of CVM as a tool to elicit nonuse values for
wildlife; the second implies that inclusion of uncertainty is necessary
for a valid CVM survey.
Uncertainty became a factor in the West River survey. For ex-
ample, even if the salt marsh was restored, respondents did not
know the probability of seeing a least tern during a visit to the river.
Thus, there was a risk involved in making a payment for that resource.
Respondents were also told that the restoration would increase risk
of flooding. But interviewers, as part of the survey design, would not
give specific details regarding where flooding would occur, or how
much more likely it would be. Respondents thus had to weigh un-
certain harms against uncertain environmental benefits.
WARM GLOW AND FAIR SHARE EFFECTS
Stevens et al. (1994) studied the change of existence values for
wildlife over time and found that they were stable. The reason for
this stability, they found, was not due to the respondents’ economic
valuation of wildlife. Instead, the WTP was for something other than
the wildlife. One such value is the satisfaction of contributing to a
“worthy cause.” Since giving to “good causes” and charities tends to
be stable for individuals and households over time, this would be
reflected in the stated willingness to pay for wildlife.
Kahneman and Knetsch (1992) have termed this the “warm
glow effect,” in which people derive value from knowing they have
contributed to a good cause. Such respondents are purchasing
the price of moral satisfaction, and one cause may be just as
worthy as another.
Some of the responses to the West River survey suggest that the
warm glow effect might have influenced respondents. When asked
to state why they were willing to pay money to restore the West
River, many people gave responses such as “to help the bird” or “to
improve the community”. These statements could imply the subjects
viewed their WTP as a “charitable” donation to a “worthy cause.”
Another factor that may influence WTP for environmental
amenities is the tendency to calculate WTP based on the perceived
total cost of the project divided by the assumed population of people
who would have to contribute. Instead of constructing an individual
existence value, respondents figure what their “fair share” would be.
The rationale is, if everyone had to pay, it would only cost me a
minimal amount. Survey respondents calculate this in two ways: out
of a sense of duty to do their fair share (Stevens et al. 1994), or from
belief that a certain amount is the maximum they should have to
pay (Schkade and Payne 1993) because total cost is spread out over
numerous individuals.
A few respondents in our survey gave WTP responses as the
maximum cost of the project divided by the population they figured
would be taxed. “If everyone in New Haven and West Haven was
charged for this, it couldn’t be more than $10 per person,” one
respondent concluded. The “fair share effect” has been offered as
evidence that WTP bids are invalid because they do not represent a
person’s true value for an environmental good. However, an alter-
nate interpretation could be that individuals motivated by their
perception of fairness would not be willing to pay more for an ame-
nity, because they would then bear an unjust proportion of the costs.
Rather than concluding that warm glow and fair share effects
render CVM unable to obtain the monetary amount of nonuse
values, it should be noted that individuals create their own valuation
priorities. Hanemann (1994) argues that people have a wide variety
of motives that help them determine their values for both market
and nonmarket goods. Why a person would be willing to pay a
certain price to restore the West River is perhaps less important than
the fact that they would be willing to pay. Such values still would
indicate a level of support for the restoration, and could be incorpo-
rated into a cost-benefit analysis.
COSTS AND BENEFITS
Respondents placed an economic value on the nonmarket envi-
ronmental benefits of the West River salt-marsh restoration. Mean
WTP could be aggregated over the population of households that
would benefit from the restoration to approximate the value of the
restoration. How would that value compare to the costs of the resto-
ration? The major potential costs of the project would be associated
with mitigating potential flooding. Fortunately, no residential areas
are likely to be affected by the minimal water elevation increase
needed to eradicate the Phragmites (Barten and Kenny, this volume).
Most land use adjacent to the restoration zone is nonresidential
(Page, this volume), and the few residential areas are much higher
than the marsh. Of greater concern would be potential impacts to
recreation facilities in the immediate flood plain, such as playgrounds,
sport fields, and the Connecticut Tennis Center.
One option to eliminate potential damage to recreational areas as
a result of opening tide gates at Route 1 would be to install an addi-
tional tidal control structure at Route 34. Thus, salt marsh could be
established between Route 1 and Route 34, while recreational facilities
north of Route 34 would be protected. Another option would be to
build dikes around areas of concern. (Dikes for a similar project in
Fairfield Connecticut cost $250,000, Steinke 1986.) Finally, self
regulating tide gates that allow tidal flushing – but close during very
high tides – could be installed to replace the existing tide gates at
Route 1 (Barten and Kenny, this volume). Self regulating tide gates
cost around $30,000 (Steinke 1986).
Other costs associated with the project would include a design
study and, possibly, removal of earth to lower the marsh surface so
that restoration could occur with less increase in water elevation.
These tasks would be performed under the Connecticut Department of
Environmental Protection’s (CT DEP) Wetland Restoration Program.
CONCLUSION
The contingent valuation survey performed well and was effec-
tive in obtaining willingness to pay (WTP) and thus value estimates.
There was strong public support for the restoration project: over
three-quarters of people interviewed supported the proposed open-
ing of the tide gates. These results suggest a large nonmarket value
associated with restoration benefits.
Yet this nonmarket value would be excluded from consideration
in traditional economic decision-making, because there is no
associated price. Without a means to estimate their value, significant
amenities would be excluded from policy decisions. Contingent
valuation is an effective method for obtaining price estimates for
nonmarket values, thus balancing the decision-making process.
Consequently, contingent valuation has a crucial role to play in
preserving and restoring environmental resources.
The end of the road with regard to the West River restoration
has not yet been reached. Further public discussion and debate lie
ahead, and there are opportunities for further research. This study,
which comprised a small-scale contingent valuation survey, could be
expanded in scope to increase its accuracy. This type of survey
should also be periodically repeated over the long term, so that
nonmarket environmental values are updated. In the case of the
West River, it would be useful to determine if economic values of
nonmarket benefits that accrue from restoration change as the
restoration proceeds.
ACKNOWLEDGMENTS
The Connecticut Sea Grant College Program and the Yale
Center for Coastal and Watershed Systems provided financial
assistance for this research. Dr. Robert Mendelsohn and David
Casagrande provided advice on survey structure. Dr. Mendelsohn
and Christina Page provided comments on the manuscript.
Thanks are also due to Katie Bannon, Deron Chang, Jonathan
Kohl, Duncan Schmitt, and Nick Slosser for their participation in
the focus group and pretest, and to Sue Ask, Iliana Ayala, Todd
Campbell, David Casagrande, Celia Hartmann, and Rob Naeser
for conducting door-to-door interviews.
APPENDIX. WEST RIVER SURVEY - TAX PAYMENT VEHICLE 1
This is a survey that will be asking for your opinions and ideas about the West River, in New Haven,
Connecticut. I appreciate your taking the time to assist me and help complete it.
1. Do you ever visit the West River or walk along roads that cross it?
Yes (go to question 2) No (go to question 6)
2. Where along the river do you go to when you visit it? If you walk along roads that cross it, which roads
do you see the river from?
3. What do you do when you visit the West River?
4. Where along the river do you do these things, and how often?
5. Do you ever see birds or other wildlife when you are at the West River?
Yes No
(After completing question 5, please skip to question 7)
6. Why don’t you visit the West River?
7. On a scale of one to ten, how clean do you think the water in the West River is? A one means you think
the water is so dirty you wouldn’t even feel safe boating in it. A three means you think it is clean enough
to boat on it. A five means you would feel safe eating a fish caught in the West River. A seven means you
would feel the water in the river is clean enough to swim in. A ten means you think the water is clean
enough to drink straight out of the river.
Nothing Boat Fish Swim Drink
1 2 3 4 5 6 7 8 9 10 don’t know
8. On a scale of one to ten, how good do you think the West River is as a recreational site? A one means you
think it has no value at all for recreation. A ten means you feel it is the best recreational site in Con-
necticut.
1 2 3 4 5 6 7 8 9 10 don’t know
9. On a scale of one to ten, how good do you think the West River is as a place for wildlife and birds to live?
A one means you think no birds or animals can live there. A ten means you think the West River is
better than any other place in Connecticut for birds and animals.
1 2 3 4 5 6 7 8 9 10 don’t know
1
Only the survey format that used a one time tax as the willingness to pay vehicle is included here. The donation vehicle survey differed only in the
wording of questions 17, 18, 20, 21, 22, and 24, in which language regarding a one-time donation was substituted for the one-time tax. Also, the
sentence that stated, “Everyone would pay their fair share, and the tax would be no greater than the costs of the project.” was not included in
questions 17 and 21 of the donation survey.
Here are pictures of the lower stretches of two Connecticut rivers.
(Show Picture A, which shows a river with a Phragmites monoculture)
(Show Picture B, which shows a Spartina salt marsh)
Most rivers in coastal Connecticut resemble either Picture A or Picture B.
10. Based on what you see in the pictures, do you prefer a river that looks like the one shown in Picture A
or Picture B?
Picture A Picture B no preference
The two rivers in the picture provide different types of habitat for wildlife and birds. Here is a photo of
a bird called the least tern. It is an endangered bird species.
(Show photo of least tern)
11. Have you ever seen a least tern before?
Yes No
If yes, where and when?
The least tern can make its home in River B, but not in River A.
12. Now that you know that River B supports the least tern, which river do you now prefer?
River A River B no preference
River B can sometimes cause flooding problems for houses and buildings that are located along the river.
River A does not cause flooding problems.
13. Now that you know that River B can cause flooding problems and River A does not, which river do you
prefer?
River A River B no preference
Historically, the lower part of most rivers in Coastal Connecticut looked like Picture B. In other words,
Picture B shows the natural state of these rivers. People have altered the natural state of these rivers in
several ways. As a result of these human actions, many rivers in Connecticut now look like the river
shown in Picture A.
14. Now that you know that River A has been altered by humans and River B is in its natural state, which
river do you prefer?
River A River B no preference
Here is some more information about River A and River B. (Hand list to interviewee.) Each of the rivers
has both advantages and disadvantages, which are listed on your sheet. (Read list out loud.)
River A River B
Is protected from flooding. Can cause flooding to adjacent properties.
Has playgrounds and soccer fields by it. Is a productive and diverse salt marsh.
Has been altered by humans. Is in its natural historic state.
An aggressive grass called Phragmites has Is not dominated by Phragmites.
taken over the river. It grows 12-15 feet high
and crowds out other plants and animals.
Provides habitat for many types of animals and
The Phragmites breeds more insects, like
birds, including endangered species like the least
ticks and mosquitoes.
tern.
The Phragmites burns frequently in the sum-
Has very little fire danger.
mer in intense fires.
15. Now that you know these things about the two rivers, which river do you prefer?
River A River B no preference
This next question will be about the West River. Historically, the lower part of the West River looked
like the river in Picture B. Here is a map of the area that used to look like Picture B.
(Show map of historic salt marsh.)
In 1919, tide gates were installed at the point where Route 1 crosses the West River
(Point to this on the map.)
These tide gates changed the West River and caused it to look like Picture A.
Suppose there was a proposal to restore the West River to its historic condition so that it would look
like the river in Picture B. The restoration would affect about 70 acres along the lower part of the river.
(Show map again.)
The restoration would restore the river to its natural, historic state, and it would also provide habitat for
the least tern, an endangered bird species. This restoration would be done by opening the tide gates.
When this was tried in other rivers along the Connecticut coast, the rivers were successfully restored to
where they looked like Picture B.
16. Do you think this project is a good idea?
Yes No no opinion
Why or why not?
(Note, if they answered Yes to question 16, go to question 17. If they answered No to question 16, go to
question 21. If they answered no preference, go to question 25.)
17. You had answered that you think this restoration project is a good idea. Suppose that in order to pay for
this project, there would be a one-time tax next year. Everyone would pay their fair share, and the tax
would be no greater than the costs of the project. However, I am interested in finding out how much
restoring the West River to its natural state is worth to you. Would you be willing to pay additional
taxes next year to restore the West River? Keep in mind that any money you pay in additional taxes will
not be available for you to spend on other things. Also keep in mind that this tax would occur only
once, during next year.
Would you be willing to pay additional taxes to restore the West River?
Yes (go to question 18) No (go to question 20)
18. What is the maximum amount that you would be willing to pay in additional taxes?
19. Why did you choose this amount?
(After completing question 19, go on to question 25.)
20. Why did you say you would not be willing to pay additional taxes to restore the West River?
(After completing question 20, go on to question 25.)
21. You had answered that you do not think the restoration project is a good idea. Suppose that the only
way to avoid doing the restoration project was to do additional construction and repair of the tide gates
where the West River crosses Route 1.
In order to pay for this repair, there would be a one-time tax next year. Everyone would pay their fair
share, and the tax would be no greater than the costs of the project. However, I am interested in finding
out how much keeping the West River in its current condition is worth to you. Would you be willing to
pay additional taxes next year to repair the tide gates? Keep in mind that any money you pay in addi-
tional taxes will not be available for you to spend on other things. Also keep in mind that this tax would
occur only once, during next year.
Would you be willing to pay additional taxes to repair the tide gates?
Yes (go to question 22) No (go to question 24)
22. What is the maximum amount that you would be willing to pay in additional taxes?
23. Why did you choose this amount?
(After completing question 23, please go to question 25.)
24. Why did you say you would not be willing to pay additional taxes to restore the West River?
25. At this point, reconfirm that the respondent either
1) expressed support for the restoration project.
2) expressed opposition to the restoration project.
3) were indifferent about the restoration project.
If they have changed their minds, ask why, and note response.
Also reconfirm their willingness to pay, either 0, or if greater than 0, the amount they stated. Again, allow
changes if they change their mind.
26. You are almost at the end of the survey. Please tell me in your own words why you think we are doing
this survey.
27. Is there anything else you would like to say about the West River?
The last few questions are for demographic purposes. The answers are private and confidential. I am
now giving you this page to fill out yourself. When you are done, please fold it and I will seal it. (Hand
the next page to them.)
28. Age (please circle one) 30. Household income (please circle one)
A Under 20 A Under $10,000
B 20-29 B $10,000-$19,999
C 30-39 C $20,000-$29,999
D 40-49 D $30,000-$39,999
E 50-59 E $40,000-$49,999
F 60-69 F $50,000-$59,999
G Over 70 G $60,000 and over
29. Gender
M F
31. Number of years of education (please circle one) 32. How would you describe your ethnicity?
A Less than 12 years A Hispanic
B Completed high school B Black and Hispanic
C Attended college C Black, non-Hispanic
D Completed college D White and Hispanic
E Attended graduate school E White, non-Hispanic
F Completed graduate school F Asian
G Native American/American Indian/
First Nations
H Other (please specify)
Now, please fold this paper and the interviewer will seal it.
Those are all the questions. Thank you very much for your time in completing this survey!
REFERENCES
Bontje, M. P. 1987. The Application of science and engineering to restore a salt marsh. In Proceedings of
the 15th annual conference on wetlands restoration and creation, 16-23. Tampa, FL: Hillsborough
Community College.
Boyle, K., and R. Bishop. 1987. Valuing wildlife in benefit-cost analysis: A case study involving endangered
species. Water Resources Research 23(5):943-950.
Brown, T. C., S. C. Barro, M. J. Manfredo, and G. L. Peterson. 1995. Does better information about the
good avoid the embedding effect? Journal of Environmental Management 44:1-10.
Cummings, R. G., D. S. Brookshire, and W. D. Schulze. 1986. Valuing environmental goods: An assessment of
the contingent valuation method. Savage, MD: Rowman & Littlefield.
Desvousges, W. H., F. R. Johnson, R. W. Dunford, S. P. Hudson, K. N. Wilson, and K. Boyle. 1993. Measuring
natural resource damages with contingent valuation: Tests of validity and reliability. In Contingent
valuation: A critical assessment, ed. by J. A. Hausman, 91-164. Amsterdam: North Holland.
Diamond P. A., and J. A. Hausman. 1994. Contingent valuation: Is some number better than no number?
Journal of Economic Perspectives 8(4):120.
Diamond, P. A., J. A. Hausman, G. K. Leonard, and M. A. Denning. 1993. Does contingent valuation measure
preferences? Experimental evidence. In Contingent valuation: A critical assessment, ed. by J. A. Hausman,
41-89. Amsterdam: North Holland.
Hanemann, W. M. 1994a. Valuing the environment through contingent valuation. Journal of Economic
Perspectives 8(4):19-43.
Kahneman, D., and J. L. Knetsch. 1992. Valuing public goods: The purchase of moral satisfaction. Journal
of Environmental Economics and Management 22:57-70.
McFadden, D. and G. K. Leonard. 1993. Issues in the contingent valuation of environmental goods:
Methodologies for data collection and analysis. In Contingent valuation: A critical assessment, ed.
by J. A. Hausman, 165-215. Amsterdam: North Holland.
Mitchell, R. C., and R. T. Carson. 1989. Using surveys to value public goods: The contingent valuation
method. Washington, DC: Resources for the Future.
Portney, P. R. 1994. The Contingent valuation debate: Why economists should care. Journal of Economic
Perspectives 8:3-17.
Post, W. 1970. Salt marsh (breeding bird census no. 34). Audubon Field Notes 24:771-772.
Roe, B., K. Boyle, and M. Teisl. 1996. Using conjoint analysis to derive estimates of compensating variation.
Journal of Environmental Economics and Management 31:145-159.
Rozsa, R., and R. A. Orson. 1993. Restoration of degraded salt marshes in Connecticut. In Proceedings of
the 20th annual conference on wetlands restoration and creation, 196-205. Tampa, FL: Hillsborough
Community College.
Samples, K. C., J. A. Dixon, and M. M. Gowen. 1986. Information disclosure and endangered species
valuation. Land Economics 62(3):306-312.
Schkade, D. A., and J. W. Payne. 1993. Where do the numbers come from? How people respond to contingent
valuation questions. In Contingent valuation: A critical assessment, ed. by J. A. Hausman, 271-293.
Amsterdam: North Holland.
Steinke, T. J. 1986. Hydrologic manipulation and restoring wetland values: Pine Creek, Fairfield, Connecticut.
In National wetland symposium: mitigation of impacts and losses, ed. by J. A. Kusler, M. L. Quammen,
and G. Brooks, 377-383. Berne, NY: Association of State Wetland Managers.
Stevens, T. H., J. Echeverria, R. J. Glass, T. Hager, and T. A. More. 1991. Measuring the existence value of
wildlife: What do CVM estimates really show? Applied Economics 67(4):390-400.
Stevens, T. H., T. A. More, and R. J. Glass. 1994. Interpretation and temporal stability of CV bids for wildlife
existence: A Panel study. Land Economics 70(3):355-363.
Whitehead, J. C. 1993. Total economic values for coastal and marine Wildlife: Specification, validity and
valuation issues. Marine Resource Economics 8:119-132.
MATTHEW UDZIELA
Matthew Udziela has been involved in planning, research, and implementation of urban revitilization initiatives,
specializing in urban streams in Oregon, Connecticut, and Baltimore, Maryland. He received a B.A. in Geography and
Environmental Studies from the University of British Columbia, Vancouver in 1993 and a M.E.S. from The Yale School
of Forestry and Environmental Studies (F&ES) in 1996. He is currently a Peace Corps volunteer in Uganda.
LYNNE L. BENNETT
Lynne Bennett, Assistant Professor of Natural Resource Economics at Yale F&ES since 1994, researches and teaches the
economics of water resource allocation and pollution control. Her primary interest is the economics of water resources
in arid climates. Prior to joining Yale F&ES, Dr. Bennett was with the Environmental and Societal Impacts Group at the
National Center for Atmospheric Research in Boulder, Colorado. She received her B.A. in International Affairs and her
M.A. in Economics, from the University of Colorado at Boulder. She received her Ph.D. in economics from the University
of Colorado and received the Universities Council on Water Resources Outstanding Water Resources Dissertation
Award in 1995.
All correspondence through: Lynne L. Bennett, 205 Prospect St., New Haven, CT 06511, Tel: (203) 432-6258, Fax:
(203) 432-3817, lynne.bennett@yale.edu
Contingent Valuation of an Urban Salt Marsh Restoration
Matthew K. Udziela
Yale School of Forestry and Environmental Studies
Lynne L. Bennett
Yale School of Forestry and Environmental Studies
ABSTRACT
This paper discusses the validity of the contingent valuation method (CVM) and presents the results of a CVM survey
we conducted to estimate the nonmarket value of restoring a salt marsh in New Haven, Connecticut. The literature
suggests that CVM is valid, but results are often biased by poor survey design. Using a survey that was extensively
pretested, we addressed potentially problematic issues such as payment vehicle bias, information bias, and hypothetical
bias. We also combined techniques from conjoint analysis. Fifty-one households near the West River in New Haven
were surveyed on Saturdays in March and April 1996. The households represented a wide range of socio-economic
characteristics. Survey respondents were provided with background information including the current condition of the river
and possible costs and benefits of restoration. Respondents were asked if they supported or opposed the restoration.
Their maximum willingness to pay either for the restoration, or to keep the river in its current condition was solicited.
The contingent valuation survey performed well and was effective in educating participants about the current
condition of the West River, the restoration measures proposed, and the possible impacts of the restoration; it was also
effective in obtaining willingness to pay responses from people interviewed. Over three-quarters of people interviewed
supported the proposed opening of the tidegates. Moreover, this support extended to a willingness to pay considerable
amounts of money for the restoration. People clearly have high values for nonmarket and even nonuse environmental
resources. Without a means to estimate their economic value, these significant amenities will be ignored and thus not
protected. Contingent valuation can be an effective method of obtaining price estimates for non-market values that
would be excluded from consideration in traditional economic decision-making because they have no associated price.
Consequently, economic valuation has a crucial role to play in preserving and restoring priceless environmental resources.
We conclude that a carefully designed contingent valuation survey is a promising method for bringing nonmarket
economic values into the cost-benefit analysis of restoring the West River salt marsh.
The West River and its flood plains occupy approximately 35 square
miles in New Haven and West Haven, Connecticut. Historically, 800
acres of this area were salt marsh (Casagrande, pp. 13-40, this volume).
It was a highly productive estuarine ecosystem with an associated bio-
logical community.
In 1919 tidal gates were installed where Route 1 crosses the river
in New Haven for the purpose of marsh reclamation. The gates
stopped Long Island Sound’s saltwater from traveling upstream,
thus reducing upstream salinity and tidal inundation. As a result,
Phragmites australis have aggressively colonized the area north of the
tide gates. Such large, man-made stands of Phragmites provide poor
wildlife habitat.
There is currently significant interest and some controversy with
regard to restoring about 70 acres of the historic salt marsh located
just upstream of the gates within West River Memorial Park. Both
the interest and the controversy revolve around the effects of opening
some of the 12 tide gates to increase tidal flushing. Other restoration
projects in Connecticut have shown that restored tidal flushing
reduces Phragmites and allows a salt-marsh community to return
(Steinke 1986, Rozsa and Orson 1993). However, opening the tidal
gates could also have negative impacts such as flooding.
For a policy decision regarding the restoration to be balanced
and objective, it is helpful to aggregate economic benefits and costs.
Economic valuation is complicated, however, by the fact that resto-
ration benefits include salt-marsh resources that do not have a mar-
ket price (e.g., scenic views and wildlife). In other words, value is not
For a policy decision regarding the
determined or set by the market. Although scenic views and wildlife
restoration to be balanced and objective,
have a significant value to the public, they risk being excluded from
it is helpful to aggregate economic benefits
economic analysis because they appear to have no economic value. and costs. Although scenic views and
The historic eradication of salt marshes provides evidence that the wildlife have a significant value to the
exclusion of nonmarket values can result in environmentally destruc- public, they risk being excluded from
tive policies (Casagrande, pp. 13-40, this volume). For this reason, we economic analysis because they appear
have focused on the nonmarket benefits and costs of restoration. to have no economic value.
The most widely recognized method for the valuation of
1
See Portney 1994, Cummings et al. 1986, and
nonmarket resources is the contingent valuation method (CVM).
Mitchell and Carson 1989 among others for
CVM involves presenting survey respondents with a hypothetical details of methodology.
market for the resource and eliciting a willingness to pay for that 2
See for example Samples et al. 1986, Boyle
resource.1 Using CVM to value natural resources is not new. CVM and Bishop 1987, and Stevens et al. 1991.
has been used in several studies of endangered species.2 Other stud-
ies have focused more generally on wildlife preservation (Stevens et
al. 1994, Desvousges et al. 1993).
CVM has been the subject of debate among economists and
others for the last decade. Even though a panel of experts convened
by the National Oceanic and Atmospheric Administration (NOAA)
in 1993 concluded that contingent valuation could produce reliable
estimates of some types of values, critics have stated that CVM has
failed to produce reliable and accurate measures of value (Diamond
and Hausman 1994, McFadden and Leonard 1993, Diamond et al.
1993). Others emphasize that most (or even all) of CVM’s perceived
failures have been due to surveys that were inadequately planned or
poorly structured, and that CVM works when properly executed
(Portney 1994, Hanemann 1994). This study helps to alleviate some
concern. We argue that a careful survey design can eliminate many
of the biases that have caused the controversy.
We used the CVM method to assess the nonmarket economic
value of restoring the salt marsh in West River Memorial Park.
Our goal was to produce a reliable estimate by adhering to survey
methods that recent studies suggest can reduce bias.
SURVEY DESIGN
The survey instrument was carefully designed and pre-tested
using focus groups and practice interviews. The survey is presented
in the Appendix. The survey was conducted as personal interviews
and consisted of four key components.
A DESCRIPTION OF THE CONTEXT AND THE COSTS AND
BENEFITS TO BE VALUED
An important factor in the effectiveness of a contingent valuation
survey is that the survey respondent must have a clear and accurate
idea of the specific good that is being valued. In the case of natural
resources, this is often a difficult concept to convey, because there
are countless ways of viewing the resource. As a result, respondents
could be considering their willingness to pay for a good that is dif-
ferent from the one being asked about in the survey. Careful survey
design can help to avoid this information bias.
For example, the West River provides foraging habitat for the least
tern, a threatened species in Connecticut that dives for fish (Lewis
and Casagrande, this volume). The degraded condition of the marsh, Least tern
however, reduces the habitat value for the tern. Least terns are common
visitors to salt marshes in the northeast (Post 1970), and have been
shown to benefit from salt-marsh restoration (Bontje 1987). Open-
3
The CV study of migratory waterfowl by
ing the tide gates and restoring the salt marsh may benefit the least
Desvousges, et al. (l993) is commonly cited
tern by increasing fish abundance and improving water quality. as an example of the severity of the
Nevertheless, a survey question that asked for the value of the least embedding effect, because respondents
failed to distinguish between 2,000, 20,000
tern only could have been ambiguous. Could the question be referring and 200,000 birds. However, Hanemann
to the bird itself, reduced chance of extinction for the bird, probabil- (1994) has noted that when the data were
trimmed to remove the extreme 10% of
ity of viewing a least tern during a visit to the West River, or the
the values, the willingness to pay for each
habitat it lives in? of the three quantities of birds was
different.
Related to this is an issue known as “part-whole bias” or the
“embedding effect,” in which respondents produce equal values of
willingness to pay for two goods, one of which is a subcomponent of
the other (Brown et al., 1995). A good of interest, known as α, is
valued by one set of subjects. Then, a larger good (Σ), which includes
α, is valued by another set of subjects. Since studies of this type have
obtained results that α ≈ Σ,3 the conclusion is that either Σ is not
valued at all (which is unlikely), or that respondents value α as a
component of Σ. In our case, people may express a desire to preserve
the least tern because they value the habitat it lives in. In addition,
respondents may be valuing other benefits of the restoration, such
as enhanced aesthetics. Again, careful survey design can reduce the
magnitude of this effect or eliminate it entirely.
This research addressed the embedding effect by phrasing the
WTP question so that the good offered was a restored salt marsh
that contained several benefits, including a specific number of
acres (70) of habitat for the least tern, habitat for other wildlife,
“naturalness” of salt marshes, and reduced fire hazard. Likewise,
costs such as increased risk of flooding were included. Thus, respon-
dents were considering an entire bundle of costs and benefits, all of
which derive from the salt-marsh restoration. Pretesting of the
survey had indicated that it would be beneficial to present respon-
dents with all of the advantages and disadvantages of both condi-
A panel of experts convened by the
tions. They could then decide which “bundle” of benefits and
National Oceanic and Atmospheric
drawbacks they preferred. This technique is called conjoint analysis
Administration (NOAA) in 1993
and has been used extensively in marketing and transportation concluded that contingent valuation
economics and more recently in natural resource valuation (Roe et could produce reliable estimates of
al. 1996). Thus our survey not only incorporated CVM questions, values.
but included conjoint analysis questions as well.
If information is comprehensive and properly presented, bias
can be reduced. Therefore, we used visual aids extensively including
photographs and maps to present a detailed description of current
conditions and the effects of restoring tidal flushing.
Interviews began with questions about environmental attitudes
and behavior. Respondents were asked how often they used the
West River, where along the river they went, and what activities they
did there, or why did they not use the river. Respondents were also
asked about their perceptions of the West River’s water quality,
recreation potential, and habitat value. This allowed people to express
their opinions and become comfortable with the survey.
Respondents were then asked whether they preferred the West
River in its current state or as a restored salt marsh. This question
was based on two photographs, one of each condition. Therefore,
initial preference was based largely on aesthetics. A series of questions
then gave successive information about the two river types. After each
new fact was presented, respondents were asked which river they
preferred. Finally, respondents were handed a list of all the advantages
and disadvantages of each river type and asked to state a final prefer-
ence based on all the information.
Interviewers next described the restoration proposal. The restored
salt marsh would have both benefits (e.g., habitat for the least tern)
and costs (e.g., increased flooding). Respondents were then asked if
they supported or opposed the restoration.
THE WILLINGNESS TO PAY QUESTION
Economic value can be elicited with a CVM survey either using a
question of willingness to pay (WTP) for a resource or willingness to
accept (WTA) the loss of a resource. WTP is a measure of the maxi-
mum amount that a person would pay for a particular good or
service, such as environmental quality or species existence, or for an
incremental change in the amount of an environmental service
provided. WTA refers to the minimum level of compensation that
person would accept for losing those resources. Traditional economic
theory suggests that the two should be equivalent. But studies have
often found WTA to be larger than WTP (in some cases orders
of magnitude larger). The NOAA panel (1993), among others,
recommend using WTP since it elicits more conservative responses
of value.
WTP is a measure of the maximum
Furthermore, a WTP response implies the respondent has no
amount that a person would pay for
property right to the resource being valued. Alternatively, willing- a particular good or service, such as
ness to accept compensation for the loss of a resource would suggest environmental quality or species
that the respondent has a property right to the resource. We did not existence, or for an incremental
change in the amount of an
feel WTA applied to the West River situation and we elicited economic
environmental service provided.
value using WTP.
The WTP question can be open-ended or presented as a dichoto-
mous choice. In an open-ended question, respondents are asked for
the maximum amount they would be willing to pay to obtain a
particular environmental amenity. With dichotomous choice,
respondents are presented with a price and asked to indicate
whether or not they would be willing to pay the stated price. The
NOAA panel recommended dichotomous choice in CVM surveys,
because respondents are asked to make a familiar decision similar to
voting or deciding whether or not to buy something. An open-ended
question, the panel argued, places respondents in the unfamiliar
situation of determining a value. However, recent comparative
studies have revealed dichotomous choice WTP bids that are much
higher than those from open-ended questions (see for example,
McFadden and Leonard 1993). We used the open-ended format in
order to obtain a more conservative estimate of New Haven’s resi-
dents’ values of salt-marsh restoration. We also chose open-ended
questions in order to eliminate starting point bias, which is typically
associated with dichotomous choice methods.
If respondents in our survey expressed support for the restoration,
they were asked the maximum amount they would pay to ensure
that it was implemented. If respondents opposed the restoration, the
next question asked how much they would be willing to pay to
ensure that the restoration was not implemented. Thus, the WTP
question measured the economic value both for and against the
restoration. If a respondent gave a range of values, the lowest value
was used for analysis.
WTP can be biased if respondents do not think in the framework
of economic choices that they make with their own money. There
may be an overestimation of price if they ignore what they will
sacrifice by paying for the environmental amenity. When interview-
ing people, we kept the realism level high by making the scenario
very specific: payment would only occur once; it would occur within
the next year; and subjects were required to make their decision on
the basis of their present income, not their expected future income.
Respondents were also reminded that any amount they were willing
to pay would reduce their ability to purchase other goods and services.
In order not to bias WTP estimates, protest bids must be identi-
fied and removed. A protest bid is a stated response of $0 to a will-
ingness to pay question even though actual value might be greater
than zero. Protest bids occur when respondents are protesting some-
thing other than the good in question. Since these responses are not
valid “zeros,” they must be eliminated from the sample. A common
example of a protest bid is the objection to the type of payment
mechanism, such as a tax. This type of protest bid creates what is
called vehicle bias. To test for vehicle bias, half of our surveys were
conducted with a one-time tax as the payment mechanism. The other
half included a one-time private donation.
Protest bids can also result from a moral objection to placing a
price on the environment. At the end of interviews, we asked re-
spondents to give the reasons for their WTP value. This question
was included to identify protest bids, though it also provided other
useful information, such as confirmation that respondents under-
stood the restoration scenario.
SOCIOECONOMIC QUESTIONS
Socioeconomic data were solicited confidentially. Respondents
were asked to fill out a form about their age, education, income, and
race. They folded the form and sealed it in an envelope, which was
attached to the survey by the interviewer. This information was used
to determine the range of socioeconomic characteristics within our
sample, and to test for influence of income on WTP.
FOLLOW-UP QUESTIONS
We asked follow-up questions to ensure that respondents under-
stood the scenario and the payment format and believed the infor-
mation presented. Respondents were also asked to give reasons for
answers of “$0” WTP, so that we could identify protest bids.
SAMPLING DESIGN
Three neighborhoods were selected to represent the greatest
range of socio-economic characteristics adjacent to the West River.
The Hill is located in southern New Haven and borders the lowest
stretch of the West River downstream from the tide gates. Households
in the Hill are mostly Hispanic or African American and represented
the lowest income. Edgewood is located to the north of the Hill, and is
upstream from the tide gates. Edgewood is racially mixed, and incomes
are higher than the Hill. Westville is in the north of New Haven;
residents are mostly White and have high incomes.
Land use adjacent to the river differs for the three neighborhoods.
Land use along the river in the Hill is mostly commercial, landfill, or
junkyards. Adjacent to the Edgewood neighborhood, the West River
flows through Edgewood Park. Westville is bisected by the river, and
some homes are located close to the river. Westville suffers major
flooding about every 30 years. The most recent flood in 1982 caused
substantial loss of property.
Interviews were conducted with teams of two. This was initially
done for safety, but it was also found that one interviewer could ask
questions while the other handled the numerous visual aids. As soon
as the interview was completed, interviewers filled out a short evalu-
ation form to record time, date, address, payment mechanism, and
their impression of the interview.
Interviewers went door-to-door to survey people in their homes
during four Saturday afternoons in March and April. Each interview
team was assigned a block and called at every house. People who
were home and willing to participate were interviewed. In order to
reduce voluntary response bias, subjects were offered $5 in cash to
be interviewed. A total of 51 households were surveyed.
Interviewers who were native speakers of Spanish conducted
impromptu translations of the survey for Spanish-speaking respon-
dents. This enabled a wider cross-section of the population to be
interviewed.
Table 1. Number of respondents in income groups by neighborhood.
Income Hill Edgewood Westville Total % Total
$10,000-$19,999 4 2 0 6 14
$20,000-$29,999 4 3 1 8 19
$30,000-$39,999 2 2 1 5 12
$40,000-$49,999 1 2 2 5 12
$50,000-$59,999 0 2 0 2 5
> $60,000 0 6 7 13 30
not indicated 2 0 2 4 9
RESULTS
Of the 51 surveys, 19 were conducted in the Hill, 19 in Edgewood,
and 13 in Westville. Household income of respondents was diverse,
with an increasing gradient from the Hill (lowest), to Edgewood and
Westville (highest, Table 1). Respondents were also racially diverse –
53% were Hispanic or African American (Table 2).
We removed protest bids in order to compute willingness to pay
(WTP). A protest bid was considered a bid of $0 that was due to a
protest against the payment mechanism (tax or donation) or moral
objection to placing a dollar value on an environmental amenity.
Protest bids were identified using follow-up questions. Nine protests
to the tax mechanism were identified and removed. We did not
detect any moral protests.
Table 2. Race of respondents.
There were three respondents who were unable to determine
their WTP. These were not protest bids, but were removed in order
Number Percent
to calculate mean WTP. Respondents who wanted to pay, but were
Hispanic 14 25
not able to pay anything were considered to have given valid bids of
Black 13 28
$0, and were included in mean WTP.
White 22 43
Of the seven responses that opposed opening the tide gates
Asian 1 2
(Table 3), five were protest bids. This left two responses: a $20 tax
not indicated 1 2
bid and a $50 donation bid. The two bids against opening the tide
gates were included in mean WTP for salt-marsh restoration as
negative values. Of the 39 responses that favored opening the tide
gates (Table 3), 4 of the tax responses were protest bids. There were
no protest bids by respondents who favored opening the gates who
took the donation survey. Three others were unable to determine
WTP. Thus, there were 32 responses with valid WTP bids for open-
ing the tide gates; 17 by donation and 15 by tax.
Table 3. Responses to the proposal to open tide gates.
For the TAX survey, with 26 people surveyed:
3 people were against opening the tidegates (11%)
22 people were for opening the tidegates (85%)
1 person had no opinion (4%)
For the DONATION survey, with 25 people surveyed:
4 people were against opening the tidegates (16%)
17 people were for opening the tidegates (68%)
4 people had no opinion (16%)
For the TOTAL:
7 people were against opening the tidegates (14%)
39 people were for opening the tidegates (76%)
5 people had no opinion (10%)
Table 4. The contingency table for the χ2 test for the relationship between income and WTP.
WTP The expected values are:
Income ≤ Med > Med Total WTP
Income ≤ Med > Med
Low 14 5 19
High 5 8 13 Low 11.28 7.72
Total 19 13 32 High 7.72 5.28
There were many $100 bids (and a few $200 bids) and many bids
of $25 or less. Only one value occurred between $25 and $100. As a
result, there is a wide difference between mean WTP ($61.41) and
the median value ($25.00), and the standard deviation is large
($100.22). We used chi-square (χ2) tests of two-way contingency
tables to determine if WTP was influenced by income or payment
mechanism (tax vs. donation). A χ2 test was used because the WTP
values had a discrete probability distribution rather than continuous
(people tended to bid $25 or $100, but not values in between). Tests
were conducted at the 5% significance level.
WTP bids were divided into two categories: bids less than the
median, and bids greater than the median. Respondents were
grouped as those having household income below the median for
the survey, and those above (Table 4). The null hypothesis (Ho) was
that income and WTP are independent. There was one degree of
freedom because we used two-by-two contingency tables. Values for
χ2 were obtained from a reference table. WTP was dependent on
income (χ2 = 3.97).
Respondents’ bids were also grouped by payment mechanism
(Table 5). WTP bids were grouped as above. The null hypothesis
(Ho) was that payment mechanism and WTP were independent.
Final WTP values appeared to be independent of whether the
purchase mechanism was by tax or donation (χ2 = 0.43).
Table 5. The contingency table for the χ2 test for the relationship between payment mechanism and WTP.
The expected values are:
Payment WTP
mechanism ≤ Med > Med Payment WTP
Total
≤ Med > Med
mechanism
Tax 8 7 15
Tax 8.91 6.09
Donation 11 6 17
Donation 10.09 6.91
Total 19 13 32
DISCUSSION
Given that our sample size was relatively small, we view this
as a “pilot project” for what will ultimately be a survey of a larger
number of households. Testing our survey design for a small
sample allows us to determine whether or not to proceed with a
larger group.
The survey’s overall performance was satisfactory. People
generally understood the content, purpose, and logic, and
respondents tended to make WTP bids in terms of real economic
decisions. One respondent mentioned goods he would sacrifice
in his determination of WTP by deciding to “give up a couple of
twelve-packs.” Other comments included: “I know my budget for
April, and this is the amount I have free.” “I’m bringing up three
children; this is all I can spare.” “I can’t pay anything because I’m
not working right now.”
Respondents often anticipated questions. For example, when
asked, “Do you think the restoration project is a good idea?”
respondents often asked how much it would cost. Respondents
were also curious as to why the initial photo with Phragmites was
unnatural, when it looked natural. They appeared satisfied to
learn a few questions later how the landscape had been altered by
humans. Such synchronization within the survey provides evidence
that its structure was cognitively sound; it made sense; it flowed
well. Nevertheless, some considerations need to be discussed.
PAYMENT MECHANISM
The χ2 test indicated that the amount respondents were willing
to pay for the restoration was independent of payment mechanism.
Protest bids, however, appeared to be more prevalent among tax
payment mechanism surveys (almost one-fifth of all tax surveys).
One respondent who was willing to make a donation of $500-$1000
told interviewers that had the question been in the form of a tax
increase, his bid would have changed to $0. The statement, “I’m
sorry, but if its dealing with taxes, then I can’t support it,” occurred
several times.
Nevertheless, the proportion of people who were willing to pay a
positive amount for the restoration with taxes (73%) was similar to
those willing to pay with donations (71%). Therefore, it does not
appear as though the subset of people who were willing to pay for
the restoration was biased by payment mechanism, and our mean
WTP value was not biased in this way.
GENDER BIAS
More males (53%) than females (37%) were interviewed. (Some
respondents left this portion of the follow-up written questionnaire
blank, explaining the discrepancy from 100%). More males than
females answered the door. Also, men tended to participate when
both men and women were present in the house. But the decision
whether or not to make a donation or support a new tax is often a
household decision, and our data analysis focused on households.
Often, surveys with individuals were quickly transformed into
household discussions. For example, a couple debated how to rate
the current condition of the river for wildlife habitat. In another
home, a wife convinced her husband to change his preference to the
salt-marsh photo because she was concerned about the least tern.
These decision-making processes are common, and thus represent
appropriate responses to this survey.
UNCERTAINTY
Most economic surveys, CVM and otherwise, focus on demand
for goods that are certain with regard to quantity, quality, and timing
of availability. For environmental goods such as wildlife, however,
these are not always certain. Threat of extinction may decrease for a
rare bird, but the extent of decrease may not be defined. Respondents
in a contingent valuation survey must determine willingness to pay
even though they may not know if the species will survive long
enough for them to experience a benefit.
Whitehead (1993) studied the effects of uncertainty on existence
values of marine and coastal wildlife in North Carolina. His two
main findings were that (1) there is evidence that “total economic
values under uncertainty are theoretically valid”, and (2) omitting
consideration of uncertainty from CVM studies will produce a
distorted value for willingness to pay. The first conclusion lends
support for the use of CVM as a tool to elicit nonuse values for
wildlife; the second implies that inclusion of uncertainty is necessary
for a valid CVM survey.
Uncertainty became a factor in the West River survey. For ex-
ample, even if the salt marsh was restored, respondents did not
know the probability of seeing a least tern during a visit to the river.
Thus, there was a risk involved in making a payment for that resource.
Respondents were also told that the restoration would increase risk
of flooding. But interviewers, as part of the survey design, would not
give specific details regarding where flooding would occur, or how
much more likely it would be. Respondents thus had to weigh un-
certain harms against uncertain environmental benefits.
WARM GLOW AND FAIR SHARE EFFECTS
Stevens et al. (1994) studied the change of existence values for
wildlife over time and found that they were stable. The reason for
this stability, they found, was not due to the respondents’ economic
valuation of wildlife. Instead, the WTP was for something other than
the wildlife. One such value is the satisfaction of contributing to a
“worthy cause.” Since giving to “good causes” and charities tends to
be stable for individuals and households over time, this would be
reflected in the stated willingness to pay for wildlife.
Kahneman and Knetsch (1992) have termed this the “warm
glow effect,” in which people derive value from knowing they have
contributed to a good cause. Such respondents are purchasing
the price of moral satisfaction, and one cause may be just as
worthy as another.
Some of the responses to the West River survey suggest that the
warm glow effect might have influenced respondents. When asked
to state why they were willing to pay money to restore the West
River, many people gave responses such as “to help the bird” or “to
improve the community”. These statements could imply the subjects
viewed their WTP as a “charitable” donation to a “worthy cause.”
Another factor that may influence WTP for environmental
amenities is the tendency to calculate WTP based on the perceived
total cost of the project divided by the assumed population of people
who would have to contribute. Instead of constructing an individual
existence value, respondents figure what their “fair share” would be.
The rationale is, if everyone had to pay, it would only cost me a
minimal amount. Survey respondents calculate this in two ways: out
of a sense of duty to do their fair share (Stevens et al. 1994), or from
belief that a certain amount is the maximum they should have to
pay (Schkade and Payne 1993) because total cost is spread out over
numerous individuals.
A few respondents in our survey gave WTP responses as the
maximum cost of the project divided by the population they figured
would be taxed. “If everyone in New Haven and West Haven was
charged for this, it couldn’t be more than $10 per person,” one
respondent concluded. The “fair share effect” has been offered as
evidence that WTP bids are invalid because they do not represent a
person’s true value for an environmental good. However, an alter-
nate interpretation could be that individuals motivated by their
perception of fairness would not be willing to pay more for an ame-
nity, because they would then bear an unjust proportion of the costs.
Rather than concluding that warm glow and fair share effects
render CVM unable to obtain the monetary amount of nonuse
values, it should be noted that individuals create their own valuation
priorities. Hanemann (1994) argues that people have a wide variety
of motives that help them determine their values for both market
and nonmarket goods. Why a person would be willing to pay a
certain price to restore the West River is perhaps less important than
the fact that they would be willing to pay. Such values still would
indicate a level of support for the restoration, and could be incorpo-
rated into a cost-benefit analysis.
COSTS AND BENEFITS
Respondents placed an economic value on the nonmarket envi-
ronmental benefits of the West River salt-marsh restoration. Mean
WTP could be aggregated over the population of households that
would benefit from the restoration to approximate the value of the
restoration. How would that value compare to the costs of the resto-
ration? The major potential costs of the project would be associated
with mitigating potential flooding. Fortunately, no residential areas
are likely to be affected by the minimal water elevation increase
needed to eradicate the Phragmites (Barten and Kenny, this volume).
Most land use adjacent to the restoration zone is nonresidential
(Page, this volume), and the few residential areas are much higher
than the marsh. Of greater concern would be potential impacts to
recreation facilities in the immediate flood plain, such as playgrounds,
sport fields, and the Connecticut Tennis Center.
One option to eliminate potential damage to recreational areas as
a result of opening tide gates at Route 1 would be to install an addi-
tional tidal control structure at Route 34. Thus, salt marsh could be
established between Route 1 and Route 34, while recreational facilities
north of Route 34 would be protected. Another option would be to
build dikes around areas of concern. (Dikes for a similar project in
Fairfield Connecticut cost $250,000, Steinke 1986.) Finally, self
regulating tide gates that allow tidal flushing – but close during very
high tides – could be installed to replace the existing tide gates at
Route 1 (Barten and Kenny, this volume). Self regulating tide gates
cost around $30,000 (Steinke 1986).
Other costs associated with the project would include a design
study and, possibly, removal of earth to lower the marsh surface so
that restoration could occur with less increase in water elevation.
These tasks would be performed under the Connecticut Department of
Environmental Protection’s (CT DEP) Wetland Restoration Program.
CONCLUSION
The contingent valuation survey performed well and was effec-
tive in obtaining willingness to pay (WTP) and thus value estimates.
There was strong public support for the restoration project: over
three-quarters of people interviewed supported the proposed open-
ing of the tide gates. These results suggest a large nonmarket value
associated with restoration benefits.
Yet this nonmarket value would be excluded from consideration
in traditional economic decision-making, because there is no
associated price. Without a means to estimate their value, significant
amenities would be excluded from policy decisions. Contingent
valuation is an effective method for obtaining price estimates for
nonmarket values, thus balancing the decision-making process.
Consequently, contingent valuation has a crucial role to play in
preserving and restoring environmental resources.
The end of the road with regard to the West River restoration
has not yet been reached. Further public discussion and debate lie
ahead, and there are opportunities for further research. This study,
which comprised a small-scale contingent valuation survey, could be
expanded in scope to increase its accuracy. This type of survey
should also be periodically repeated over the long term, so that
nonmarket environmental values are updated. In the case of the
West River, it would be useful to determine if economic values of
nonmarket benefits that accrue from restoration change as the
restoration proceeds.
ACKNOWLEDGMENTS
The Connecticut Sea Grant College Program and the Yale
Center for Coastal and Watershed Systems provided financial
assistance for this research. Dr. Robert Mendelsohn and David
Casagrande provided advice on survey structure. Dr. Mendelsohn
and Christina Page provided comments on the manuscript.
Thanks are also due to Katie Bannon, Deron Chang, Jonathan
Kohl, Duncan Schmitt, and Nick Slosser for their participation in
the focus group and pretest, and to Sue Ask, Iliana Ayala, Todd
Campbell, David Casagrande, Celia Hartmann, and Rob Naeser
for conducting door-to-door interviews.
APPENDIX. WEST RIVER SURVEY - TAX PAYMENT VEHICLE 1
This is a survey that will be asking for your opinions and ideas about the West River, in New Haven,
Connecticut. I appreciate your taking the time to assist me and help complete it.
1. Do you ever visit the West River or walk along roads that cross it?
Yes (go to question 2) No (go to question 6)
2. Where along the river do you go to when you visit it? If you walk along roads that cross it, which roads
do you see the river from?
3. What do you do when you visit the West River?
4. Where along the river do you do these things, and how often?
5. Do you ever see birds or other wildlife when you are at the West River?
Yes No
(After completing question 5, please skip to question 7)
6. Why don’t you visit the West River?
7. On a scale of one to ten, how clean do you think the water in the West River is? A one means you think
the water is so dirty you wouldn’t even feel safe boating in it. A three means you think it is clean enough
to boat on it. A five means you would feel safe eating a fish caught in the West River. A seven means you
would feel the water in the river is clean enough to swim in. A ten means you think the water is clean
enough to drink straight out of the river.
Nothing Boat Fish Swim Drink
1 2 3 4 5 6 7 8 9 10 don’t know
8. On a scale of one to ten, how good do you think the West River is as a recreational site? A one means you
think it has no value at all for recreation. A ten means you feel it is the best recreational site in Con-
necticut.
1 2 3 4 5 6 7 8 9 10 don’t know
9. On a scale of one to ten, how good do you think the West River is as a place for wildlife and birds to live?
A one means you think no birds or animals can live there. A ten means you think the West River is
better than any other place in Connecticut for birds and animals.
1 2 3 4 5 6 7 8 9 10 don’t know
1
Only the survey format that used a one time tax as the willingness to pay vehicle is included here. The donation vehicle survey differed only in the
wording of questions 17, 18, 20, 21, 22, and 24, in which language regarding a one-time donation was substituted for the one-time tax. Also, the
sentence that stated, “Everyone would pay their fair share, and the tax would be no greater than the costs of the project.” was not included in
questions 17 and 21 of the donation survey.
Here are pictures of the lower stretches of two Connecticut rivers.
(Show Picture A, which shows a river with a Phragmites monoculture)
(Show Picture B, which shows a Spartina salt marsh)
Most rivers in coastal Connecticut resemble either Picture A or Picture B.
10. Based on what you see in the pictures, do you prefer a river that looks like the one shown in Picture A
or Picture B?
Picture A Picture B no preference
The two rivers in the picture provide different types of habitat for wildlife and birds. Here is a photo of
a bird called the least tern. It is an endangered bird species.
(Show photo of least tern)
11. Have you ever seen a least tern before?
Yes No
If yes, where and when?
The least tern can make its home in River B, but not in River A.
12. Now that you know that River B supports the least tern, which river do you now prefer?
River A River B no preference
River B can sometimes cause flooding problems for houses and buildings that are located along the river.
River A does not cause flooding problems.
13. Now that you know that River B can cause flooding problems and River A does not, which river do you
prefer?
River A River B no preference
Historically, the lower part of most rivers in Coastal Connecticut looked like Picture B. In other words,
Picture B shows the natural state of these rivers. People have altered the natural state of these rivers in
several ways. As a result of these human actions, many rivers in Connecticut now look like the river
shown in Picture A.
14. Now that you know that River A has been altered by humans and River B is in its natural state, which
river do you prefer?
River A River B no preference
Here is some more information about River A and River B. (Hand list to interviewee.) Each of the rivers
has both advantages and disadvantages, which are listed on your sheet. (Read list out loud.)
River A River B
Is protected from flooding. Can cause flooding to adjacent properties.
Has playgrounds and soccer fields by it. Is a productive and diverse salt marsh.
Has been altered by humans. Is in its natural historic state.
An aggressive grass called Phragmites has Is not dominated by Phragmites.
taken over the river. It grows 12-15 feet high
and crowds out other plants and animals.
Provides habitat for many types of animals and
The Phragmites breeds more insects, like
birds, including endangered species like the least
ticks and mosquitoes.
tern.
The Phragmites burns frequently in the sum-
Has very little fire danger.
mer in intense fires.
15. Now that you know these things about the two rivers, which river do you prefer?
River A River B no preference
This next question will be about the West River. Historically, the lower part of the West River looked
like the river in Picture B. Here is a map of the area that used to look like Picture B.
(Show map of historic salt marsh.)
In 1919, tide gates were installed at the point where Route 1 crosses the West River
(Point to this on the map.)
These tide gates changed the West River and caused it to look like Picture A.
Suppose there was a proposal to restore the West River to its historic condition so that it would look
like the river in Picture B. The restoration would affect about 70 acres along the lower part of the river.
(Show map again.)
The restoration would restore the river to its natural, historic state, and it would also provide habitat for
the least tern, an endangered bird species. This restoration would be done by opening the tide gates.
When this was tried in other rivers along the Connecticut coast, the rivers were successfully restored to
where they looked like Picture B.
16. Do you think this project is a good idea?
Yes No no opinion
Why or why not?
(Note, if they answered Yes to question 16, go to question 17. If they answered No to question 16, go to
question 21. If they answered no preference, go to question 25.)
17. You had answered that you think this restoration project is a good idea. Suppose that in order to pay for
this project, there would be a one-time tax next year. Everyone would pay their fair share, and the tax
would be no greater than the costs of the project. However, I am interested in finding out how much
restoring the West River to its natural state is worth to you. Would you be willing to pay additional
taxes next year to restore the West River? Keep in mind that any money you pay in additional taxes will
not be available for you to spend on other things. Also keep in mind that this tax would occur only
once, during next year.
Would you be willing to pay additional taxes to restore the West River?
Yes (go to question 18) No (go to question 20)
18. What is the maximum amount that you would be willing to pay in additional taxes?
19. Why did you choose this amount?
(After completing question 19, go on to question 25.)
20. Why did you say you would not be willing to pay additional taxes to restore the West River?
(After completing question 20, go on to question 25.)
21. You had answered that you do not think the restoration project is a good idea. Suppose that the only
way to avoid doing the restoration project was to do additional construction and repair of the tide gates
where the West River crosses Route 1.
In order to pay for this repair, there would be a one-time tax next year. Everyone would pay their fair
share, and the tax would be no greater than the costs of the project. However, I am interested in finding
out how much keeping the West River in its current condition is worth to you. Would you be willing to
pay additional taxes next year to repair the tide gates? Keep in mind that any money you pay in addi-
tional taxes will not be available for you to spend on other things. Also keep in mind that this tax would
occur only once, during next year.
Would you be willing to pay additional taxes to repair the tide gates?
Yes (go to question 22) No (go to question 24)
22. What is the maximum amount that you would be willing to pay in additional taxes?
23. Why did you choose this amount?
(After completing question 23, please go to question 25.)
24. Why did you say you would not be willing to pay additional taxes to restore the West River?
25. At this point, reconfirm that the respondent either
1) expressed support for the restoration project.
2) expressed opposition to the restoration project.
3) were indifferent about the restoration project.
If they have changed their minds, ask why, and note response.
Also reconfirm their willingness to pay, either 0, or if greater than 0, the amount they stated. Again, allow
changes if they change their mind.
26. You are almost at the end of the survey. Please tell me in your own words why you think we are doing
this survey.
27. Is there anything else you would like to say about the West River?
The last few questions are for demographic purposes. The answers are private and confidential. I am
now giving you this page to fill out yourself. When you are done, please fold it and I will seal it. (Hand
the next page to them.)
28. Age (please circle one) 30. Household income (please circle one)
A Under 20 A Under $10,000
B 20-29 B $10,000-$19,999
C 30-39 C $20,000-$29,999
D 40-49 D $30,000-$39,999
E 50-59 E $40,000-$49,999
F 60-69 F $50,000-$59,999
G Over 70 G $60,000 and over
29. Gender
M F
31. Number of years of education (please circle one) 32. How would you describe your ethnicity?
A Less than 12 years A Hispanic
B Completed high school B Black and Hispanic
C Attended college C Black, non-Hispanic
D Completed college D White and Hispanic
E Attended graduate school E White, non-Hispanic
F Completed graduate school F Asian
G Native American/American Indian/
First Nations
H Other (please specify)
Now, please fold this paper and the interviewer will seal it.
Those are all the questions. Thank you very much for your time in completing this survey!
REFERENCES
Bontje, M. P. 1987. The Application of science and engineering to restore a salt marsh. In Proceedings of
the 15th annual conference on wetlands restoration and creation, 16-23. Tampa, FL: Hillsborough
Community College.
Boyle, K., and R. Bishop. 1987. Valuing wildlife in benefit-cost analysis: A case study involving endangered
species. Water Resources Research 23(5):943-950.
Brown, T. C., S. C. Barro, M. J. Manfredo, and G. L. Peterson. 1995. Does better information about the
good avoid the embedding effect? Journal of Environmental Management 44:1-10.
Cummings, R. G., D. S. Brookshire, and W. D. Schulze. 1986. Valuing environmental goods: An assessment of
the contingent valuation method. Savage, MD: Rowman & Littlefield.
Desvousges, W. H., F. R. Johnson, R. W. Dunford, S. P. Hudson, K. N. Wilson, and K. Boyle. 1993. Measuring
natural resource damages with contingent valuation: Tests of validity and reliability. In Contingent
valuation: A critical assessment, ed. by J. A. Hausman, 91-164. Amsterdam: North Holland.
Diamond P. A., and J. A. Hausman. 1994. Contingent valuation: Is some number better than no number?
Journal of Economic Perspectives 8(4):120.
Diamond, P. A., J. A. Hausman, G. K. Leonard, and M. A. Denning. 1993. Does contingent valuation measure
preferences? Experimental evidence. In Contingent valuation: A critical assessment, ed. by J. A. Hausman,
41-89. Amsterdam: North Holland.
Hanemann, W. M. 1994a. Valuing the environment through contingent valuation. Journal of Economic
Perspectives 8(4):19-43.
Kahneman, D., and J. L. Knetsch. 1992. Valuing public goods: The purchase of moral satisfaction. Journal
of Environmental Economics and Management 22:57-70.
McFadden, D. and G. K. Leonard. 1993. Issues in the contingent valuation of environmental goods:
Methodologies for data collection and analysis. In Contingent valuation: A critical assessment, ed.
by J. A. Hausman, 165-215. Amsterdam: North Holland.
Mitchell, R. C., and R. T. Carson. 1989. Using surveys to value public goods: The contingent valuation
method. Washington, DC: Resources for the Future.
Portney, P. R. 1994. The Contingent valuation debate: Why economists should care. Journal of Economic
Perspectives 8:3-17.
Post, W. 1970. Salt marsh (breeding bird census no. 34). Audubon Field Notes 24:771-772.
Roe, B., K. Boyle, and M. Teisl. 1996. Using conjoint analysis to derive estimates of compensating variation.
Journal of Environmental Economics and Management 31:145-159.
Rozsa, R., and R. A. Orson. 1993. Restoration of degraded salt marshes in Connecticut. In Proceedings of
the 20th annual conference on wetlands restoration and creation, 196-205. Tampa, FL: Hillsborough
Community College.
Samples, K. C., J. A. Dixon, and M. M. Gowen. 1986. Information disclosure and endangered species
valuation. Land Economics 62(3):306-312.
Schkade, D. A., and J. W. Payne. 1993. Where do the numbers come from? How people respond to contingent
valuation questions. In Contingent valuation: A critical assessment, ed. by J. A. Hausman, 271-293.
Amsterdam: North Holland.
Steinke, T. J. 1986. Hydrologic manipulation and restoring wetland values: Pine Creek, Fairfield, Connecticut.
In National wetland symposium: mitigation of impacts and losses, ed. by J. A. Kusler, M. L. Quammen,
and G. Brooks, 377-383. Berne, NY: Association of State Wetland Managers.
Stevens, T. H., J. Echeverria, R. J. Glass, T. Hager, and T. A. More. 1991. Measuring the existence value of
wildlife: What do CVM estimates really show? Applied Economics 67(4):390-400.
Stevens, T. H., T. A. More, and R. J. Glass. 1994. Interpretation and temporal stability of CV bids for wildlife
existence: A Panel study. Land Economics 70(3):355-363.
Whitehead, J. C. 1993. Total economic values for coastal and marine Wildlife: Specification, validity and
valuation issues. Marine Resource Economics 8:119-132.
MATTHEW UDZIELA
Matthew Udziela has been involved in planning, research, and implementation of urban revitilization initiatives,
specializing in urban streams in Oregon, Connecticut, and Baltimore, Maryland. He received a B.A. in Geography and
Environmental Studies from the University of British Columbia, Vancouver in 1993 and a M.E.S. from The Yale School
of Forestry and Environmental Studies (F&ES) in 1996. He is currently a Peace Corps volunteer in Uganda.
LYNNE L. BENNETT
Lynne Bennett, Assistant Professor of Natural Resource Economics at Yale F&ES since 1994, researches and teaches the
economics of water resource allocation and pollution control. Her primary interest is the economics of water resources
in arid climates. Prior to joining Yale F&ES, Dr. Bennett was with the Environmental and Societal Impacts Group at the
National Center for Atmospheric Research in Boulder, Colorado. She received her B.A. in International Affairs and her
M.A. in Economics, from the University of Colorado at Boulder. She received her Ph.D. in economics from the University
of Colorado and received the Universities Council on Water Resources Outstanding Water Resources Dissertation
Award in 1995.
All correspondence through: Lynne L. Bennett, 205 Prospect St., New Haven, CT 06511, Tel: (203) 432-6258, Fax:
(203) 432-3817, lynne.bennett@yale.edu