Marine Policy 117 (2020) 103933

Contents lists available at ScienceDirect

Marine Policy
journal homepage: http://www.elsevier.com/locate/marpol

Estimating the damage cost of plastic waste in Galapagos Islands: A

contingent valuation approach
Manuel A. Zambrano-Monserrate a, *, Maria Alejandra Ruano b
a
Universidad Espíritu Santo, Ecuador
b
Escuela Superior Polit�ecnica del Litoral, ESPOL, Facultad de Ciencias Sociales y Humanísticas, Campus Gustavo Galindo Km 30.5 Vía Perimetral, P.O. Box 09-01-
5863, Guayaquil, Ecuador

A R T I C L E I N F O A B S T R A C T

Keywords: Plastic waste affects various ecosystems in the world. The oceanic islands are one of the most affected places by
Galapagos islands this type of material. A significant case of plastic pollution occurs in the Galapagos Islands. The Galapagos
Plastic pollution Islands, located west of Ecuador, have one of the most biodiverse ecosystems on the planet. It is the home of
Economic value
unique species in the world, and since 1978 it has been considered as a Natural World Heritage. However, the
Contingent valuation
Ecuador
Galapagos Islands are threatened by plastic pollution. Therefore, this study aims to quantify the cost of envi­
ronmental damage generated by plastic waste in the Galapagos Islands. The non-market economic valuation
approach was used to implement contingent valuation. It was determined that the Willingness to Pay of Ecua­
dorian families, to reduce plastic pollution, was between US$ 4.90 - US$ 14.51 per year, with a median of US$
7.65. This estimated value can serve as a reference to demand compensation from those who are generating
pollution. At the end of the document, public policies are discussed based on the results obtained.

1. Introduction Patrimony of Humanity by UNESCO in 1978. They are the place of a

unique ecosystem with terrestrial animals, marine life, and plants that
Plastic pollution is a serious environmental problem that affects are not found anywhere else on the planet. The Galapagos habitat
various ecosystems in the world [1]. Pollution from plastic waste is the inspired the evolutionary theory of Charles Darwin and today remains
accumulation of plastic products in the environment that produces one of the most important biological landmarks in the world [6,7].
adverse effects on the habitat of wildlife, or humans [2]. The oceans are Despite its biological importance, some of its islands are polluted
one of the ecosystems most affected by plastic pollution [3]. The plastic with plastic waste. A large amount of this waste does not come from the
waste that ends up in the oceans can persist for decades, which seriously Galapagos Islands productive activities but instead comes from South
compromises the sustainable development of these ecosystems. Also, of America and Central America, in particular northern Peru and southern
all the plastic that enters the oceans, only about 1% can be seen, the Ecuador [8]. Local residents, as well as volunteers, have tried to keep the
other 99% disappears into the depths, which makes the problem less beaches of the Galapagos Islands clean.2 However, there is so much
visible [4]. garbage that accumulates that it has become an impossible task. In the
One of the most striking cases of plastic pollution occurs in the first quarter of 2019 alone, eight tons of garbage was collected compared
Galapagos Islands. The Galapagos Islands are located in Ecuador, to 24.23 tons throughout 2018 and 6.47 tons in 2017 [9].
approximately 1200 km west of mainland Ecuador (Fig. 1). They are Accumulated trash affects most species in the archipelago. For
made up of three cantons1 (Isabela, Santa Cruz, and San Cristobal). Each example, species such as the Mancon Cormorant (Phalacrocorax harrisi)
canton, in turn, is made up of a set of islands and islets. In total, there are build nests with bags and disposable diapers found on the beaches.
13 islands and more than 200 islets, which make up 1667 km of rocky Other species, such as sea turtles, mistake plastic bags with jellyfish and
coast mainly [5]. The Galapagos Islands were declared Natural are sometimes trapped or entangled in these wastes.3 In more severe

* Corresponding author.
E-mail address: [email protected] (M.A. Zambrano-Monserrate).
1
Ecuador is divided into 24 provinces, and the provinces are divided into cantons. There are 221 cantons.
2
Since 1996 artisanal fishermen clean the most remote islands, but a record of the waste has only been kept for three years.
3
Entanglement due to oceanic plastic pollution is a threat to at least 243 marine species, in addition to being the deadliest [54].

https://doi.org/10.1016/j.marpol.2020.103933
Received 29 December 2019; Received in revised form 25 February 2020; Accepted 12 March 2020
Available online 31 March 2020
0308-597X/© 2020 Elsevier Ltd. All rights reserved.
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

Fig. 1. Galapagos Islands Location. Elaboration: The authors.

2
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

cases, animals such as blue-footed booby (Sula nebouxii) consume quality (31%), and dog feces (24%). 87% of British respondents
microplastic,4 which is a deadly chemical element for different species5 expressed a positive willingness to pay. British tourists were willing to
[4]. pay £ 1.03 per adult per visit, which gave an average of £ 0.90 for the
Some scientists mention that the most effective solution is to attack entire sample. Also, Brouwer et al. [18] analyzed the social costs of
the problem at the source; cleaning the beaches is only a partial mea­ marine litter along the coasts of three European countries (Greece,
sure. It is necessary to determine where the garbage comes from and Bulgaria, and the Netherlands). Using the choice experiment method,
thus design appropriate strategies to prevent contamination of the ar­ they found that Bulgarian beach visitors are willing to pay significantly
chipelago [4]. However, before implementing appropriate strategies more than Greek and Dutch visitors for both marine plastics washed
(policies and regulations), it is crucial to estimate the cost of environ­ ashore and cigarette butts left behind by visitors. Their results have
mental damage generated by these wastes in the ocean. The estimation important implications for the size and transferability of the social
of this cost represents an economic reference for environmental resto­ (estimated) costs of marine litter in Europe.
ration or compensation measures. In relation to the restoration and conservation of coastal ecosystems,
The non-market economic valuation (EV) allows these costs to be the work of Park et al. [19] stands out. They evaluated the conservation
quantified through an indicator measured in monetary units [10–12]. value of tidal planes in Shinan, Korea. In general, Korean households
The contingent valuation method (CVM) is one of the most used ap­ were willing to pay a significant amount for the conservation of these
proaches to quantify the costs of environmental damage [13]. While ecosystems. They estimated that the average WTP for the preservation of
many studies have used this methodology to quantify environmental the Shinan Tidal Flat was approximately KRW 3100 (US$ 2.75) per year
damage to air and water, few studies have used this method to estimate per household. They used the CVM. Along the same lines, Tonin [20]
the cost of damage (caused by plastic) to the marine environment [14]. used the CVM to estimate the willingness to pay for biodiversity resto­
The work of Choi and Lee [13] stands out, who estimated the Willing­ ration and conservation scenarios in some unique coral habitats in the
ness to Pay (WTP) of the residents of Seoul (Korea) to remove micro­ North Adriatic Sea, Italy. They found that households are willing to
plastics from the ocean in that locality. They found that families are spend, on average € 35.42 one time only. They emphasize that prior
willing to pay an average of US$ 2.59 annually. knowledge and awareness of marine biodiversity are key determinants
After the literary review, no previous work on the subject is evi­ for the WTP.
denced in the studied location. Therefore, this research aims to estimate
the cost of environmental damage of plastic waste in Galapagos. The 3. Methodology
main results show that families are willing to pay on average between
US$ 4.90 - US$ 14.51 per year, with a median of US$ 7.65. The rest of 3.1. Contingent valuation
the document is structured as follows: literature review is presented in
section 2. The methodology is described in section 3. The results are The Galapagos Islands are a natural resource that lacks an explicit
presented in section 4. Section 5 concludes. market per se. As such, the valuation of a natural resource is different
from the valuation of market goods [21]. Lo and Jim [22] explain that
2. Literature review natural resources, not traded directly in the markets, can be assigned
monetary value by using the non-market Economic Valuation (EV)
As mentioned in the previous section, few studies have used the CVM method. Contingent Valuation (CV) is one of the most popular and used
to quantify environmental damage (caused by plastic) to the marine non-market EV methods [23].
environment. However, other studies have used contingent valuation The CV method analyzes people’s preferences directly for environ­
and other direct and indirect methods to find economic values and mental goods or services [24]. This methodology is generally imple­
promote the conservation of coasts and beaches. In this section, some mented using questionnaires designed following economic principles. In
studies with these characteristics are reviewed. the questionnaire, a hypothetical environmental improvement is pre­
Beharry-Borg and Scarpa [15] economically valued a change in the sented, and implicit compensation is required between personal income
quality of coastal waters in Tobago. Using the choice experiment and environmental quality [22]. People who participate in the ques­
method, they found that the specific individual means of the WTP esti­ tionnaire are asked to indicate their maximum WTP to avoid negative
mates vary significantly between two heterogeneous groups: those who environmental change (In this particular case, the deterioration of the
practice snorkeling and those who do not. The authors stress the Galapagos Islands due to plastic pollution).
importance of including individual preferences in the context of a
developing country where it is necessary to prioritize policy recom­ 3.2. Focus groups, pilot surveys and questionnaire design
mendations. This is due to limited financial resources and conflicting
objectives for the management of natural resources. Along the same Previous data collection, pilot surveys, and focus groups were con­
lines, Loomis and Santiago [16] evaluated the improvement of water ducted in order to assess the questionnaire, following the recommen­
quality in four beaches of Puerto Rico. Using the CVM and choice dations by Kanninen [25] and Haab and McConnell [26]. The focus
experiment method, they found that the average WTP was US$ 54 and groups were made up of 12 people over 18 years old who were interested
US$ 51 per visit, respectively. They concluded that there are no signif­ in the problem of plastic pollution in the Galapagos Islands. The selected
icant differences in WTP estimates between the two methods. people were chosen according to the “snowball6” non-probabilistic
Blakemore and Williams [17] used the CVM and travel cost method sampling method. Additionally, 30 pilot surveys were conducted
to value the beach of Olu Deniz, Turkey, economically. According their face-to-face with open-ended questions to people over 18. Both focus
results, the majority of beach users (70%) were British, who used the groups and pilot surveys were conducted in three cities: Quito, Guaya­
beach for recreational activities and to enjoy open spaces and land­ quil, and Cuenca. At this stage, the hypothetical improvement scenario
scapes. They found that the main dislikes are garbage (41%), water was validated, and the frequency and schemes of the bids were defined.
The final questionnaire consisted of four parts. The first part evalu­
ated the relationship of people with the environment and natural
4
Saltwater, the heat of the sun, and time degrade plastics forming
microplastics.
5 6
There are documented encounters of 693 wildlife species with marine First, a group of people was randomly selected to be interviewed. A person
debris, impacted through ingestion, entanglement, transportation and habitat interested in environmental problems was chosen from this group. In turn, he
alteration [54]. was asked to refer other people with the same characteristics.

3
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

resources. The second part first assessed and measured the knowledge 3.4. Bias control and payment vehicle
and importance of the population towards the problem of plastic
pollution in the Galapagos Islands. Subsequently, in this same section, CV studies highlight the importance of controlling the different
the respondent was informed about the problem of plastic pollution in biases that originate due to the methodology. As such, to control the
the Galapagos Islands. In the third part, the hypothetical improvement hypothetical bias,9 the valuation question was created to make the re­
scenario and the various strategies to mitigate the different biases were spondents feel like they were paying the agreed money. Even more, the
raised and evaluated. In the last section, questions were asked about the survey reminded them that they had limited income and had to consider
socioeconomic characteristics of the population.7 their opportunity cost. Several studies back up that these types of stra­
tegies reduce the hypothetical bias [24,27,33].
3.3. Sampling and measurement of WTP On the other hand, to control the strategic bias10, the survey speci­
fied that if more than 50% of respondents agree to pay for the
This study reduced the sampling area to three cities in Ecuador: improvement, the policy would be implemented, and citizens would pay
Quito, Guayaquil, and Cuenca. These cities are considered the most for it. While, if more than 50% of respondents do not agree, the program
important in Ecuador, in addition to being the most populated. An would not continue. This procedure suggests to respondents that their
expert market research company conducted the survey. The sampling answers would affect the implementation of the policy, reducing the
technique that was used was the Proportionate Stratified Sampling. This presence of excessively large or small WTP values.
technique, belonging to the probabilistic sampling family, consists of Respondents who answered “no” to the general proposal were asked
dividing the entire population under study into different subgroups or a follow-up question about their main reason for not wanting to
strata so that an individual (families in this case) can only belong to a contribute to the program. The response options were: a) the competent
stratum. The stratum variable considered in this case was the cities. The authorities are already spending enough to fight the problem of plastic
sample size of each stratum was proportional to the population size of pollution in the Galapagos Islands, b) I do not trust that the money raised
the stratum. Thus, the optimal sample size for Guayaquil was 445 ob­ will be managed appropriately, c) It is not my job to finance this type of
servations, 451 for Quito, and 104 for Cuenca. The company randomly projects, d) My current economic situation does not allow me to pay, e)
selected a family from each stratum. Heads of households or spouses Plastic pollution is not a problem in the Galapagos Islands, f) There are
capable of making financial decisions and paying taxes were more important environmental problems than plastic pollution in the
interviewed. Galapagos Islands. The responses to the first three options were
This research used face-to-face interviews,8 as it has proven to be the considered as protest responses, while the other three options are true
most reliable approach in CV studies [27]. One of the points in favor is zeros [34,35]. Labao et al. [34] mention that reasons other than the
that interviewers can clearly explain the hypothetical market to re­ assumption that the good does not have value for the respondent and
spondents; therefore, they may have enough information to discern if money restrictions should be considered protest responses.
they are willing to pay for the proposed improvement scenario [24]. According to the results from the focus groups, the value for the
The sample was collected as follows: the interviewers visited the annual payment was chosen. Participants in the focus group said they
households in the sample and asked if they would participate or not in preferred annual fees since it was easier to picture it as a real situation
the survey. If they agreed, the interviewers started the survey. Other­ than other types of payment that could turn out as more complex or
wise, they were excluded from the sample, and the company extracted unrealistic. Egan et al. [36] discuss the advantages of using annual fees
new stratified homes from the sample. This procedure was repeated until over different types. The payment vehicle chosen in our study was an
the number of optimal observations for each city was obtained. The income tax. This type of tax is very realistic and fits very well with the
surveys were conducted between October 20, 2019, and November 22, annual payments that family heads can afford [13]. Also, it is used as a
2019, and follow up calls were placed to randomly selected respondents financial source for government policy, and it is closely related to the
to gauge the interviewers’ performance. policy of reducing plastics in the oceans and beaches.
The valuation question was raised as a consultative referendum.
Respondents answered a question about their maximum WTP for a 4. Results
public (hypothetical) program that reduces plastic pollution in Gal­
apagos. In this study, the dichotomous choice format, presented by The most important results of this research are described below.
Bishop and Heberlein [28]; was used. It was recommended for the first First, some descriptive statistics on the relationship of people with the
time by the NOAA Panel to evaluate the credibility and reliability of the environment and natural resources are shown, particularly towards
CV method [29]. This approach is compatible with incentives, avoids the plastic pollution in the Galapagos Islands. Acceptance rates are also
lack of response and outliers, and represents a lower cognitive burden analyzed at each bid level. Subsequently, empirical estimates are shown
for the respondent [30]. Moreover, according to Borzykowski et al. [31]; using different parametric models. Finally, the WTP of people to reduce
this approach perfectly mimics a market situation, so it is relatively easy plastic pollution in the Galapagos Islands is estimated.
to understand.
Initially, respondents were asked if they were willing to pay for a
public (hypothetical) program that reduces plastic pollution in the
Galapagos Islands. If the respondents answered “yes,” then they were
given a bid where the respondents had to answer “yes” or “no” (in order
to accept or decline) the bid. Six different bids were established based on 9
The situation in which the interviewee could declare an exaggeratedly high
the results of the focus groups and the pre-test surveys: 0.25, 0.50, 1, 3,
WTP because the scenario of improvement is hypothetical [55].
5, and 10 US$. Each respondent was presented with a single randomly 10
Respondents may think that if the improvement is carried out, it will be
assigned bid from one of these six payment levels. Many studies, financed by the beneficiaries, according to the WTP declared in the surveys. If
including that of Zografakis et al. [32]; allow this random assignment. so, the interviewee will likely choose the least credible amount as a response.
On the other hand, the interviewee may be convinced that if the implementa­
tion of the improvement program is carried out, it will be independent of their
7
The complete questionnaire can be found as supplementary material, which response. As [56] points out, “it is interesting for the person, from a selfish point of
was translated from Spanish. view, to give false signals, to pretend to have a lower interest than what is really had
8
In the CV methodology, the term “face-to-face interviews” is used to refer to in a certain collective activity,” to offer, in short, a strategic response, not an
the process of people responding to surveys. honest one.

4
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

4.1. Descriptive statistics of negative WTP. However, it does not solve the problem of infinite WTP
[31]. Also, mixed models like Spike Log-normal could be used. The Spike
The majority of respondents (71%) said that natural resources are model proposed by Kristro €m [41]; assumes that the data follow a
very important. However, only a small percentage (18%) of the total two-part distribution, with a mass of probability at a price of zero, fol­
respondents said they participated as volunteers in an environmental lowed by a continuous probability distribution at nonzero prices [42].
organization. Also, only 29% of those interviewed said they had The estimates of the three models described above are shown in
collaborated financially with an environmental organization in the last Table 1. The main independent variable is the bid value that was pre­
12 months. Attendance at training/talks related to the care of the sented to the respondents. Additional control variables were included
environment natural resources was also low. Only 33% of the heads of (see details in the Appendix). The inclusion of control variables should
household interviewed said they had attended such events. not alter the mean of the WTP since the latter is evaluated on the average
Regarding the problem of plastic pollution in the Galapagos Islands, of these variables. However, it allows controlling the heterogeneous
72% of respondents revealed that they did not know that the islands characteristics of the population [31]. In all models, the sign of the
have this problem. However, 65% of household heads consider the issue variable “bid” is negative and statistically significant. This result in­
of plastic pollution in the area as very important. Fig. 2 dicates that the greater the bid, the lower the probability of people
Fig. 3 shows the relationship between different levels of bids and paying. This result is consistent with the findings of Ferreira and Mar­
acceptance rates. For example, when the bid was US$ 0.25, the average ques [43] in Portugal, Jianjun et al. [44] in China, and Khong et al. [45]
acceptance rate was 70.5%. In contrast, when the bid increased to US$ in Vietnam.
10, the bid rate was reduced to 37.4%. In general, the higher the bid, the People’s relationships with natural resources influence their WTP.
lower the acceptance rate. On the other hand, the ratio of negative re­ Respondents, who said that natural resources are very important or
sponses is shown in square brackets. The relationship with the bid is important, are more likely to pay than those who consider this type of
positive. These findings are consistent with economic theory and show resources as not important. This finding is consistent with the results of
that the proposed program has a good level of acceptance. Tonin [20]; who emphasizes that the awareness/importance that people
give to natural resources is a key factor in explaining their WTP. Also,
those people who collaborated economically with an environmentalist
4.2. Empirical analysis
organization in the last 12 months are more likely to contribute mone­
tarily than those who did not. Likewise, respondents who attended
Most CV studies often eliminate protest responses from the analysis.
training/talks related to the care of the environment and/or natural
However, the elimination of protest responses is only justified if the
resources during the last year are more likely to pay than those who did
group of protesters is not significantly different from the rest of the
not attend. Finally, those people who said that the problem of plastic
sample [37,38] and if these observations are a random part of the
pollution in the Galapagos Islands is a very important/important issue
sample that can be excluded without affecting the results [39]. If the
are more likely to pay than those who consider the issue as “not
above is not fulfilled, protest bidders are likely to declare a WTP value
important."
different from the rest of the sample. Therefore, eliminating these re­
The effects of sociodemographic and economic variables on WTP are
sponses would be falling into a significant sample selection bias. Ghosh
also analyzed. This study demonstrates that income positively and
et al. [58] recommend estimating a two-step Heckman probit model to
significantly affects WTP. This finding is consistent with the studies by
account for this possible bias. Thus, this procedure was initially imple­
Dribek and Voltaire [35] in Tunisia [44], in [47] in Norway. However, it
mented in our study. The equation of interest is the probit equation of
contradicts the findings of Cazabon-Mannette et al. [46] in Tobago, who
the calibrated dichotomous choice question.11 For the selection equa­
found no evidence that the income affects the WTP of people by pre­
tion, the dependent variable was the general valuation question.12 The
serving sea turtles in that locality. Other variables considered in the
null hypothesis of the likelihood ratio test of independent equations
present study were: “sex,” “age,” and “education.” There is no evidence
(rho ¼ 0) could not be rejected (X2 ¼ 0.33; p-value ¼ 0.6213), indicating
that these variables significantly affect WTP. Studies by other authors
that there is no evidence of selection bias.13
who have used CV have found similar results. For example, the educa­
After checking the above, two simple parametric models were esti­
tion of people was not significant in the Chen and Qi [48] studies in
mated: one logistic (with symmetric distribution) and another Log-
China and Indab [49] in the Philippines. Likewise, age did not affect the
Normal (with asymmetric distribution). Also, a mixed parametric
WTP of those interviewed in the study by Cook et al. [50] in Iceland. Sex
model was estimated: Spike Log-Normal (with asymmetric distribution).
was a non-significant variable in Tonin [20] research in Italy. Finally,
Previous studies have shown that the estimation of WTP (both mean and
there are no significant differences between the families of different
median) is sensitive to the assumption of distribution assumed [40].
cities in relation to their WTP.
The logistic model is the most used out of the simple parametric
At the end of each estimate, some goodness of fit tests are shown. The
models since the estimation of WTP is relatively simple compared to
likelihood ratio (LR) statistic is significant at 1% in the Logistic model
other models. Also, the results estimated by a logistic model usually
and Log-Normal model; this means that the coefficients of the slopes are
coincide with those estimated by non-parametric models [30]. However,
statistically significant jointly to explain the probability that people are
these types of models, being symmetrical, could yield a negative or
willing to pay or not. The same interpretation follows the Wald test for
infinite WTP. Both results from economic theory are not acceptable,
the Spike Log-Normal model. Also, count R2 is presented, which shows
since if an individual does not value a good, his WTP should be zero, and
the percentage of observations that the model correctly classified. For
not negative.14 Besides, the WTP should not be infinite, since this
example, in the Logistic model, 72.1% of the observations were correctly
regularly depends on the income of people, which is finite [30]. A
classified. Additionally, the probability of Pearson’s goodness-of-fit test
parametric model with an asymmetric distribution like the log-normal
is shown. In the Logistic and Log-Normal models, this value is higher
could better fit the true WTP. This type of model solves the problem
than 0.05, so there is a good fit of the estimated models. Finally, the
Akaike (AIC) and Bayesian (BIC) information criteria are presented.
11 That model, with the lowest information criteria, is considered the best
Excluding protest responses.
12 goodness-of-fit. In this case, the Log-Normal model meets that condition.
Including protest responses.
13
The full results of this procedure are not shown to save space but are
available upon request. 4.3. Estimation of willingness to pay (WTP)
14
A negative result could only be accepted if the public program can be
considered as a deterioration. The calculation of the WTP (mean and median) depends on the

5
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

Fig. 2. Statistics on the relationship of people with natural resources and the environment. Elaboration: The authors.

functions of the estimated parameters, procedures such as the delta

method15 are inappropriate since they produce symmetric confidence
intervals [31]. Therefore, non-symmetric confidence intervals were
estimated using Krinsky and Robb simulations, as recommended by Park
et al. (1991), and Haab and McConnell [26]. Table 2 shows that the
mean WTP of the three models evaluated differs significantly, similarly,
the median WTP of asymmetric models.
After considering the information criteria shown in Table 1, the
model with the best fit is the Log-Normal. When the adjusted distribu­
tion is asymmetric (as in the case of Log-Normal), the median, rather
than the mean, is preferred as a measure of central tendency [38].
Therefore, it is concluded that the WTP for reducing plastic pollution in
the Galapagos Islands is between US$ 4.90 - US$ 14.51 per year, with a
median of US$ 7.65.
Fig. 3. Acceptance rates. Elaboration: The authors. Negative WTP ratio be­
tween [ ]. 5. Discussion

Around the world, humans produce about 1.3 billion tons of plastic
assumption of the distribution assumed [40]. For example, the WTP that
arise from the “log” distributions is higher, due to the asymmetry and waste per year, a figure that will increase to 2.2 billion by 2025. In
countries like Ecuador, whose garbage collection services are limited,
the correct bias of these distributions. This result is standard, as recog­
nized in Bengochea-Morancho et al. [51]. Thus, Table 2 shows that the some of these plastic wastes inevitably end up in the oceans or on the
beaches, where they have the potential to affect wildlife and human
mean WTP of the logistic model is much lower than the mean WTP of the
non-symmetric models. On the other hand, it is observed that the health [52].
In recent years, one of the places that have been most affected by
average of the logistic model is similar to the median of the log-normal
model. One possible explanation for this is that the median blocks the plastic pollution is the Galapagos Islands. The Galapagos Islands are one
of the most biodiverse places on the planet, and although 97% of its
asymmetric effect of the log-normal distribution, so its value could
approach the logistics average. The confidence intervals for each model islands are inhabited by humans, plastic waste from all over the world
reaches its shores. They reach the islands at such a high rate that it is
were estimated to determine if these differences are significant.
Because WTP measurements (mean and median) are non-linear almost impossible to keep their coastal areas clean. Many of the endemic
species that inhabit the islands live in piles of garbage, confusing plastic

15
nlcom in Stata.

6
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

Table 1 study aims to determine the WTP of Ecuadorian families to reduce

Models estimation. plastic pollution in the Galapagos Islands. In this sense, the non-market
Logistic Log- Spike Log- EV allows quantifying in monetary terms the cost of environmental
Normal Normal damage generated by this waste on the islands. Thus, the WTP was
Bid 0.3665* – – estimated to be between US$ 4.90 - US$ 14.51 per year, with a median of
Log(Bid) – 0.3389* 0.4635* US$ 7.65. This estimated value can serve as a reference to demand
Importance of natural resources compensation from those who are generating pollution.
Less important 0.4583** 0.1851** This study also analyzed the effect of environmental behavior vari­
Important 0.5138** 0.2073** 0.2415**
Very important 0.7243* 0.2895* 0.3161*
ables (of families) on WTP. For example, it was determined that heads of
Charitable contribution 0.7912* 0.4211* 0.2542** household who consider natural resources very important or important,
Attendance to talks 0.6173* 0.3789* 0.3816* are more likely to pay, to reduce plastic pollution, than those who
Importance of plastic pollution consider these resources not important. This can be explained by the
Less important 0.2528*** 0.2463*** 0.2072***
level of sensitivity of the population towards the subject. In addition,
Important 0.5162* 0.5010* 0.4983*
Very important 0.3392** 0.3592** 0.3174** respondents who claim to contribute financially to an environmental
Sex 0.0212 0.0241 0.0184 organization in the last year are more likely to pay than those who did
Age 0.0128 0.0051 0.0188 not.
Income From the results of this research, some recommendations for public
Between US$ 400 – US$ 900 0.1029 0.1312 0.1510
Between US$ 901 – US$ 1700 0.1895** 0.1681** 0.2133**
policy can be derived. For example, it was determined that respondents
Between US$ 1701 – US$ 2800 0.2183* 0.1868* 0.2572* who attended talks related to the care of the environment/natural re­
Between US$ 2801 – US$ 3500 0.2293* 0.1923* 0.3102* sources are more likely to pay than those who did not. However, only a
Higher than US$ 3500 0.2865* 0.2071* 0.3772* small percentage of the heads of family (33%) said they attended this
Education
type of event. Thus, competent authorities should strengthen awareness
High school 0.0584 0.0312 0.0273
University 0.1021 0.0592 0.0637 programs on the care of the environment/natural resources, and
Postgraduate 0.1283 0.0693 0.0601 encourage the population to attend these programs.
City As mentioned by Monteiro et al. [53]; discussions on the conserva­
Guayaquil 0.0342 0.0382 0.0323 tion of these types of ecosystems must be followed by urgent and effi­
Cuenca 0.0730 0.0821 0.0839
cient management regulations. Palliative actions (such as continuous
Constant 0.4090 0.1523 0.4089
LR chi2 102.05* 104.40* – cleaning of the island), as well as preventive actions, can be combined.
Wald chi2 – – 190.91* On this last point, in the Galapagos Islands, it is prohibited to use plastics
Count R2 (%) 72.1 72.2 – such as straws, t-shirt covers, polyethylene containers, and
Pearson (Prob > chi2) 0.2550 0.1436 –
non-returnable plastic bottles (Government Galapagos, 2018[9]).
AIC 1.097 1.094 1.192
BIC 3547.81 3550.16 3329.01 However, and as previously mentioned, most of the plastic waste in the
Galapagos Islands comes from sources outside its inhabitants. Therefore,
***p < 0.1, **p < 0.05, *p < 0.01. in addition to the policy of banning the use of plastics, there is a need to
develop monitoring systems that determine the exact source of plastic
waste. Only in this way can the problem be attacked efficiently (at the
Table 2
WTP estimation.
source).
This research encourages future studies on the subject, either deep­
Logistic Log-Normal Spike Log-Normal
ening the results of this study or extending research on plastic waste to
Mean WTP 4.96 594.76 29.61 other locations of interest. This research contributes to the literature by
LB 4.28 76.39 18.48 quantifying the environmental damage generated to one of the most
UB 6.05 58503.66 78.45
ASLa 0.000 0.000 0.000
biodiverse places on the planet, such as the Galapagos Islands. In the
Median WTP n.a.b 7.65 19.45 long term, turning this site into a place without plastic pollution would
LB – 4.90 14.89 be a goal of sustainable development.
UB – 14.51 35.73
ASLa 0.000 0.000
–
CRediT authorship contribution statement
Krinsky and Robb (95%) Confidence Interval for WTP measures (Nb of reps:
10,000). Manuel A. Zambrano-Monserrate: Conceptualization, Methodol­
a
Achieved Significance Level for testing H0: WTP � 0 vs. H1: WTP>0LB: ogy, Data curation, Formal analysis, Writing - original draft, Writing -
Lower bound; UB: Upper bound. review & editing. Maria Alejandra Ruano: Writing - original draft,
b
Logistic distributions are symmetrical and hence mean WTP equals the
Writing - review & editing.
median.

Acknowledgements
waste with food, or getting trapped in the rubble [4].
Although numerous studies have analyzed plastic pollution in the All persons who have made substantial contributions to the work
world, few studies have focused on analyzing the effect of plastic reported in the manuscript (e.g., technical help, writing and editing
pollution on oceanic islands [53]. Even less, they have used economic assistance, general support), but who do not meet the criteria for
valuation approaches. Only recent work by Choi and Lee [13] stands authorship, are named in the Acknowledgements and have given us their
out, who estimated the WTP of the residents of Seoul (Korea) to remove written permission to be named. If we have not included an Acknowl­
microplastics from the ocean in that locality. They found that families edgements, then that indicates that we have not received substantial
are willing to pay an average of US$ 2.59 annually. contributions from non-authors.
Considering the global importance of the Galapagos Islands, this

7
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.marpol.2020.103933.

Appendix. Variables considered for the estimation of the WTP

Variable Description Code

Bid Bid proposed to the respondent in US$ 0.25, 0.50, 1, 3, 5, 10.

Importance of natural Rating given by respondents to the natural resources 0 ¼ Not important
resources 1 ¼ Less important
2 ¼ Important
3 ¼ Very important
Charitable contribution If in the last 12 months the respondent has collaborated economically with an environmental organization. 0 ¼ No
1 ¼ Yes
Attendance to talks If in the last 12 months the respondent has attended training/talks related to the care of the environment and/or 0 ¼ No
natural resources. 1 ¼ Yes
Importance of plastic Rating given by respondents to the importance of plastic pollution in the Galapagos Islands. 0 ¼ Not important
pollution 1 ¼ Less important
2 ¼ Important
3 ¼ Very important
Sex Sex of the respondent. 0 ¼ Women
1 ¼ Men
Age Age of the respondent in years –
Income Monthly family income of the respondent. 0 ¼ Lower than US$ 400
1 ¼ Between US$ 400– US$ 900
2 ¼ Between US$ 901– US$
1700
3 ¼ Between US$ 1701– US$
2800
4 ¼ Between US$ 2801– US$
3500
5 ¼ Higher than US$ 3500
Education Highest level of education achieved by the respondent. 0 ¼ Elementary school
1 ¼ High school
2 ¼ University
3 ¼ Postgraduate
City City to which the surveyed family belongs 0 ¼ Quito
1 ¼ Guayaquil
2 ¼ Cuenca

References [15] N. Beharry-Borg, R. Scarpa, Valuing quality changes in Caribbean coastal waters
for heterogeneous beach visitors, Ecol. Econ. 69 (5) (2010) 1124–1139.
[16] J. Loomis, L. Santiago, Economic valuation of beach quality improvements:
[1] B. Worm, H.K. Lotze, I. Jubinville, C. Wilcox, J. Jambeck, Plastic as a persistent
comparing incremental attribute values estimated from two stated preference
marine pollutant, Annu. Rev. Environ. Resour. 42 (2018) 1–26.
valuation methods, Coast. Manag. 41 (1) (2013) 75–86.
[2] M.A. Zambrano-Monserrate, M.A. Ruano, Do you need a bag? Analyzing the
[17] F. Blakemore, A. Williams, British tourists’ valuation of a Turkish beach using
consumption behavior of plastic bags of households in Ecuador, Resour. Conserv.
contingent valuation and travel cost methods, J. Coast Res. (2008) 1469–1480.
Recycl. 152 (2020) 104489.
[18] R. Brouwer, D. Hadzhiyska, C. Ioakeimidis, H. Ouderdorp, The social costs of
[3] E.M. Jepsen, P.N. de Bruyn, Pinniped entanglement in oceanic plastic pollution: a
marine litter along European coasts, Ocean Coast Manag. 138 (2017) 38–49.
global review, Mar. Pollut. Bull. 145 (2019) 295–305.
[19] S.Y. Park, S.H. Yoo, S.J. Kwak, The conservation value of the Shinan Tidal Flat in
[4] Galapagos, 2019. (Retrieved from: https://www.galapagos.gob.ec).
Korea: a contingent valuation study, Int. J. Sustain. Dev. World Ecol. 20 (1) (2013)
[5] H.M. Snell, P.A. Stone, H.L. Snell, A summary of geographical characteristics of the
54–62.
Galapagos Islands, J. Biogeogr. 23 (5) (1996) 619–624.
[20] S. Tonin, Estimating the benefits of restoration and preservation scenarios of
[6] M.K. Tanner, N. Moity, M.T. Costa, J.R.M. Jarrin, O. Aburto-Oropeza, P. Salinas-de-
marine biodiversity: an application of the contingent valuation method, Environ.
Le�on, Mangroves in the Galapagos: ecosystem services and their valuation, Ecol.
Sci. Pol. 100 (2019) 172–182.
Econ. 160 (2019) 12–24.
[21] H. Choi, Y. Koo, Using contingent valuation and numerical methods to determine
[7] G.J. Edgar, S. Banks, R. Bensted-Smith, M. Calvopi~ na, A. Chiriboga, L.E. Garske,
optimal locations for environmental facilities: public arboretums in South Korea,
S. Salazar, Conservation of threatened species in the Galapagos Marine Reserve
Ecol. Econ. 149 (2018) 184–201.
through identification and protection of marine key biodiversity areas, Aquat.
[22] A.Y. Lo, C.Y. Jim, Protest response and willingness to pay for culturally significant
Conserv. Mar. Freshw. Ecosyst. 18 (6) (2008) 955–968.
urban trees: implications for Contingent Valuation Method, Ecol. Econ. 114 (2015)
[8] E. Van Sebille, P. Delandmeter, A.J. Schofield, B.D. Hardesty, J. Jones, A. Donnelly,
58–66.
Basin-scale sources and pathways of microplastic that ends up in the Gal� apagos
[23] W.M. Hanemann, Valuing the environment through contingent valuation, J. Econ.
Archipelago, Ocean Sci. (2019) 1341–1349.
Perspect. 8 (4) (1994) 19–43.
[9] Galapagos, 2019. Retrieved from: https://www.galapagos.gob.ec.
[24] R.C. Mitchell, R.T. Carson, Using Surveys to Value Public Goods: the Contingent
[10] J.B. Marre, O. Th�ebaud, S. Pascoe, S. Jennings, J. Boncoeur, L. Coglan, Is economic
Valuation Method, Rff Press, 2013.
valuation of ecosystem services useful to decision-makers? Lessons learned from
[25] B.J. Kanninen, Optimal experimental design for double-bounded dichotomous
Australian coastal and marine management, J. Environ. Manag. 178 (2016) 52–62.
choice contingent valuation, Land Econ. 69 (2) (1993) 138–146.
[11] M.A. Zambrano-Monserrate, C.A. Silva-Zambrano, M.A. Ruano, The economic
[26] T.C. Haab, K.E. McConnell, Valuing Environmental and Natural Resources: the
value of natural protected areas in Ecuador: a case of Villamil Beach National
Econometrics of Non-market Valuation, Edward Elgar Publishing, 2002.
Recreation Area, Ocean Coast Manag. 157 (2018) 193–202.
[27] X. Guo, H. Liu, X. Mao, J. Jin, D. Chen, S. Cheng, Willingness to pay for renewable
[12] M.A. Zambrano-Monserrate, M.A. Ruano, Does environmental noise affect housing
electricity: a contingent valuation study in Beijing, China, Energy Pol. 68 (2014)
rental prices in developing countries? Evidence from Ecuador, Land Use Pol. 87
340–347.
(2019) 104059.
[28] R.C. Bishop, T.A. Heberlein, Measuring values of extramarket goods: are indirect
[13] E.C. Choi, J.S. Lee, The willingness to pay for removing the microplastics in the
measures biased? Am. J. Agric. Econ. 61 (5) (1979) 926–930.
ocean–The case of Seoul metropolitan area, South Korea, Mar. Pol. 93 (2018)
[29] K. Arrow, R. Solow, P. Portney, E. Leamer, R. Radner, H. Schuman, Contingent
93–100.
valuation methodology report. Report of the NOAA Panel on contingent valuation,
[14] K. Ostberg,
€ L. Hasselstr€
om, C. Håkansson, Non-market valuation of the coastal
Fed. Regist. 58 (1993) 4602–4614.
environment–Uniting political aims, ecological and economic knowledge,
[30] I.J. Bateman, R.T. Carson, B. Day, M. Hanemann, N. Hanley, T. Hett, R. Sugden,
J. Environ. Manag. 110 (2012) 166–178.
Economic Valuation with Stated Preference Techniques: A Manual. Economic
Valuation with Stated Preference Techniques: a Manual, 2002.

8
M.A. Zambrano-Monserrate and M.A. Ruano Marine Policy 117 (2020) 103933

[31] N. Borzykowski, A. Baranzini, D. Maradan, Scope effects in contingent valuation: [44] R. JianJun He, W. Wang, H. Gong, Valuing cultivated land protection: a contingent
does the assumed statistical distribution of WTP matter? Ecol. Econ. 144 (2018) valuation and choice experiment study in China, Land Use Pol. 74 (2018) 214–219.
319–329. [45] T.D. Khong, A. Loch, M.D. Young, Inferred valuation versus conventional
[32] N. Zografakis, E. Sifaki, M. Pagalou, G. Nikitaki, V. Psarakis, K.P. Tsagarakis, contingent valuation: a salinity intrusion case study, J. Environ. Manag. 243
Assessment of public acceptance and willingness to pay for renewable energy (2019) 95–104.
sources in Crete, Renew. Sustain. Energy Rev. 14 (3) (2010) 1088–1095. [46] M. Cazabon-Mannette, P.W. Schuhmann, A. Hailey, J. Horrocks, Estimates of the
[33] R. Cummings, L. Taylor, Unbiased value estimates for environmental goods: a non-market value of sea turtles in Tobago using stated preference techniques,
cheap talk design for the contingent valuation method, Am. Econ. Rev. 89 (3) J. Environ. Manag. 192 (2017) 281–291.
(1999) 649–665. [47] M.A. Skeie, H. Lindhjem, S. Skjeflo, S. Navrud, Smartphone and tablet effects in
[34] R. Labao, H. Francisco, D. Harder, F.I. Santos, Do colored photographs affect contingent valuation web surveys–No reason to worry? Ecol. Econ. 165 (2019)
willingness to pay responses for endangered species conservation? Environ. 106390.
Resour. Econ. 40 (2) (2008) 251–264. [48] B. Chen, X. Qi, Protest response and contingent valuation of an urban forest park in
[35] A. Dribek, L. Voltaire, Contingent valuation analysis of willingness to pay for beach Fuzhou City, China, Urban For. Urban Green. 29 (2018) 68–76.
erosion control through the stabiplage technique: a study in Djerba (Tunisia), Mar. [49] A.L. Indab, Willingness to pay for whale shark conservation in Sorsogon,
Pol. 86 (2017) 17–23. Philippines, in: Marine and Coastal Ecosystem Valuation, Institutions, and Policy in
[36] K.J. Egan, J.R. Corrigan, D.F. Dwyer, Three reasons to use annual payments in Southeast Asia, Springer, Singapore, 2016, pp. 93–128.
contingent valuation surveys: convergent validity, discount rates, and mental [50] D. Cook, K. Eiríksd�ottir, B. Davíðsd�ottir, D.M. Krist�
ofersson, The contingent
accounting, J. Environ. Econ. Manag. 72 (2015) 123–136. valuation study of Heiðm€ ork, Iceland–Willingness to pay for its preservation,
[37] J.M. Halstead, A.E. Luloff, T.H. Stevens, Protest bidders in contingent valuation, J. Environ. Manag. 209 (2018) 126–138.
Northeastern J. Agric. Resour. Econ. 21 (2) (1992) 160–169. [51] A. Bengochea-Morancho, A.M. Fuertes-Eugenio, S. del Saz-Salazar, A comparison
[38] E. Strazzera, R. Scarpa, P. Calia, G.D. Garrod, K.G. Willis, Modelling zero values of empirical models used to infer the willingness to pay in contingent valuation,
and protest responses in contingent valuation surveys, Appl. Econ. 35 (2) (2003) Empir. Econ. 30 (1) (2005) 235–244.
133–138. [52] B. Borrel, Tackling Plastic Pollution in the Gal�apagos, 2019. Recuperado de: https:
[39] I. Grammatikopoulou, S.B. Olsen, Accounting protesting and warm glow bidding in //www.wwf.org.ec/?uNewsID¼343790.
Contingent Valuation surveys considering the management of environmental [53] R.C. Monteiro, J.A.I. do Sul, M.F. Costa, Plastic pollution in islands of the Atlantic
goods–An empirical case study assessing the value of protecting a Natura 2000 ocean, Environ. Pollut. 238 (2018) 103–110.
wetland area in Greece, J. Environ. Manag. 130 (2013) 232–241. [54] S.C. Gall, R.C. Thompson, The impact of debris on marine life, Mar. Pollut. Bull. 92
[40] H. Aizaki, T. Nakatani, K. Sato, Stated Preference Methods Using R, Chapman and (1–2) (2015) 170–179.
Hall/CRC, 2014. [55] J. Hausman, Contingent valuation: from dubious to hopeless, J. Econ. Perspect. 26
[41] B. Kristr€
om, Spike models in contingent valuation, Am. J. Agric. Econ. 79 (3) (4) (2012) 43–56.
(1997) 1013–1023. [56] P.A. Samuelson, The pure theory of public expenditure, Rev. Econ. Stat. 36 (1954)
[42] J. Orgill-Meyer, M. Jeuland, J. Albert, N. Cutler, Comparing contingent valuation 387–389.
and averting expenditure estimates of the costs of irregular water supply, Ecol. [58] R. Ghosh, Y. Goyal, J. Rommel, J. Sagebiel, Are small firms willing to pay for
Econ. 146 (2018) 250–264. improved power supply? Evidence from a contingent valuation study in India,
[43] S. Ferreira, R.C. Marques, Contingent valuation method applied to waste Energy Pol. 109 (2017) 659–665.
management, Resour. Conserv. Recycl. 99 (2015) 111–117.