Download:
pdf |
pdfWRITE-UP OF PRETEST RESULTS
Purpose
The Coral Reef Valuation Study measures the public’s preferences and valuation for protecting
and restoring Hawaii’s coral reef ecosystem. The results of the study will provide important
information to the National Oceanic and Atmospheric Administration’s Office of Response and
Restoration, the State of Hawaii’s resource managers, and other federal agencies that are actively
managing the coral reef resources of Hawaii. To ensure that the information being developed is
as useful as possible, these managers and other stakeholders have been involved in the design
and development process undertaken to date.
The goal of the overall study is to obtain estimates of the general U.S. population’s preferences
and willingness-to-pay (WTP) to protect coral reef ecosystems in the Main Hawaiian Islands
(MHI) and the Northwestern Hawaiian Islands (NWHI). The survey focuses on two of the most
widespread threats to the reef ecosystems: overfishing and ship groundings. The survey presents
three methods of protection: (1) restoration of the coral reef ecosystems of the MHI through
establishing a special category of Marine Protected Areas (MPAs) known as Marine Reserves or
no-fishing areas; (2) prevention of the future decline in the overall health of coral ecosystems
from overfishing in the NWHI1 (again through designation of no-fishing areas); and
(3) restoration of coral habitats after vessel groundings. The survey uses a stated choice
framework to evaluate respondents’ willingness to trade off these actions against each other at
different costs and against each other by taking no action.
As part of this study, we conducted a pretest of the survey instrument using the Knowledge
Network’s (KN’s) Internet Panel. The pretest provided an opportunity to evaluate the survey
instrument and obtain some simple results to help refine the main survey. This attachment
presents the results of the pretest.
Response rate
We calculated response rates using the panel recruitment response rate, the household profile
rate, and the survey completion rate in order to provide the final response rate. The panel
recruitment response rate reported by KN was 28.2%. The household profile rate — the
percentage of households recruited where an adult completed the demographic profile survey —
reported by KN was 57.1%. The survey completion rate, or the percentage of sampled cases that
completed the web survey, reported by KN was 69.2% (216 completed surveys out of the
312 surveys sent out). To calculate the overall response rate, we multiplied the panel recruitment
response rate by the household profile rate and the survey completion rate, to yield a response
rate of 11.1%.
1. This includes incorporating areas within the boundaries of the Northwestern Hawaiian Island Ecosystem
Reserve established by President Clinton in Executive Order 13178 and as modified in Executive Order 13196.
2-1
Methods
The pretest consisted of a small-scale survey, designed for implementation in the Internet mode,
with a sample of U.S. households (216 responses). KN administered the pretest survey to a
random sample of its Internet Panel. We designed the pretest to provide information on two key
issues: how well the survey would work under full field conditions and whether the preliminary
range of dollar values used in the pretest would be suitable for the final survey. As is standard in
survey development, the pretest results, along with the results from focus groups and cognitive
interviewers and the research team’s previous experience with nonmarket valuation of public
goods, will inform the dollar values to be used in the final survey.
KN administered the pretest survey in two phases. In phase 1, KN verified that all survey
programming, skip patterns, and other survey design features worked properly. Upon completion
of the first 50 surveys, we analyzed the responses to ensure compliance with established quality
assurance control measures. Upon successful completion of phase 1 of the pretest, KN
administered phase 2 to the remaining sample to garner 156 additional responses. Given the
selected implementation mode and budget constraint, this sample size was feasible and provided
sufficiently large numbers of observations to support a simple summary statistical analyses of the
data (e.g., means, medians, standard deviations, maximums, and minimums).
Given the pretest sample size, the pretest was limited to 6 survey versions, with respondents
randomly assigned to a version. Next, dollar amounts were assigned by version: some versions
had low costs and others had medium and high costs2 to balance the design across the dollar
range of interest.
Results
General Social Survey comparison
The instrument began with questions from the General Social Survey (GSS). The GSS is
comprised of standard demographic and attitudinal questions, as well as questions about special
interest topics. The GSS is conducted periodically and is considered authoritative when it comes
to trends in demographic characteristics and attitudes of U.S. residents. The latest data currently
available are from the 2006 GSS survey, however, we used results from the 2004 GSS survey
since the coral reef survey was conducted in 2004. We drew on the GSS for questions about
attitudes toward government spending on a number of social issues, including environmental
issues. The GSS has several versions of this question. We used two of the versions — one short
and one long — to compare our respondents to those who completed the GSS in 2004. We
placed these questions at the beginning of the pretest instrument, before coral reefs had even
been mentioned, both to serve as a warm-up to the survey and to provide information to evaluate
potential differences between the respondents and the general public. Table 2.1 summarizes how
the coral reef pretest respondents answered these questions compared to the GSS respondents in
2004 for the short version.
2. Based on preliminary interviews and a range of $0 (status quo) to $101 to be tested in the pretest.
2-2
Table 2.1. Comparison of the coral reef survey results to the 2004 GSS results for the short
version
Category
Space exploration
Survey
Coral Reef Pretest
GSS
The environment
Coral Reef Pretest
GSS
Health
Coral Reef Pretest
GSS
Assistance to big cities Coral Reef Pretest
GSS
Law enforcement
Coral Reef Pretest
GSS
Drug rehabilitation
Coral Reef Pretest
GSS
Education
Coral Reef Pretest
GSS
Are we spending too much, about the right amount,
or too little on these categories?
About the right
Too much
amount
Too little
47%
39%
14%
40%
42%
13%
4%
32%
64%
7%
27%
63%
7%
17%
76%
7%
14%
77%
38%
41%
21%
28%
41%
17%
7%
48%
45%
9%
36%
52%
14%
48%
38%
11%
34%
49%
5%
16%
78%
5%
17%
77%
Table 2.2 summarizes how the coral reef pretest respondents answered these questions compared
to the GSS respondents in 2004 for the long version.
Few differences exist between how our respondents answered these questions compared to the
2004 GSS participants. Tests for statistical significance do not show any significant differences
for the long version. For the short version, however, there are significant differences between the
GSS and coral respondents for drug rehabilitation and assistance to big cities categories. For
example, in the short version, more of our respondents feel that the government spends too much
on drug rehabilitation and on assistance to big cities. Our proposed design includes GSS
questions so that we can continue to compare our respondents to the population as a whole.
Part I: Set-up
The full survey is presented and its features are discussed in detail elsewhere in this document.
As a brief summary, once the GSS questions were administered, respondents were presented
with some background information on the purpose of the survey and the sponsors of the survey.
Other issues such as whether the respondent had audio capabilities on the computer system were
included for optional voice instructions during later parts of the survey. No survey questions
were asked in this section.
2-3
Table 2.2. Comparison of the coral reef survey results to the 2004 GSS results for the
long version
Are we spending too much, about the right
amount, or too little on these categories?
About the right
Too much
amount
Too little
37%
50%
14%
39%
43%
13%
9%
30%
61%
7%
29%
62%
5%
23%
72%
4%
17%
77%
16%
45%
40%
12%
37%
40%
3%
35%
62%
5%
35%
57%
11%
31%
58%
9%
33%
54%
5%
23%
72%
5%
22%
71%
Category
The space exploration
program
Survey
Coral Reef Pretest
GSS
Improving and protecting the Coral Reef Pretest
environment
GSS
Improving and protecting the Coral Reef Pretest
nation’s health
GSS
Solving the problems of big Coral Reef Pretest
cities
GSS
Halting the rising crime rate Coral Reef Pretest
GSS
Dealing with drug addiction Coral Reef Pretest
GSS
Improving the nation’s
Coral Reef Pretest
education system
GSS
Part II: Introduction
After being provided with some basic coral reef facts, respondents were prompted to answer
several questions on coral reefs. Table 2.3 summarizes how often respondents have read about or
seen TV programs about coral reefs.
Table 2.3. How often respondents have read about or seen TV programs about coral reefs,
either in U.S. waters or elsewhere
Response
Never
Sometimes, but not often
Often
Very often
Percent respondents
24.7%
65.6%
7.4%
2.3%
Respondents were then asked how many times they have ever been to a coral reef in the United
States or elsewhere. Table 2.4 summarizes these responses.
For the respondents who had visited a coral reef, the locations of visits are summarized in
Table 2.5.
2-4
Table 2.4. The number of times respondents have been to a coral reef in the United States
or elsewhere (for example, to fish, snorkel, scuba dive, or view marine life)
Response
Never
Once
A few times (2-4 times)
Several times (5-10)
Many times (more than 10 times)
Percent respondents
63.3%
12.6%
16.7%
4.2%
3.3%
Table 2.5. Summary of locations where respondents have visited a coral reef
Response
Florida
Puerto Rico or the U.S. Virgin Islands
Other Caribbean, Gulf of Mexico, or Atlantic locations
Hawaii
Pacific Ocean locations other than Hawaii
Other location
Percent of respondents who
have visited a coral reef
36.7%
12.7%
45.6%
41.8%
24.1%
8.9%
About 3% of the respondents have lived in Hawaii. Of the respondents who have never lived in
Hawaii, 27% have visited Hawaii. Table 2.6 summarizes how likely it will be that respondents
will travel to Hawaii in the next 10 years.
Table 2.6. Summary of how likely is it that respondent will go to Hawaii
Response
Not at all likely
Somewhat likely
Very likely
Don’t know
Percent respondents
35.0%
37.9%
16.4%
10.7%
Part III: Overfishing
After receiving a brief overview of overfishing in the MHI and NWHI and potential solutions to
overfishing, respondents were prompted to answer several questions about fishing practices in
Hawaii. Table 2.7 summarizes whether respondents feel that protecting jobs of commercial
fishermen and protecting recreational fishing are more important than protecting Hawaiian coral
reefs, as well as whether it is important for the federal government to take an active role in trying
to protect the reefs.
2-5
Table 2.7. Summary of how respondents feel about these statements
Response
Strongly agree
Somewhat agree
Neither agree nor disagree
Somewhat disagree
Strongly disagree
Protecting jobs of
Protecting recreational
It is important for
commercial fishermen
fishing is more
the federal government to
is more important than
important than
take an active role in
protecting Hawaiian
protecting Hawaiian trying to protect Hawaiian
coral reefs
coral reefs
coral reefs
2.8%
2.3%
46.5%
7.9%
4.2%
31.2%
20.5%
14.9%
13.0%
36.3%
20.5%
4.7%
32.6%
58.1%
4.7%
Table 2.8 summarizes the degree to which respondents oppose or support increasing federal
taxes to expand no-fishing zones in the MHI versus the NWHI. Just less than half of the
respondents at least somewhat to strongly support the idea of increasing federal taxes to expand
no-fishing zones around the NWHI. Nearly 25% of the respondents at least somewhat to strongly
oppose increasing taxes for that purpose. Tests for significance show that there is no significant
difference in the overall pattern between how respondents feel about increasing taxes to expand
no-fishing zones around the MHI and the NWHI. In general, respondents support increasing
federal taxes to expand no-fishing zones slightly more for around the MHI than for the NWHI.
Table 2.8. Summary of how respondents feel about increasing federal taxes to expand nofishing zones around the Main/Northwestern Hawaiian Islands
Response
Strongly oppose
Somewhat oppose
Neither oppose nor support
Somewhat support
Strongly support
Percent respondents
Main Hawaiian Islands
Northwestern Hawaiian Islands
10.7%
11.7%
7.9%
13.1%
27.6%
25.4%
34.1%
31.5%
19.6%
18.3%
Part IV: Ship accidents
Part IV of the survey begins with an introduction to ship accidents, which are one cause of
physical injury to coral reefs. Table 2.9 summarizes whether respondents have heard about, read
about, or seen where ship accidents have injured coral reefs. Tests for significance show that
there is a significant difference in respondents’ WTP increased taxes for increasing no-fishing
zones around the MHI when compared to WTP to protect against ship accidents.
Respondents were then asked how they feel about increasing federal taxes to restore coral reefs
injured by ships around the MHI. Table 2.10 summarizes their responses to this question.
2-6
Table 2.9. Summary of whether respondents have heard about, read about, or seen where
ship accidents have injured coral reefs in Hawaii or elsewhere
Response
Yes
No
Don’t know
Percent respondents
22.3%
64.7%
13.0%
Table 2.10. How respondents feel about increasing federal taxes to restore coral reefs
injured by ships around the Main Hawaiian Islands
Response
Strongly oppose
Somewhat oppose
Neither oppose nor support
Somewhat support
Strongly support
Percent respondents
14.9%
18.1%
27.0%
31.2%
8.8%
Part V: Choice questions
The final section of the survey included a series of choice questions asking respondents to rank
alternatives. In the jargon, each alternative is defined by a “bundles of attributes.” We used four
attributes: the percentage of coral reef ecosystems in the MHI protected as no-fishing zones, the
percentage of coral reef ecosystems in the NWHI protected as no-fishing zones, whether ship
injuries to coral reefs are repaired, and the cost in higher taxes. Each attribute, in turn, could take
on two levels: the status quo (i.e., 1% of coral reef ecosystems protected for the MHI, 5% of
coral reef ecosystems protected for the NWHI, no new program to repair ship injuries, and zero
cost) and changes from the status quo (i.e., 25% of coral reef ecosystems protected for the MHI,
100% protected for the NWHI, a new program to repair ship injuries, and a positive dollar
amount). Each alternative consisted of some specified combination of these attributes. For
example, an alternative might consist of 1% protection for the MHI (the status quo), 100%
protection for the NWHI (a change from the status quo), no repair of coral reefs injured by ships
(the status quo), and an annual cost of $40.
Each choice question involved three alternatives. Alternative A was always the status quo for all
attributes: no new no-fishing zones in the MHI or the NWHI, no additional efforts to restore
damages from vessel groundings, and no additional taxes. Alternatives B and C posed some
combination of management actions beyond the status quo and some increase in taxes. For each
choice question, respondents were asked to identify their most preferred and their least preferred
alternative. In this way, a complete ranking of alternatives for each of the choice questions could
be identified. Six versions of the survey were developed, each with a different combination of
choices in each of the three choice set questions. Table 2.11 illustrates how this worked by
summarizing the first choice question in each of the six versions and the frequency with which
alternatives were chosen as most and least preferred. In each of the six versions, later choice
questions varied the attributes in each alternative and the cost. Each version was designed to
2-7
Table 2.11. Selection of alternative protection programs, by version, for Choice Set 1
Main Hawaiian Northwest Hawaiian
Islands: % reef
Islands: % reef
protected
protected
Version 1
Alternative A
Alternative B
Alternative C
Version 2
Alternative A
Alternative B
Alternative C
Version 3
Alternative A
Alternative B
Alternative C
Version 4
Alternative A
Alternative B
Alternative C
Version 5
Alternative A
Alternative B
Alternative C
Version 6
Alternative A
Alternative B
Alternative C
Reefs
repaired
from ship
injuries
Added taxes to
your household
each year
Percent
preferring this
alternative
the most
Percent
preferring this
alternative
the least
1%
25%
25%
5%
100%
5%
No
Yes
No
$0
$80
$30
18%
61%
21%
53%
29%
18%
1%
25%
1%
5%
100%
100%
No
Yes
No
$0
$40
$15
12%
62%
26%
72%
19%
9%
1%
25%
25%
5%
100%
100%
No
Yes
No
$0
$101
$100
28%
56%
17%
72%
22%
6%
1%
25%
1%
5%
100%
100%
No
Yes
Yes
$0
$40
$20
30%
51%
19%
54%
43%
3%
1%
25%
1%
5%
100%
100%
No
Yes
No
$0
$60
$20
16%
63%
21%
78%
19%
3%
1%
25%
25%
5%
100%
5%
No
Yes
Yes
$0
$70
$40
27%
36%
36%
63%
34%
3%
2-8
avoid inconsistencies across the choice question. For example, within each version, alternatives
that would do less always had a lower cost. The setup and responses to the second and third
choice questions are reported below.
With the exception of respondents who received version 6 of choice set 1, a majority of
respondents chose Alternative B as their most preferred alternative. In all versions, a majority of
respondents chose Alternative A, the status quo, as their least preferred alternative. This points
towards the conclusion that the dollar values for Alternatives B and C used in the pretest were
likely too low, a conclusion that received further support in responses to the other choice sets
across the various survey versions.
The survey includes an open-ended question following the first choice question to further probe
why respondents chose a particular level of protection for Hawaiian reefs as their most
preferred.3
To analyze the open-ended responses, we developed descriptive categories to group each of the
open-ended responses. Categories were developed by first looking at the full set of open-ended
responses, and then by categorizing them. Responses could fall into multiple categories.
Table 2.12 shows the final categories and provides a rationale for selecting each category.
Table 2.12. Description of open-ended response categories for the choice question
Category
1. It’s the right thing to do/We have
an obligation to do something.
Rationale for choosing the category
Respondent’s comments fell into this category if they specifically
discussed the need to protect coral reefs as a responsibility for each
citizen of the U.S., OR if they discussed the importance of protecting
reefs because it is the right thing to do.
2. We need to protect reefs for future Respondents felt that we NEED to protect coral reefs for future
generations.
generations to enjoy and use.
3. There is a trade-off between long- Respondents wrote about the need to pay for protection now in order to
term and short-term costs for
avoid further damage in the future. They also commented on the
protection.
difference in price for protecting reefs now and in the future.
4. Does not want more taxes/Thinks Respondents simply did not want to pay more taxes. Some felt that the
we should consider other alternatives tax was too high and that we should consider other alternatives first.
Others felt that money would be better spent elsewhere.
first/The cost is too high/Money
should be spent on other things.
5. Something needs to be done, but
This category of responses came from individuals who agreed that
not too much.
something needed to be done to help protect the reefs. They felt that
doing something was preferred to doing nothing.
6. Find a compromise between cost These individuals rationalized their responses by weighing the cost of
and protection.
protection against the benefits of protection.
3. The follow up to the choice question read, “Please provide a brief comment that helps us understand why
you chose the alternative as most preferred and as least preferred in the previous question.”
2-9
Table 2.12. Description of open-ended response categories for the choice question
Category
7. We need to choose the alternative
with the best benefit for the reef
ecosystem and the individuals that
depend on them.
8. Protest/Does not believe the
program could work.
9. Other
Rationale for choosing the category
Respondents felt that the reef ecosystem is important enough for
citizens to pay more in taxes to protect them.
These respondents were not convinced that any of the alternatives solve
the reef problem or thought that protection should not cost as much.
Responses that could not be combined into a specific category.
Across all six versions, the first choice question had full implementation of all the management
actions (i.e., 25% protection for the MHI, 100% protection for the NWHI, implementation of a
program to repair injuries from ship accidents). Alternative C returned at least one of these
attributes to the status quo and had a low dollar price, which made it easy to match respondents’
most preferred choice with their responses to the open-ended question about why they made that
choice.
About 75% of respondents chose both a most and least preferred alternative in answering the
first choice question. Most of these respondents chose B as their most preferred option and A,
status quo, as their least preferred option (56.4%). Table 2.13 reports the frequency of responses
by category and choice pattern.
Table 2.13. Frequency of responses for respondents’ most preferred and least preferred
alternatives, by category
Category
1. It’s the right thing to do/We have an
obligation to do something.
2. We need to protect reefs for future
generations.
3. There is a trade-off between longterm and short-term costs for
protection.
4. Does not want more taxes/Thinks
we should consider other alternatives
first/The cost is too high/Money should
be spent on other things.
5. Something needs to be done, but not
too much.
6. Find a compromise between cost
and protection.
7. The best benefit for the reef
ecosystem and the individuals that
depend on them.
A most
A most
B most
B most
C most
C most
preferred, preferred, preferred, preferred, preferred, preferred,
B least
C least
A least
C least
A least
B least
preferred preferred preferred preferred preferred preferred
0
0
30
1
0
1
0
0
12
0
0
0
0
0
9
0
1
0
22
1
0
0
1
3
0
0
0
0
13
2
0
0
9
0
5
7
0
0
23
3
0
0
2-10
Table 2.13. Frequency of responses for respondents’ most preferred and least preferred
alternatives, by category
Category
8. Protest/Do not believe the program
could work.
9. Other.
Total
A most
A most
B most
B most
C most
C most
preferred, preferred, preferred, preferred, preferred, preferred,
B least
C least
A least
C least
A least
B least
preferred preferred preferred preferred preferred preferred
2
3
27
1
2
4
0
10
93
1
0
5
0
0
20
1
2
16
Below we discuss the verbatim responses according to respondents’ most preferred and least
preferred management option:
Alternative A most preferred/Alternative B least preferred: These respondents tended to fall
into Category 4. That is, they tended to object to taxes or feel that costs were excessive, or other
such lines of reasoning. Of all the responses falling into Category 4, 81.5% chose A as most
preferred and B as least preferred.
Alternative A most preferred/Alternative C least preferred: Only four respondents
fell into this group.
Alternative B most preferred/Alternative A least preferred: Ninety-four respondents chose
Alternative B as their most preferred option and Alternative A as their least preferred option. They
represent 93.8% of responses in Category 1 (30 out of 32), 100% of responses in Category 2
(12 out of 12), 90% of responses in Category 3 (9 out of 10), 43% of Category 6 responses (9 out
of 21), 88% of responses in Category 7 (23 out of 26), and 59% of responses in Category 9 (10 out
of 17).
Alternative B most preferred/Alternative C least preferred: Only 3% (n = 5) fell into this
group. Answers to the open ended responses tended to focus on justifying B as most preferred and
do not provide insights on why doing nothing was preferred to alternatives that did not do as much
as Alternative B.
Alternative C most preferred/Alternative A least preferred: A total of 19 respondents (12.1%)
fell into this group. In their open ended responses, respondents in this group tended to fall into
Categories 5 and 6.
Alternative C most preferred/Alternative B least preferred: A total of 16 respondents (9.7%)
fell into this group. Their open ended responses were spread out among the categories, with some
tendency to fall into Categories 6 and 8.
Tables 2.14 and 2.15 summarize respondents’ rankings to the alternatives made in response to
the second and third choice set questions, respectively.
2-11
Table 2.14. Selection of alternative protection programs, by version, for Choice Set 2
Main Hawaiian
Northwest
Reefs
Islands: % reef Hawaiian Islands: repaired from
protected
% reef protected
ship injuries
Version 1
Alternative A
Alternative B
Alternative C
Version 2
Alternative A
Alternative B
Alternative C
Version 3
Alternative A
Alternative B
Alternative C
Version 4
Alternative A
Alternative B
Alternative C
Version 5
Alternative A
Alternative B
Alternative C
Version 6
Alternative A
Alternative B
Alternative C
Added taxes to
your household
each year
Percent
preferring this
alternative
the most
Percent
preferring this
alternative
the least
1%
1%
1%
5%
100%
100%
No
Yes
No
$0
$50
$40
21%
62%
18%
67%
24%
9%
1%
1%
1%
5%
5%
100%
No
Yes
No
$0
$5
$15
9%
32%
59%
79%
9%
12%
1%
1%
25%
5%
5%
5%
No
Yes
No
$0
$1
$50
16%
24%
59%
69%
3%
28%
1%
1%
25%
5%
5%
100%
No
Yes
No
$0
$5
$30
30%
11%
59%
62%
11%
27%
1%
1%
1%
5%
5%
100%
No
Yes
No
$0
$10
$20
14%
22%
65%
79%
5%
16%
1%
1%
25%
5%
100%
5%
No
Yes
Yes
$0
$30
$60
21%
55%
24%
67%
3%
30%
2-12
Table 2.15. Selection of alternative protection programs, by version, for Choice Set 3
Main Hawaiian
Northwest
Reefs repaired
Islands: % reef Hawaiian Islands:
from
protected
% reef protected
ship injuries
Version 1
Alternative A
Alternative B
Alternative C
Version 2
Alternative A
Alternative B
Alternative C
Version 3
Alternative A
Alternative B
Alternative C
Version 4
Alternative A
Alternative B
Alternative C
Version 5
Alternative A
Alternative B
Alternative C
Version 6
Alternative A
Alternative B
Alternative C
Added taxes to
your household
each year
Percent
preferring this
alternative
the most
Percent
preferring this
alternative
the least
1%
25%
1%
5%
100%
100%
No
No
No
$0
$70
$40
26%
59%
15%
63%
28%
9%
1%
25%
25%
5%
100%
5%
No
No
Yes
$0
$35
$25
15%
56%
29%
68%
24%
9%
1%
25%
1%
5%
5%
100%
No
No
Yes
$0
$50
$51
22%
5%
73%
73%
14%
14%
1%
25%
25%
5%
5%
100%
No
No
Yes
$0
$15
$85
31%
36%
33%
51%
8%
41%
1%
25%
1%
5%
5%
100%
No
No
No
$0
$30
$80
18%
34%
47%
76%
5%
18%
1%
25%
25%
5%
100%
5%
No
No
Yes
$0
$50
$40
24%
39%
36%
64%
33%
3%
2-13
Econometric modeling of responses to the choice questions
Using the data from the pretest, we estimated a rank order logit model, regressing the probability
of selecting a specific alternative on the levels of protection from overfishing around the MHI
and NWHI, whether ship injuries would be repaired, and the bid amount. Table 2.16 summarizes
these results.4 All coefficients have the expected sign and are highly significant. A strong
preference for addressing problems of overfishing in the MHI is apparent from the results.
Support is lowest for repairing ship injuries, with no-fishing areas for the NWHI receiving
somewhat more support.
Table 2.16. Rank order logit model results
Main Islands
Northwest Islands
Ship Injuries
Household Cost
Parameter
estimate
0.0975
0.0191
0.0148
-0.0314
Standard
Error
0.0040
0.0014
0.0017
0.0030
Standard
Deviation
0.1780
0.0598
0.0745
0.1333
Z
24.15
14.09
8.73
-10.39
P>|z|
< 0.001
< 0.001
< 0.001
< 0.001
Despite these encouraging results, any attempt to use them to estimate dollar values could be
quite misleading. The problem is that large numbers of respondents chose alternatives with dollar
costs toward the high end. This indicates that the distributions of probabilities coming out of the
model are likely to suffer from “fat tails” on the right side that are not supported by data. Some
higher dollar costs will be used in the final study to remedy this problem.
Attitudes about choice questions
We asked respondents six follow-up questions regarding their attitudes about the choice
questions. Table 2.17 summarizes responses to these attitudinal questions.
Table 2.18 summarizes significant correlations between responses to these attitudinal questions
and answers to other questions in the survey. All relationships reported here are significantly
correlated, at a 10% level or better. The correlations reveal consistent attitudes across questions;
for example, those who believed that cost should not be a factor when protecting the
environment supported paying higher taxes to restore reefs elsewhere in the survey.
4. In this model, no-fishing zones are entered in continuous terms expressed as the percentage of total coral
reefs protected. As a check on results, we also ran an effects coded model, rather than a continuous model, and
the results were the same. The effects coded model coefficients are the marginal values of moving from one
level to the next — which are the same as the coefficient reported above multiplied by the unit changes.
2-14
Table 2.17. Summary statistics for responses to the six attitudinal follow-up questions
Attitudinal statement
Costs should not be a factor when
protecting the environment
I found it difficult to select my most
preferred alternative
There was not enough information for
me to make informed decisions about
doing more to protect coral reefs in
Hawaii
I am concerned that the federal
government cannot effectively manage
coral reefs
I should not have to pay more federal
taxes to protect coral reefs around
Hawaii
The public’s views as expressed in this
survey should be important to the
government when it chooses how to
manage coral reefs in Hawaii
NOBS
Mean
Median
StdDev
213
2.77
3
0.08
214
3
3
0.09
213
3.56
3
0.07
213
2.58
3
0.07
214
2.94
3
0.09
210
1.97
2
0.08
Scale
1 = strongly agree
2 = somewhat agree
3 = indifferent
4 = somewhat disagree
5 = strongly disagree
Those who agreed with the statement “Costs should not be a factor when protecting the
environment” also tended to support increasing federal taxes to expand no-fishing zones around
the MHI and NWHI and restore coral reefs injured by ships around the MHI. They tended to
disagree with the statement that they should not have to pay more federal taxes to protect coral
reefs and to agree that the views of the public as expressed in the survey should be important in
making decisions about Hawaii’s reefs.
Those who found it difficult to select their most preferred alternative tended to oppose increasing
taxes to restore coral reefs around the MHI. They also believed that there was insufficient
information to make informed decisions and did not want to pay more taxes to protect reefs.
If respondents did not believe there was sufficient information to make informed decisions, they
tended to be less likely to expect to visit Hawaii in the next 10 years and opposed using federal
taxes to restore reefs injured by ships. They also tended to agree that the federal government
cannot effectively manage reefs and that they should not have to pay more federal taxes to
protect the reefs.
Respondents who were concerned that the federal government cannot effectively manage coral
reefs had a higher tendency to have visited coral reefs in the United States.
2-15
Table 2.18. Summary of significant correlations between attitudinal and other variables
Attitudinal statement
Q19a. Costs should not be a factor
when protecting the environment
Questions with significant
correlations
Q10a (-)a
Q11a (-)
Q14a (-)
Q19e (-)
Q19f (+)
Q19b. I found it difficult to select
my most preferred alternative
Q14a (+)
Q19c (+)
Q19e (+)
Q19c. There was not enough
information for me to make
informed decisions about doing
more to protect coral reefs in
Hawaii
Q6 (+)
Q14a (+)
Q19_4 (+)
Q19_5 (+)
Q19d. I was concerned that the
federal government cannot
effectively manage coral reefs
Q19e. I should not have to pay
more federal taxes to protect coral
reefs around Hawaii
Q2 (-)
Q2 (+)
Q6 (+)
Q10a (+)
Q11a (+)
Q14a (+)
Respondents who agreed with this
statement also tended to:
Support higher taxes to expand no-fishing
zones and restore reefs from ship strikes
Disagree with NOT paying more to
protect coral reefs around Hawaii
Agree the public’s opinions should be
considered in management decisions
Oppose increasing taxes to restore reefs
after ship injuries
Believe there was not enough information
for informed decisions
Not want to pay higher taxes to protect
coral reefs around Hawaii
Less likely to visit Hawaii in the next 10
years
Oppose increasing taxes to restore reefs
after ship injuries
Concerned about the government’s ability
to effectively manage reefs
Not want to pay higher taxes to protect
coral reefs in Hawaii
Never have visited a coral reef in the US
Never have visited a coral reef in the US
Not expect to visit Hawaii in the next 10
years
Oppose paying higher federal taxes to
increase no-fishing zones and restore reefs
Not expect to visit Hawaii in the next 10
years
Q19f. The public’s views as
Q6 (-)
expressed in this survey should be
important to the government when
it chooses how to manage coral
reefs in Hawaii
The scales for Q10a, Q11a, and Q14a ran from 1 for “strongly disagree” to 5 for “strongly agree.” The scales
in Q19 ran in the opposite direction. Hence, the expected signs or the correlation are negative.
Respondents who agreed that they should not have to pay more federal taxes to protect coral
reefs around Hawaii were less likely to have visited a reef in the U.S., less likely to expect to
visit Hawaii in the next 10 years, opposed to raising federal taxes to expand no-fishing zones
around the MHI and NWHI, and opposed to raising federal taxes to restore reefs injured by ships
around the MHI.
2-16
Finally, respondents who agreed that the public’s views should be considered by the government
when managing coral reefs were more likely to expect to visit Hawaii in the next 10 years.
Influence of news media
As the study was nearing the pretest stage, President Bush announced that a large area in the
NWHI would become a national monument. This meant that no commercial fishing would be
allowed, a fact that conflicted with what we told respondents about the possible need for further
protection from overfishing in the NWHI. This had the potential of undermining the credibility
of the choice questions in the eyes of respondents and we asked the following question to
evaluate this risk: Have you heard or read anything about the Northwestern Hawaiian Islands in
the past year?
Only nine respondents (4.2%) answered this question. Four of the nine participants who
answered were aware that the NWHI had some type of protection (e.g., a sanctuary, monument,
or national park).
Examples of their responses include the following.5
“That the government created a large no fishing sanctary North of Hawaii.”
“Just that a national park was being created there.”
“Pres. Bush signed an evironmental bill that would protect the islands.”
Only one of these participants knew that the U.S. President designated the NWHI as a national
monument. The other five responses varied. Some participants learned about the islands from
watching television shows like PBS. The remaining responses represent how participants feel
about the threat to the islands − the MHI and NWHI − as well as general knowledge about the
number of islands in Hawaii.
“Now is the time to protect the Northwestern Hawaiian Islands, not wait until they are
compromisd.”
“Just that they are not in near as much danger as the main islands.”
“I very recently learned that there were 122 islands in Hawaii, and I did not know there
were so many. These must mostly be in the Northwestern Islands.”
Hence, we concluded that the national monument announcement would not do much to
undermine the validity of our study as it was designed.
5. All responses are presented verbatim as we received them. The respondents entered this information
themselves.
2-17
Closing comments from respondents
At the end of the survey, we asked respondents if they had any additional comments to help us
understand their views about coral reefs in Hawaii and their responses to the survey. A question
like this at the end of a survey allows respondents to have one final opportunity to tell us any
idea or concerns that might otherwise have been unstated, or to restate a previous opinion. This
question also may provide clues about whether participants’ responses to the choice questions
are consistent with their views about coral reefs in Hawaii.
To analyze the open-ended responses, we developed descriptive categories to group them. We
developed categories by first looking at the full set of open-ended responses, and then developed
potential categories. Table 2.19 shows the final categories and provides a description of the
category meaning.
Table 2.19. Description of response categories for additional comments
Category
1. Needs more information/never thought
about coral reefs in Hawaii before.
2. Distrust of the federal government.
3. Feel that we need to maintain a balance in
nature.
4. Feel that we have an obligation to protect
reefs/we should protect reefs just like any
other national resource.
5. Does not believe the program could
work/does not care/protest.
6. Other people have a responsibility to
pay/cannot afford more taxes.
7. Other
Rationale for choosing the category
Respondents did not feel that they could make an informed
decision without more information, or respondents had never
thought about the coral reefs in Hawaii before taking the
survey.
Respondents did not trust the federal government to put the
money to good use (i.e., protecting the reefs).
These respondents feel like we need to make greater efforts
to balance the effects of consumption of natural resources by
taking measures to protect the reef ecosystem.
Respondents feel that we need to protect coral reefs,
regardless of the cost we incur.
Respondents do not feel that the program will have its
desired effects on the reef ecosystem, or they do not care
about reefs.
Respondents in this category feel that we need to consider
the source of the problem (e.g., over-fishing, ship strikes)
and tax the people causing the problem rather than creating a
federal tax, or that the people visiting Hawaii should pay for
the problem, or that they could not pay more in taxes.
Responses that we could not combine into a specific
category.
Only 23% (50 of 216) of the respondents provided comments for this question. Table 2.20
presents the frequency of responses for each category.
2-18
Table 2.20. Any closing comments respondents had about the survey
Category
1. Needs more information/never thought about coral reefs in Hawaii
before.
2. Distrust of the federal government.
3. Feel that we need to maintain a balance in nature.
4. Feel that we have an obligation to protect reefs/we should protect
reefs just like any other national resource.
5. Does not believe the program could work/does not care/protest.
6. Other people have a responsibility to pay/cannot afford more taxes.
7. Other.
Number of
responses
Frequency of
responses
5
5
6
10%
10%
12%
13
3
7
11
26%
6%
14%
22%
Conclusions
Overall, the pretest supported the soundness of the survey instrument and supported the use of
internet mode administration. Comparisons with the GSS national survey showed that
respondents to the Coral Reef survey were for the most part very similar to the national GSS
respondents in terms of attitudes toward social policy issues. Answers to the choice questions
were sufficiently coherent to support simple econometric models with significant coefficients of
the expected signs on the various attributes despite the relatively small sample size. Comparing
verbatim responses to the open-ended questions with responses to the first choice question
showed a few anomalies, but not many. Our conclusion is that the instrument is ready to be
finalized and that finalizing it should be straightforward.
The most surprising result was the lack of correspondence between attitudes toward government
spending of our respondents and those of the national sample from the GSS, but this does not
appear to have any implications for the soundness of our instrument.
2-19
| File Type | application/pdf |
| File Title | C:\PRA\OMB83I pre-ps.WP6.wpd |
| Author | rroberts |
| File Modified | 2009-03-12 |
| File Created | 2009-03-12 |