Spatial heterogeneity and trans-boundary pollution in cross drainage basin: a
contingent valuation study on Xijiang River drainage basin in South China?
Jie He1, Anping Huang2, Luodan Xu2
This version May 6th 2014
(Preliminary version, please do not cite)
Abstract
This article aims to examine whether and how the location of a city along a river can
affect its residents’ Willingness to Pay (WTP) for a river water quality improvement
project. The underlying mechanism is the transboundary river water pollution spillover
which may reduce people’s expectation about the utility increase that they can obtain
from cleaner water, therefore leads to under-reported WTP. Our analysis is based on a
CVM survey conducted in 20 cities of 4 Provinces of South China located in the Xijiang
river basin during July-August 2012. To test this hypothesis, the water quality of the part
of Xijiang River belonging to an upstream city is included into the determination
function of the WTP of a respondent living in its direct downstream city. Our results
demonstrate that besides the city-specific river water quality and other macroeconomics
characteristics, the upstream city’s water quality is also a significant and negative
determinant in the formation of WTP. A downstream city’s respondents report lower
WTP when their direct upstream is more heavily polluted. We also found the negative
externality on WTP of downstream cities to be decreasing with the distance that
separates them from their direct upstream city. This negative externality is also proven
to be mitigated by the negotiation power of a downstream city with respect to its direct
upstream counter-partner. Our results reveal that higher is the negotiation power of the
downstream city (measured by relative ratio of per capita gross regional product (GRP),
GRP growth rate and population size), lower will be the impact of negative externality
of transboundary pollution on WTP. The under-reported WTP due to the transboudary
river pollution reminds us of the potential huge social benefit that can be created by a
river basin level water quality management project, which, if functioning efficiently, can
largely reduce the concerns of the downstream people about the pollution spillover from
the upstream cities. Our study also provides interesting insight about the huge potential
of the hotly debated pilot payment for ecological service regime between upstream and
downstream regions, considered as a market-based solution for transboundary river
water pollution control in China.
Keywords: transboundary water pollution, river, negative externality, spatial, contingent
valuation, China
1
2
University of Sherbrooke, Quebec, Canada. Jie.He@usherbrooke.ca.
University of Sun Yat-sen, China.
1. Introduction
The stated preference non-market valuation methods are very often used to assess public
preference for environmental quality improvement policies or projects. Although the spatial
characteristics may play an important role in willingness-to-pay (WTP) determination (e.g.
Bockstael, 1996; Johnston et al. 2002, Bateman et al. 2006, etc.), the increasing attention to
include consideration of the spatial characteristics in stated preference studies are relatively
recent, the earliest ones can be traced to Desvousges et al. (1987) and Bockstael (1996).
Johnston et al. (2002) suggested that “[the] omissions [of spatial factors] may be particularly
troubling in case where the survey instrument itself provides cartographic details that
respondents may associate with particular spatial characteristics”. Many conventional WTP
determinants, varying from initial environmental condition to some specific economic
characteristics (Bateman, 2009), may present large spatial heterogeneities, are therefore able
to diverge the value of WTP of people living in different places for a same environmental
quality improvement project.
The spatial preference heterogeneity can be potentially important for the river water
quality related valuation studies. As large scale generic resources, rivers often flow through
boundaries between political jurisdictions (communities, cities, provinces and even nations).
If the different sides of the boundaries apply diverged environmental policies, combined
with initial difference in geographical and economic conditions, water quality across
boundaries may not be uniformly and continuously distributed but more patchy and discrete
in their presence (Jørgensen et al., 2012). Previous studies such as Tait et al. (2012), Moore
et al. (2012) and Lanz and Provins (2013) have already provided some solid evidences about
how the WTP of a respondent to be dependent on the very spatially precise initial
environmental quality or other related characters, under the terms of “spatial scope
sensitivity”.
In this paper, besides the “spatial scope sensitivity”, we will study another dimension of
factors that may also affect people’s WTP in a spatial way, this is the transboundary pollution,
a phenomenon closely related to the public good characteristic of river water. The flow of
rivers creates up- and downstream regions. The administrative boundaries between regions,
however, do not prevent pollution acrossing regional borders. Without a uniform water
quality standard applied to the whole river system, which respects the basic Samuelson
condition and equalises the marginal benefit from water quality control summed over all
regions along the river to the marginal abatement cost, a polluting upstream region, either not
experiencing the full benefits of pollution control or not bearing the full cost of pollution
damage, may not exert sufficient control of their pollution discharge into the river. Even
worse, Oates (2001) further indicates such a situation can lead to a race to the bottom of the
regional water quality control policies. In such situation, given a same hypothetical water
pollution improvement scenario, further a person living in the downstream part of the river,
stronger could be his/her doubt about the possibility of enjoying the proposed better water
quality, therefore lower would be his/her WTP, everything else being equal.
Not yet investigated by the valuation literatures, the existence of the phenomenon of riverlevel transboundary pollution has already been confirmed on international and province/state
level in the studies directly focusing on the spatial distribution patterns of river water
pollution concentration. Sigman (2002) used the biochemical oxygen demand (BOD) data
measured by 291 river monitoring stations in 49 countries during 1979-1990 from
GEMS/Water 1 and found that the BOD indicator to be significantly higher at stations
upstream of borders than other comparable stations, at least for the stations located in non
EU (European Union) countries. As most US federal environmental policies give the
responsibility of regulation implementation and enforcement to state level, Sigman (2005)
also investigated the potential trans-boundary spillover phenomena inside of the US, using a
composite water pollution index based on five major pollutants compiled from the data of
618 monitoring stations on rivers in the US from 1973-1995. Based on a difference-indifference logic, Sigman (2005) revealed that, all else equal, a water quality index to be
significantly 4% worse at the stations downstream from a state authorized with the
discretion to implement and enforce environmental regulation. Using GIS water quality
panel data of 321 monitoring stations across Brazil and the data about the modification of
jurisdictional boundaries of 5500 Brazilian counties, Lipscomb and Mobarak (2007) studied
whether water quality across jurisdictional boundaries deteriorates due to the increase in
pollution close a river’s exist point out of a jurisdiction. Their results also confirmed a
pollution increase of 2.3% for every kilometer closer a river get to the exiting border is
within 5 kilometers. The transboundary pollution phenomena are not only limited to river
water pollution, but also in air pollution aspect. Gray and Shadbegian (2004) examined the
determinants of environmental regulatory activities on air and water pollution of 409 US
pulp and paper mills during 1985-1997 and found both the inspection and monitoring efforts
reduce significantly for plants located close to a state border, since the out-of-state residents
receive a larger share of benefit of these pollution control activities. Helland and Whitford
(2003) also report a transboundary spillover case for air pollution in US, illustrated by
higher toxic chemical releases in border counties.
The principal hypotheses in this paper will therefore be that the geographical location of a
respondent along a river basin can significantly affect his/her demand for river water quality
improvement, not only through the current river water pollution situation and other related
social and economic condition of the regions where reside the respondents, but also through
the geographical location of the regions along the river relative to other regions. To test
these hypotheses, this paper reports a study using Contingent Valuation method (CVM) to
examine the determinant factors for people’s willingness to pay (WTP) for a hypothetical
river water quality improvement project. Our CVM survey was conducted in 20 cities of 4
Provinces of South China belonging to Xijiang river basin during July-August 2012. Xijiang
Rivier is an interesting case for spatial-related study since the regions along this river present
high disparity in economics development and environmental pollution during the last several
decades. Dichotomous choice and MBDC format WTP questions were randomly presented
to local residents of the 20 cities. City-level macroeconomics and river water quality data are
used in econometrics model along with individual-level socio-economic characteristics to
measure the heterogeneity of WTP between people living in different cities for the proposed
water quality improvement. Moreover, our paper will pay special attention to test whether
and how the water quality and polluting economic activities of the upstream cities have
significant impacts on the preference and perception of the downstream cities for a same
river quality conservation project.
One of our central finding is that, although in general, respondents living in the cities
facing more polluted river water are willing to pay more, all else being equal, those living in
downstream of Xijiang river report significantly lower WTP when their upstream is heavily
polluted. We also test the robustness of our results by adjusting the influence of water
quality of the upstream city with the distance from the direct downstream city to it and with
1
UN’s Global Environmental Monitoring System Water Quality Monitoring Project.
the relative negotiation power of direct downstream city with respect to it, and obtain quite
stable results. These findings actually reveal the necessity to include the inter-relationship
between regions located on a same river in CVM study while studying a river basin level
water quality improvement project, in the aims to take care of the negative externality of
river water pollution.
The paper is organized as follows: in Section 2, we reviewed spatial factor related stated
preference valuation literature. Section 3 outlined the river basin level transboundary
pollution situation in China. Section 4 presented the background of Xijing river basin.
Section 5 and 6 provided some details about our survey and about data cleaning works.
Section 7 discusses the principal hypotheses to test under the scenario of river transboundary
pollution. The main models, results of our estimations and the robustness check will be
given in the section 8 and 9. Finally, we will provide some discussion in section 10 about the
pertinence and policy implication of our findings for China’s pilot ecosystem service
compensation mechanism which aim to reinforce transboundary coordination in river
pollution control. Finally, we will conclude in Section 11.
2. Literature Review
Spatial factors get increasing attention in recent stated preference (SP) environmental
valuation studies. 1 Such research interest is mainly motivated by the potential spatial
preference heterogeneity. Considering the spatial preference heterogeneity can affect both
the size of the economic jurisdiction in which relevant people have positive WTP (Bateman
et al. 2006; Hanley et al. 2003) and the sample mean WTP (Bateman et al. 2006; Pate and
Loomis, 1997; Bockstael, 1996): the two factors on which the aggregation of total benefit
for non-market goods commonly relies; more and more authors believe without controlling
spatial heterogeneity can come at the cost of the precision of aggregate WTP. A good
example is Bateman et al. (2006), in which the authors found the aggregate WTP based on
the “economic jurisdiction” and spatially controlled individual WTP comes out to be sixteen
times smaller than the simple aggregate WTP based on “geographical jurisdiction” and
sample mean individual WTP.
The most frequently discussed spatial heterogeneity in the literature is the so-called
“distance-decay” (Loomis, 2000), which suggests WTP to decline with the increase of the
distance between a respondent’s place of residence and the site providing ecosystem services.
(Schaafsma et al. 2013). Various reasons have been proposed in the past decade to
explain/interpret distance-decay. Sutherland and Walsh (1985) believe the information about
the valuated site motivates positively people’s WTP, a longer distance can lead to higher
cost to obtain the information about the site, and therefore reduces people’s WTP. Hanley et
al. (2003) pointed out the necessity to differentiate between user and non-user, and found
more rapid distance decay exists for use value than for non-use values. Bateman et al. (2006)
underlined the potential difference in the nature of distance decay between the equivalent
loss studies which examine preservation value and the compensation surplus studies which
estimate the environmental improvement value. As it is more possible to have non-users to
convert to user in environmental improvement project and the conversion from non-user to
user is more likely to occur for households nearer to the resource, Bateman et al. (2006)
therefore suggested that the distance decay can be observed among either user or non-user,
1
Past literatures already studied numerous spatial attributes and their impact on people’s preference for
ecological goods and services provided by large-scale ecosystems as river (Desvousges et al. 1987;
Holmes et al. 2004; Zhiqiang et al. 2002), forest, wetland and other eco-system (Broch et al.,2012;
Guimarães et al.,2011; Zhongmin et al.,2003), etc.
at least for the compensation surplus studies. Other authors tried to explain the distance
decay by the availability of substitutes (Brouwer and Slangen, 1998; Hanley et al. 2003;
Jorgensen et al., 2013 and Schaafsma et al. 2013). As the number of substitutes often
increases with the distance from the valued site, greater is the distance, lower will be the
demand for the valued site, and therefore lower will be WTP. Schaafsma et al. (2012) tried
to further refine this logic by identifying the directional heterogeneity in the availability of
the substitutes, which may affect the distance decay phenomenon in a heterogeneous way in
different directions. Finally, Some authors (Sutherland and Walsh, 1985; Rolfe and Windle,
2010, Jorgensen et al. 2013, among others) also mentioned the importance to distinguish
between generic non-market resources (i.e. the wetland in a country) and an iconic ones (i.e.
Yellow Stone national park or the Great Barrier Reef), distance decay affects less likely
these goods with large scale importance, as people’s awareness and knowledge about them
are less sensitive to the distance and there are few significant substitutes.
Parallel to the phenomena of distance-decay, spatial scope sensitivity (Lanz and Provins,
2013) is another aspect of spatial preference heterogeneity that starts to attract researchers’
interests in SP studies. More pertinent for generic natural resource with relatively large
geographical scale and non-homogeneously distributed quality/quantity, spatial scope
sensitivity interests in understanding how an individual’s WTP varies according to the
spatial variability in the quality/quantity of an environmental resource itself. Johnston et al.
(2002) examined how the WTP for rural land development project is dependent on whether
or not the developments located adjacent to main roads, etc., Moore et al. (2012) found
respondents’ WTP for a universal water clarity improvement in Green Bay, Wisconsin to be
dependent on the initial clarity of the water closest to their residence. Since the dirtiest part
of water in the bay is close to population center, they found their “geospatial referencing
model” give much higher aggregate willingness to pay for the project then the base mode
which suggests households to be a single entity defined only by its average water quality in
the bay. Tait et al. (2012) combined the spatially related water quality data into their choice
experiment study via the Geographical Information System (GIS). They found that the
respondents who live in the vicinity of low quality waterways are generally willing to pay
more for the improvements relative to those who live near to high quality waterways. Broch
et al. (2013) studied how farmers’ willingness to provide a good or services are affected by
spatially heterogeneous in species richness, human population density, share of are with
special groundwater interests, share of the forest cover and hunting resources, their
conclusion also confirms the necessity to include spatial variations in designing conservation
policies.
The studies interested in spatial scope sensitivity therefore focalize more on the
characteristics of the valued goods and services themselves and the geographical variability
of these characteristics. These mark their differences from the studies of distance-decay,
which give more attention to the distances separate the individuals from the valued
goods/services and availability of their substitutes. Naturally, the spatial scope sensitivity
related studies depend more heavily on the availability of the geographical data describing
the spatial differences in either quality or quantity of the valued goods and service. It is also
these geographical data that enable the studies to provide precise WTP estimation in more
detailed geographical scale.
Another dimension of factors that can provoke spatial preference heterogeneity, to which
we can group the preference heterogeneity caused by river drainage basin trans-boundary
pollution problem that we will study in this paper, is the potential inter-relationship between
people of different regions. An improvement project applied on a generic natural resource
can be viewed in different ways by respondents living in different place according to how
the provision of the natural resource can be spatially differentiated and how this spatial
differentiation can affect the utility improvement for a respondent located in a specific
location. Brouwer et al. (2010) provided such an example. By assessing the preference
heterogeneity related to the spatial distribution of water quality improvement throughout a
river basin via a choice experiment study, they found that irrespective of the sub-basin
where people live, the improvement of water quality to a good state is valued equally by
local and non-local residents. However when improving water quality on level further to
very good quality, local residents place an additional value on this improvement if it occurs
in their own sub-basin. Another interesting example is Lanz and Provins (2013), which
examined explicitly how the preference of respondent changes with the location of
improvement project by including the coverage of policy as an attribute in their choice
experiment design. Besides finding logical strong preferences for improvement in local
environmental amenity, they also observed that with the enlargement of the spatial coverage
of the project, preference heterogeneity between regions increases, accompanied by a
declining in the sample mean WTP. Their explanation is that “for some attributes,
respondents can display negative surplus for the provision of improvement in other nearby
regions, which signals competition for the appropriation of rivalrous benefits” (Lanz and
Provins, 2003, P107).
Our paper enters naturally to the spatial scope sensitivity literatures since people’s WTP is
assumed to be directly related to their location along the river, but the particularity or the
innovative point of this study compared to the existing spatial related studies is that we will
put more attention to the inter-relationship between cities/regions along the same river and
how this inter-relationship can affect directly people’s WTP. This is according to our
knowledge the first paper discussing this aspect of spatial preference heterogeneity,
Moreover, our study also brings an interesting case for the empirical river basin
transboundary pollution literatures as Sigman (2002, 2005) etc., although the transboundary
phenomenon is analyzed in our paper indirectly.
Our study also differentiates from the other existing non-market valuation studies aiming
at valuing the water related ecological services in China and other developing countries and
regions, such as Wei et al.(2006) in the North China Plain, W. Yang et al.(2008) in
Hangzhou, China; Tong et al.(2007) in Wenzhou, China, Zhiqiang et al.(2002) in Heihe
river basin in China, Zhongmin et al.(2003) in Eijna region of China, Choe et al.(1996) in
Philippines, Vásquez et al.(2009) in Mexico, Boadu(1992) in Ghana, Hammitt et al.(2001)
in Taiwan., etc. Those previous studies were more focused on specific regions/areas, instead
of taking the whole river basin into consideration, even less have they discussed the
potential role played by spatial factors in value/preference determination.
3. River basin level transboundary pollution in China
Most of China’s large-scale river basins are composed of several regional jurisdictions
(province/region/city). It is therefore evident that decisions taken by one regional
jurisdictions can have implications across administrative boundaries, just as what have been
reported for the international case or those from the US and Brazil. From the point of the
view of environmental regulation, this risk seems to be confirmed for China, because
although in general environmental policy is formulated centrally and local jurisdictions have
only the authority to set their own environmental standard more stringent than that of
national level, the implementation responsibilities are devolved to the branch office of
Ministry of Environment Protection (MEP), operating at the provincial, municipal and
county levels (Hills and Robert, 2000).
The monitoring and enforcement efficiencies of these branch offices of MEP are largely
affected by the complexity and fragmentation of the different authorities. Yu (2011)
describes the complex relationship between the Ministry of Water Resource Management
(MWRM) which deals with water quantity and water utilization and the MEP which
coordinates and solves trans-boundary environmental pollution disputes. He believes there
are either overlaps or gaps between the competences of the two authorities, which may
largely compromise the efficiency of their efforts in trans-boundary river water pollution
control. Hills and Man (1998) indicated that the policy enforcement difficulties in China can
be further compound by the networking relationship (guanxi in Chinese) between regulators
and polluters, which can result in erratic and inconsistent enforcement of environmental
protection measures. Yu (2011) believe such a possibility is especially true in the current
context where the budget of lower branches of MEP is dependent on local governments,
who may give higher priority to economic growth instead to environment protection.
Another particularity of China is that, given the lack of water resources, it is very common
for each jurisdiction to construct water conservation structures, such as dams or floodgates,
which may further enable each region to control the distribution of pollutants within its
jurisdiction by managing the pollutant flow more directly to the downstream jurisdiction
(Zhao et al. 2012)
These particularities, along with lack of efficient transboundary pollution agreements and
concrete dispute settlement mechanisms, have already caused severe water pollution at
trans-boundary interfaces (Wang, 1999; Zhao et al. 2005; Yu, 2012; Zhao et al. 2012).
Examples are the blue-green algae out-break in Taihu Lake Basin in May 2007 (Zhao et al.
2012) and the very well documented transboundary pollution between Guangdong Province
and Hong Kong via the Pearl River drainage basin (Hills and Roberts, 2001; Hills et al. 1998,
Marsden, 2011, among others), etc.
4. Background of Xijiang River Basin
Xijiang River is the main stream and longest tributary of Pearl River, the largest river
system in southern China. It flows for 2217 kilometers from the north of Yunnan province
eastward across 2 southern provinces of China: Guizhou province, Guangxi province and
flows through the Pearl River delta in Guangdong province and enters the South China Sea
near Macau.
Xijiang River basin accounts for 78% of Pearl River basin (409,480 km²) , and it drains
the majority of Southern provinces, as well as parts of Southwest of China, According to the
6th nationwide population census of China (conducted in 2010) , the population living inside
of Xijiang River basin amounts to 100 million. The regions belonging to the basin are very
important in term of economic activities, with an annual gross output of 3790 billion Yuan
(2010) and an annual growth rate over 10% until 2012, this river basin covers the most
developed area of Southern China.
During last 35 years, Xijiang River Basin, like other regions in China, experienced
increasing inequality in economics development. Taking the provinces Guangdong and
Guangxi as example: before the beginning of economic reform, in 1978, Guangdong’s gross
regional output (GRP) per capita was 370 Yuan (220.24 USD), about 1.5 times higher than
that of Guangxi (246 Yuan, equal to 146.43 USD). However in 2008, Guangdong’s GRP per
capita has reached to 37589 RMB (5408.49 USD), which is about 2.5 times higher than that
of Guangxi (14966 RMB, equalt to 2156.48 USD).1 Along with the widen disparity in per
1
The conversion between RMB and USD are conducted with official exchange rate, which was 1
capita GRP, we also observe increasing gap between these two provinces in economics
structure, urbanization level, economics openness, etc.
Uneven development between regions naturally leads to very different interpretation of
regional government about the relationship between environment and economy. Although
the environmental conscience start to awake in some richest eastern coastal provinces/cities,
many inland western regions are still willing to endorse environmental damage to attract
investment in polluting productive sectors. Such a situation may further exacerbates the
above mentioned transboudary pollution problem, since a particularity of Xijiang river basin
is that most of the poor inland provinces/cities, which are willing to sacrifice environment
for growth, are located in the upstream of the Xijiang river basin. Such a situation might
unfortunately be further enhanced by the fact that the two inland provinces (Yunnan and
Guizhou) located to the upstream of Xijiang River basin are rich in nonferrous metal
reserves, whose extraction is very polluting, particularly for water resources, either directly
by discharging of mining wastewater or indirectly by extraction-related deforestation and
soil-erosion/contamination. The accumulated industrial wastewater pollution further
intervenes with huge domestic wastewater discharge burden and great water demand
pressure of the alongside cities. All these factors have contributed to the important
accumulation of deterioration of Xijiang River water quality, especially for the downstream
cities.
5. Survey
To measure the WTP for the ecological goods and services provided by the Xijiang River,
we conducted in July-Auguest 2012 a large-scale cross-region CVM survey in 20 cities of
four provinces (Guangdong, Guangxi, Guizhou and Yunnan) along the Xijiang River.
The design of the questionnaire benefited from a series of focus-group composed of
researchers, students and experts (both academic and operational) working on hydrology and
geography fields. The discussions covered the content and form of questionnaire,
appropriateness and efficiency of wording, rationality of the hypothetical scenario,
feasibility and credibility of the proposed payment vehicle, etc. Moreover, pre-test was
conducted in Guangzhou city with the input of over 50 respondents randomly solicited.
Their comments about the questionnaire have been included in final questionnaire.
The final version of the questionnaire can be divided into 3 sections: (1) general
understanding and attitudes of respondents toward the current condition of environmental
and Xijiang River water pollution around sampled city, (2) description of hypothetical
scenario proposing water quality improvement and elicitation question of willingness to pay,
(3) social-economic and demographical information of respondents and their anticipation of
changes in income and welfare if the hypothetical scenario turns to reality.
Based on the discussion of focus group and pilot study, tax as a payment vehicle was
rejected, since few people shown their empathy to a new tax and doubt about its efficiency
in environment improvement. We therefore used voluntary donation as payment vehicle for
hypothetical water quality improvement project.
To facilitate the understanding of the respondents about the different levels of water
quality, we use in the survey the water quality ladder initially proposed by Mitchell et
Carson (1989). The river basin level uniform water quality improvement target is fixed at
the swimmable level (C level as illustrated in the annex 1), which corresponds to the level II
USD=1.68 CNY in 1978 and 1USD=6.95CNY in 2008.
in the classification of the Ministry of Environment Protection of China. More recent water
quality ladder design, such as the 4D version image proposed in Hime et al. (2009) has also
been in our consideration while designing questionnaire, we finally took the simpler but
classical version of Mitchell and Carson (1989) due to its simplicity. Moreover, given the
current bad water quality of Xijiang, some elements proposed in the water quality ladder of
Hime et al. (2009) comes out to be less relevant, since a large part of the ecological service
are currently very weak or absent in Xijiang river. Furthermore, few recent experience with
cleaner river water environment may also affect respondents’ understanding of certain
ecological services proposed in the water quality ladder of Hime et al. (2009). Appendix 1
provides the water quality ladder used in our survey.
Two types of elicitation format were employed in split sample in our survey. Randomly,
50% of respondents have encountered a double bounded dichotomous choice (DC2) WTP
question and the other 50% have answered their WTP question in a multiple bounded
discrete choice (MBDC) matrix. Both questionnaire versions share the same 14 bid price
range varying from 0 yuan till 500 yuan. In DC2 version questionnaire, respondents were at
first asked “would you like to donate ___ Yuan for a program that will improve water
quality of Xijiang river to be clean enough for swimming (level C)?” Depending on his/her
answers to this question, a higher/lower bid price will be further proposed. Table 1 present
the 14 groups of alternative bids used in the double biding regimes. The MBDC
questionnaire invited respondents to fill out the following matrix with their level of certainty
to accept each of the 14 bid price, c.f. Table 2.
To ensure a uniform and efficient sampling strategy in the 20 cities, we decided to
conduct intercept face-to-face survey in public locations. The waiting areas of Outpatient
section of hospitals was identified as appropriate places, given their large traffic flow and
relative balanced attendance of people of different social classes. 30 students from 4
universities (Sun Yat-sen University of Guangdong Province, Zunyi Medical College of
Guizhou Province, Guangxi Normal College of Guangxi Province, Yunnan University of
Yunnan Province) were recruited and trained for the survey.
Stratified random sampling was used to ensure good balance of representativeness of the
20 surveyed cities. For each city, its sample size was determined as roughly proportional to
its total population size. The details of the sample size in the 20 cities are given in Table 3.
In total, we got 2223 returned questionnaires. After deleting the respondents under 18 years
old and those having lived in sampled city for less than 3 years (to ensure respondent to have
enough knowledge about situation of environmental problem in the city), we obtain a total
database of 2133 usable questionnaires, in which 1090 were presented with a DC2 version
questionnaire and the remaining 1023 were interviewed by MBDC questionnaire.
The comparison in respondents’ socio-demographic characteristics between two
subsamples in table 4 shows quite good similarity, except the average education years, with
the people answered MBDC questionnaire to be 0.7 years longer.
Table 5 presents comparison between provinces and with population census data. Clearly,
across the 4 provinces, the biggest differences are in average income and education level.
Coherent with China’s regional disparity patterns: the respondents from Guangdong
province, the richest province in South China, enjoy the highest income and education level.
People from Yunnan province, owing to rich natural resources, also report relatively high
income but have lowest education level. The respondents from Guizhou and Guangxi in
general have much lower income, although the education level is only slightly lower than
Guangdong.
With respect to census statistics, the gender structure gives good coherence in three over
four provinces, except for Yunnan, whose male respondents are over-represented (62.37%
vs. 50.85%). Well educated people are overrepresented in all the 4 provinces. This bias can
be explained by three reasons. The first is that our survey is only carried out only in urban
areas, where the population is in general more educated. Secondly, the survey mode that we
chose may discourage less educated people, who may not want to be interviewed by
university students. Finally, people may have tendency to over-report their education level
since the interviewers are college students. A similar but less important bias of our sampling
is age. Table 5 shows that our sample was heavily skewed towards the younger generation,
with the majority of people being between 20 and 40 years old. This might also be explained
by our survey mode.
6. Data cleaning for WTP answers
Details about data cleaning for the answers given to WTP question are illustrated in table
6. Among the 1023 respondents of MBDC question, 119 respondents expressed negative
probability to accept the project even at zero bid. We also find 99 respondents affirming
their extremely high demand by giving positive responses even at the highest bid price, 1000
Yuan. Another 120 returned questionnaires present missing values in MBDC matrix, which
make the answers non-usable. We only find one disordered answers to MBDC matrix WTP
question, this shows relatively good understanding of the respondents about the logic of the
MBDC questionnaire. For the 1090 respondents of DC questionnaire, 143 respondents gave
negative probability to accept the project at zero bid and 37 respondents gave positive
answer at the highest bid price. There are also 63 missing answers.
The negative answers at the zero bid and the extremely high demand WTP answers can be
further classified into different categories based on the answers that respondents gave to the
two follow-up questions. As illustrated in Table 7, among the 119 negative zero answers,
102 can be considered as protest (9.74%), with a majority of respondents (87 persons, 8.3%)
expressed their doubt about the proper usage of the fund and another 15 persons (1.44%)
believe it should be the responsibility of government to support financially the project. The
rest 17 respondents expressed their weak demand: little importance of the project (3 persons),
the potential heavy financial burden of such a project to their family expenditure (14
persons), we therefore consider their WTP as zero. For the DC2 data, we identified 82
protest answers: 76 expressed their low confidence in the usage of the fund and 6 others
believe it should be government’s responsibility. The other 61 persons can be considered as
giving real zero WTP.
Among the various reasons to explain the extremely high demand responses (c.f. Table 8),
we find 73 (=31+42) in MBDC format questionnaire and 35 (=9+26) in DC questionnaire
can be considered as unreliable. We can only confirm two respondents in DC survey and 26
(=5+8+13) in MBDC survey as expressing valid high demand for the project.
The frequency (in %) of the five probability responses in MBDC for each bid price is
summarized in Table 9. The percentage of ‘definitely yes’ answers is decreasing fast from
about 92.6% at the price of zero to less than 1.2% at the price of 200 yuan. While the
percentage of ‘definitely no’ answers increases steadily with the price offered, from 0% at
price of zero to about 95.2% when the price increases to 1000 yuan. Between 20 and 150
yuan, about 50% of the respondents chose the ‘probably yes/no’ or ‘not sure’ response
options, showing the respondents actually have relatively important uncertainty in their
answers.
Table 10 presents the percentage of “yes” responses for each group of alternative bid price
proposed in DC2 version WTP question. The percentage number in bold type presents the
acceptance rate for the initial bids. The number listed above (below) in the same column
reports the acceptance rate for the bid-decreasing (-increasing) path. Vertically comparing,
we can observe the logical negative relationship between the acceptance rate and the bid
price, especially for the bidding game whose initial bid superior to 20 yuan. Comparing
horizontally, we can see that for a same bid price, presented before or after the initial bid can
lead to different acceptance rate. This finding is consistent with that reported in Bateman et
al. (1999).
7. Hypotheses to be tested
Based on the discussion in introduction and literature review, we would like to test the
following five hypotheses in this paper. First of all, considering spatial difference in current
river water pollution situation, we follow the logic of spatial scope sensitivity similar to
Moore et al. (2012) and Tait et al. (2012) and assume WTPi,k of a respondent i of city k to be
dependent on the current water pollution situation of the part of river flowing through his/her
city, Pk. Logically, higher is the pollution higher should be the WTP, we therefore expect a
positive coefficient for this variable.
(⏟) for all i living in city k
Secondly, due to the transboundary water pollution problem, we also assume that the
WTP of a respondent i living in city k to be dependent on the pollution situation of the direct
upstream city j. A more severe water pollution situation in the upstream city will discourage
respondent’s WTP reporting. We therefore expect a negative correlation between the WTPi,k
and the upstream city’s river water, Pj.
( ⏟ ) for all i living in city k, j is the direct upstream city of k
Thirdly, we further assume that the transboundary water pollution’s impact on WTP of
respondent i living in city k described in H2 to follow the pattern of “distance-decay”.
Therefore the negative externality of water pollution situation of the upstream city j on the
WTP of individual i in city k decreases with the distance along the river that separates the
two cities j and k. To capture this distance-decay effect, we will use a cross-term between Pj
and the river distance between j and k Dj,k, where the distance enter as an adjustment term
. When city j and k share a frontier which the Xijiang river flows through, the cross term
takes the whole impact of Pj, as in H2. When there are some river distance between j and k,
the negative externality of Pj on WTPi.k will be mitigated by the distance that separates the
two cities. Moore et al. (2012) used a similar term in her measurement of distance decay. We
expect a negative coefficient for this cross-term.
(
⏟
) for all i living in city k, j is the direct upstream city of k
We also believe that respondent i’s estimation on the probability to realise the target water
quality improvement is also dependent on the negotiation power of the downstream cities j
with the upstream city k. Higher is the negotiation power of city j, lower will be the negative
externality impact of transboundary pollution on WTP. As suggestion in H4, with an expect
negative coefficient, a higher negotiation power of city j will reduce the impact of P j on
WTPi,k. We propose to identify three dimensions of comparative indicators to measure the
relative negotiation power of city k with respect to j: population size, GDP growth rate and
per capita GDP.
(
⏟
) for all i living in city k, j is the direct upstream city of k
Finally, we can also combined the two adjustment effect of the transboundary pollution on
WTP together, this gives us the fifth hypothesis:
(
⏟
) for all i living in city k, j is the direct upstream
city of k
8. Estimation
Table 11 presents descriptive statistics for the independent variables that we intent to
include into the WTP determination function. All individual level variables are directly
obtained from the survey. The city-level economic and structural characteristics are
compiled by authors based on the data from China Urban Statistics Yearbook for year 2011.
The city-level spatial variables are calculated by authors according to geographical
information. The water pollution data come from several sources, the details are provided in
the Annex 2.
The estimation of our WTP determination model is separately conducted for the two split
samples. For MBDC data, we use the generalized double-bounded maximum likelihood
interval model initially proposed by Welsh and Poe (1998), which, applied to each of the
positive certainty levels, can be differentiated into three models: definitely yes, probably yes
and not sure; and the two stage Random Valuation Model (RVM) approach proposed by
Wang and He (2011) to estimate the WTP determination function. 1 The DC2 data are
analyzed by both the simple Probit model which only uses the responses to the first bid, and
the double bounded model initially proposed by Hanemann et al.(1991).
1
This method considers the responses given by an individual in the MBDC matrix can be interpreted as the
verbal likelihood for him/her to accept the corresponding bid prices, therefore advocate, in a first stage, to use
the information to estimate individual WTP distribution, whose mean value signifies the real individual WTP
and variance measures the individual uncertainty. The benchmark verbal likelihood coding strategy for our
analysis is 1.00 for “Definitely Yes,” 0.75 for “Probably Yes,” 0.50 for “Not Sure,” 0.25 for “Probably No,”
and 0.00 for “Definitely No”. The details about the distributions of the estimated individual mean WTP (μ i)
and the standard variance (σi) of the entire sample are given in the annex 3. The estimated individual mean
WTP is then regressed, in a second stage, directly on the identified WTP determinants, in a linear function
form. The results corresponding to Wang and He (2012) presented in the main text are those from the second
stage.
Tables 12.1-12.3 present the results of estimations for the five hypothesis, using the
model of “probably yes” based on Welsh and Poe (1998) approach, the second stage of
Wang and He (2011) approach and the double-bound model for DC2 data of Hamemann et
al. (1991) approach. The other variants of the Welsh and Poe (1988) and Hanemann (1991)
approaches are given in the annex. For each model, the results testing the five hypotheses
are presented. As we use three proxies of relative negotiation power (relative population size,
⁄
⁄
, relative economic growth rate
and relative income level
1
⁄
, the results corresponding to the H4 and H5 are presented each in three
columns, each corresponding to one of the three proxy of the relative negotiation power. A
general observation from the estimation results of the 3 tables, along with the other three
tables in the annex, is the quite similar coefficients for most of the individual-level variables
(Environmental perception, socio-economics factors and relationship with the river). As
expected, income (income_level) is significantly and positively correlated with the mean
WTP in all 6 models. People reporting income to be significantly higher than need are also
willing to pay significantly higher WTP. The negative coefficients of the dummy variable
(resp_gov) reveals the respondents who believe the environmental protection to be
responsibility of government are in general willing to pay less. The level of willingness
(will_serve, varying from 5 to 1) to contribute to environmental protection programs is
found to be significantly and positively associated with WTP, which signifies respondents
with strong will to contribute to environmental protection are willing to pay more. The
positive and significant coefficient of the variable quality_deg indicates that respondents
who believe the quality of the river water in their cities experienced obvious deterioration
have in general a higher WTP. Although not always significant, the positive coefficient of
d_fish also confirms that respondents who consume fish at least once a month are willing to
pay more for the river water quality improvement project. Logically, people living far from
the river seems to report in general lower WTP (c.f. the negative coefficient before far from
the river). The three city-level macro-economic variables in general report quite significant
coefficients. Respondents living in the cities with higher share of secondary industry (c.f.
share2nd) and with higher population density (c.f. pop_density) report lower WTP. Not
always significant, people living in the cities enjoying higher economic growth rate seems to
have higher WTP (c.f. gdp_growth).
The key variables related to the five hypotheses are grouped together in the tables 13,
where only the sign of the coefficients of the critical variables and their statistical
significance are reported. According to our results, all the five hypotheses can be considered
as valid, although the statistic significance of the results from DC2 data stays relatively low.
As assumed in H1, the coefficients associated with the river water quality in the city where
respondents live are found to be positive and significant in most of cases. This signifies that
for a same river water quality improvement target, respondents do base their valuation of the
project on the current water quality of their city, worse was the quality, higher will be their
WTP. This finding, echoing to the results of Moore et al. (2012) and Tait et al. (2012),
reveals the existence of the spatial scope sensitivity in CV studies. The LR test reported at
the bottom of the tables 12.1-12.3 compares the global explanation power of models H2-H5
to that of H1. The related statistics reveal higher explanation power of the models including
⁄
⁄
We tested both the
and
in our estimation and found that using
⁄
gave more coherent results. We can explain this by the following logic: compared to its
upstream counter-partner, poorer a downstream city is, higher negotiation power will it has. This higher
negotiation power will reduce the negative impact of upstream water pollution situation on the WTP of
individual living in the downstream city.
1
the hypotheses H2-H5, confirm the necessity to include upstream city’s water quality into
the WTP determination function.
The inclusion of the upstream city’s river water quality (c.f. degu_upper), as expected in
H2, is found to contribute negatively to the WTP and suggests a worse water quality in the
direct upstream city can lead the WTP of people living in downstream city to decrease, all
else being equal. This finding can be considered as an indirect evidence of the existence of
transboundary pollution problem in Xijiang river basin. The adjustment of the upstream
city’s river quality by distance (H3), negotiation power (H4) and both factors (H5) reveal
quite stable and expected results. The inclusion of the
into the product with upstream
city water quality does not change neither the sign nor the significance of the coefficient of
the upstream water quality related term, The negative coefficient before this cross-term
confirms H3 that the negative externality of the transboundary pollution on WTP decreases
with the distance from the upstream city. Very similar to the case of inclusion of distance
adjustment term, the inclusion of each of the three measurement of the relative negotiation
power between downstream and upstream cities, which are relative population size (
;
relative per capita GDP (
and relative GDP growth rate (
does not modify
the negative coefficient before the variable of upstream city water quality, which therefore
confirms H4 that if the downstream city possesses, compared to its direct upstream city,
relatively larger negotiation power, measured by either a larger population size, a lower per
capita GDP or a faster economic growth rate, a negatie externality of transboundary
pollution on WTP will be weaker. The combination of the two dimensions of adjustment
does not change the results, therefore also provides coherent confirmation for the H5.
9. Robustness check
To test the robustness of the above-described negative externality of transboundary
pollution in WTP determination, we use another three variables as substitute of the water
quality of upstream city. The first one is the distance weighted sum of industrial output of all
upstream cities along the Xijiang River. The reason to choose this variable is based on the
fact that industrial wastewater emission is an important source of river water pollution and
that the waste water emission is positively correlated with industrial output. For a
respondent, without precise information about the pollution flowing from its upstream, it is
natural to extrapolate it via the total industrial production scale along the upstream of the
river. We can use total industrial output to measure it. At the same time, the autopurification function of the river will also allow the concentration of pollutant to decrease,
we therefore use inverse distance (
) as a weight to simulate the auto-purification
procedure, where j (j=1,2, …,S) signifies all the cities located in the upstream of the river of
city k. The industrial output variable used in the paper is therefore calculated as:
∑
S signifies the total number of upstream cities of the direct upstream city of city k.
Parallel to the upstream pollution proxy related to industrial production scale, another
important source of river water pollution is the everyday life waste water. We therefore also
replace the upstream water quality by the distance weighted sum of population of all
upstream cities along the Xijiang River. The distance weighted sum of population as proxy
of upstream city water quality is therefore calculated as:
∑
The third substitute variable for upstream city water quality is the simple interger count
numbers (1, 2,…) which indicate the total number of upstream cities along the Xijiang River
before it flows through the city k, (stagek). stagek can be a good proxy of the negative
externality related to transboudary river pollution for two reasons: firstly, given the water
pollution concentration indicator’s cumulative nature, more cities located in the upstream of
a river, more pollution will be accumulated in the river before it flow by the city k. Secondly,
from the point of view of respondent, facing the risk of transboundary pollution, the
probability for a water quality improvement project to be accomplished should also be
dependant on the number of upstream cities: more cities located in the upstream of a river,
more difficulties will it be to cooperate in controlling water pollution and higher will be risk
of free-riding of upstream cities.
The estimation results are listed in Table 14-16, where the sign of the key variables and
their statistical significances are given. For details of the whole model estimation in each
robustness test, please refer to the annex. We can see that the signs of coefficient of key
substitute variables and their adjusted variants report quite similar and stable coefficients as
in table 13. More specifically, weighted sum of industrial output/population and the stage
both correlate negatively with respondents’ WTP, which reflect the impact of the negative
externality of transboundary pollution on people’s WTP. Similar as the results in Table 13,
this transboundary water pollution related externality also decreases with the distance and
the negotiation power of the downstream cities, similar to our main finding. Therefore we
can conclude that the three robust tests converge to the conclusion found in the last section.
10. Policy discussion
The estimation results of the last section can be used to simulate the city-level average and
aggregate WTP for the hypothetical water quality improvement project under two scenarios.
We call the first one Business as Usual (BaU) in which we assume the transboundary
pollution spillover between up and downstream cities affects directly a person’s WTP as
revealed by our estimation models. The second scenario, to be named as Integrate, assumes
the existence of an efficient river basin level water quality integrate management system
which ensures a good collaboration between between up- and down-stream regions in river
water pollution control, therefore removes concerns of the respondents about the negative
externality caused by transboundary pollution spillover, so in this scenario, we suppose the
coefficients related to the transboundary pollution to become zero.
Table 17 and 18 report the simulation results for the 9 cities along the main tributary of the
Xijiang river from Qujing until Zhuhai, upstream to downstream. In this paper, we use the
estimation results of H5 where the relatively negotiation power measured by GDP growth
rate ratio between upstream and downstream cities to calculate the mean WTP for each city.1
The comparison between the senarios BaU and Integrate reveals the average impact of
transboundary river pollution spillover on average WTP and aggregate WTP for each city
along the river. Clearly, if the concerns about transboundary spillover can be removed, the
1
The choice of the H5 estimation is based on its highest value of the Maximum Likelihood and LR test.
mean WTP of a representative person in the downstream cities will be higher. Further
downstream is the city located along the river, larger will be the increase in the mean WTP.
Multiplying the mean WTP with the total population of corresponding city give the total
WTP (c.f. Table 18). As we can see, the increase in the social benefit in the scenario
Integrate compared to BaU can be quite big. Although the absolute values of social benefit
increase of the project seems to be variables between different estimation models, the relative
percentage increase is at about 10-62%, depending on the estimation models and approaches.
The comparison of the aggregate WTP of these two scenarios actually reveals the potential
social benefit that can be created by a river basin-level integrated water quality management
system, which has already proven its efficiency in reducing river level transboundary
pollution, based on international experiences in the Netherlands (Yu, 2011) and in Quebec
(Fournier et al. 2013). Although the so-called River Basin Committees exist in China for
most of the large rivers, these institutes are currently subordinate to the Ministry of Water
Resource Management (MWRM) and only take care of region level water resource
distribution but not the transboundary pollution problems. Similar as the local environmental
protection office, this committee is not financially self-dependent. Their subordinate
relationship with respect to local government reduces largely their efficiency in reducing the
transboundary negative externality in either the inter-region water resource distribution or
river water pollution (Yu, 2011). The huge social net benefice revealed by our study actually
indicates that it is financial viable to empower the river basin committees with more
transboundary river pollution control responsibilities.
Our simulation results about the net social benefit of integrate water quality management
system can also provide interesting insight for the currently hotly debated mechanism of
Payment for Ecological Service (PES) (Wang, 2004, XXX, XXX), whose idea is currently
tested via several pilot projects in several trans-province level large water bodies, such as
Taihu Lake (Jiangsu and Zhejiang Province) and Xin’anjiang River (Anhui and Zhejiang
province), etc. Concretely, PES aims at ensuring the payment made by people/regions located
in downstream to the people/regions located in upstream of river for their efforts/sacrifices to
assure better river-related ecological services, either in terms of concrete water quality
protection measures or in terms of real water pollutant abatement. Such a payment
mechanism, either ensured by bilateral or multilateral transboundary agreements 1 or by a
river basin committee, provides a potentially feasible remedy for the market failure caused by
river transboundary pollution spillover by compensating the environmental protection efforts
of upstream regions. From this angle, our estimations of the total social benefit differences
between the BaU and integrate water quality management scenarios provide an upper bound
for the amount that a downstream regional is willing to pay for the effort exerted by the
upstream regions to ensure the river not to be additionally polluted by their activities.
1
Such agreements exist already between certains provinces, good examples are the one between the
province of Zhejiand and Jiangsu for Taihu Lake, XXX for Heihe River, etc. (Zhao et al. 2013, Yu, 2011,
etc.)
11. Conclusion
This paper uses contingent valuation method to assess the economic benefits of water
quality improvement project of Xijiang river basin in southern China and gives a detailed
analysis on how the transboundary river pollution spillover can affect people’s willingness
to pay for a river water quality improvement project. Our results suggest that besides the
individual-level characteristics and city-level economic and initial environmental conditions,
the spatial location of the cities along the river with respect to the other cities also exerts
significant influence on people’s WTP for water improvement project. Precisely, our results
reveal that worse is the pollution in the direct upstream city, lower will be the WTP of the
people living in the direct downstream, all else being equal. This WTP diminishing effect is
also found to be adjustable by distance between the direct up and downstream cities and by
the relative negotiation power of the downstream city with respect to its direct upstream peer.
These findings reveal another pertinent dimension of spatial factors, which is the interrelationship between regions and their impact on the benefits that each region can enjoy
from environmental quality improvement.
Although we can consider our results as an indirect proof about the existence of the
negative externality caused by transboundary river pollution spillover for the case of China,
we admit that it is still necessary to provide direct transboundary water pollution evidence
for the case of China, especially for the discussion of the necessity about establishing a
river-basin level water management system in China. Previous similar studies are Sigman
(2002, 2005) and unpublished working paper of Lipscomb and Mobarak (2007). To conduct
such a study, detailed and reliable water quality data from the monitoring station along the
Xijiang River and their location with respect to different the regional boundaries will be
necessary.
Following the logic of a cost-benefit analysis, our result of the net benefit gain via an
integrate basin level water quality management system can also be compared with the
potential economic cost/loss that different regions/cities along the river need to bear to
respect the uniform water quality targets. The existing theoretical analysis as Van der Laan
and Moes (2012) and Ni and Wang (2002) seem to confirm a net gain in social benefit.
Although such a comparison is not possible for the current study since we do not have
information about the related economic cost/loss of different regions along the river, the
provision of the information about the potential gain in our paper offers such a possibility of
comparison for future studies.
Reference
Bateman, I.J.(2009. Bringing the real world into economic analyses of land use value:
Incorporating spatial complexity. Land Use Policy 26, S30–S42.
Bateman, I.J.(Day, B.H.(Georgiou, S.(Lake, I.(2006. The aggregation of environmental
benefit values: Welfare measures, distance decay and total WTP. Ecological Economics 60,
450–460.
Bateman, I.J.(Langford, I.H.(1997. Non-users' Willingness to Pay for a National Park: An
Application and Critique of the Contingent Valuation Method. Regional Studies 31, 571–
582.
Boadu, F.O.(1992. CONTINGENT VALUATION FOR HOUSEHOLD WATER IN
RURAL GHANA. Journal of Agricultural Economics 43, 458–465.
Bockstael, N.E.(1996. Modeling Economics and Ecology: The Importance of a Spatial
Perspective. American Journal of Agricultural Economics 78, 1168–1180.
Broch, S.W.(Strange, N.(Jacobsen, J.B.(Wilson, K.A.(2012. Farmers' willingness to
provide ecosystem services and effects of their spatial distribution. Ecological Economics.
Brouwer, R.(Martin-Ortega, J.(Berbel, J.(2010. Spatial preference heterogeneity: a choice
experiment. Land Economics 86, 552–568.
Campbell, D.(Hutchinson, W.(Scarpa, R.(2009. Using choice experiments to explore the
spatial distribution of willingness to pay for rural landscape improvements. Environment
and Planning A 41, 97–111.
Campbell, D.(Scarpa, R.(Hutchinson, W.G.(2008. Assessing the spatial dependence of
welfare estimates obtained from discrete choice experiments. Lett Spat Resour Sci 1, 117–
126.
Choe, K.(Whittington, D.(Lauria, D.T.(1996. The Economic Benefits of Surface Water
Quality Improvements in Developing Countries: A Case Study of Davao, Philippines. Land
Economics 72, 519–537.
Desvousges, W.H.(Smith, V.K.(Fisher, A.(1987. Option price estimates for water quality
improvements: A contingent valuation study for the monongahela river. Journal of
Environmental Economics and Management 14, 248–267.
Guimarães, M.H.(Sousa, C.(Garcia, T.(Dentinho, T.(Boski, T.(2011. The value of
improved water quality in Guadiana estuary—a transborder application of contingent
valuation methodology. Lett Spat Resour Sci 4, 31–48–48.
Hammitt, J.K.(Liu, J.-T.(Liu, J.-L.(2001. Contingent valuation of a Taiwanese wetland.
Envir. Dev. Econ. 6.
Hanemann, M.(Loomis, J.(Kanninen, B.J.(1991. Statistical efficiency of double-bounded
dichotomous choice contingent valuation. American Journal of Agricultural Economics
1255–1263.
Holmes, T.P.(Bergstrom, J.C.(Huszar, E.(Kask, S.B.(Orr, F.(III, 2004. Contingent
valuation, net marginal benefits, and the scale of riparian ecosystem restoration. Ecological
Economics 49, 19–30.
Johnston, R.J.(Swallow, S.K.(Bauer, D.M.(2002. Spatial factors and stated preference
values for public goods: considerations for rural land use. Land Economics 78, 481–500.
Jørgensen, S.L.(Olsen, S.B.(Ladenburg, J.(Martinsen, L.(Svenningsen, S.R.(Hasler,
B.(2013). Spatially induced disparities in users“ and non-users” WTP for water quality
improvements—Testing the effect of multiple substitutes and distance decay. Ecological
Economics, 92(2013): 58-66.
Kanbur, R.(Zhang, X.(1999. Which regional inequality? The evolution of rural–urban and
inland–coastal inequality in China from 1983 to 1995. Journal of Comparative Economics
27, 686–701.
Kanbur, R.(Zhang, X.(2005. Fifty years of regional inequality in China: a journey through
central planning, reform, and openness. Review of Development Economics 9, 87–106.
Menz, T.(2011. Do people habituate to air pollution? Evidence from international life
satisfaction data. Ecological Economics 71, 211–219.
Moreno-Sanchez, R. Maldonado, J.H. Wunder, S. Borda-Almanza, C.(2012.
Heterogeneous users and willingness to pay in an ongoing payment for watershed protection
initiative in the Colombian Andes. Ecological Economics 75, 126–134.
Pate, J.(Loomis, J.(1997. The effect of distance on willingness to pay values: a case study
of wetlands and salmon in California. Ecological Economics 20, 199–207.
Sutherland, R.J.(Walsh, R.G.(1985. Effect of Distance on the Preservation Value of Water
Quality. Land Economics 61, 281–291.
Tong, C.(Feagin, R.A.(Lu, J.(Zhang, X.(Zhu, X.(Wang, W.(He, W.(2007. Ecosystem
service values and restoration in the urban Sanyang wetland of Wenzhou, China. Ecological
Engineering 29, 249–258.
Van der Laan, G.(Moes, N.(2012. Transboundary externalities and property rights: An
international river pollution model. Working papers.
Vásquez, W.F.(Mozumder, P.(Hernández-Arce, J.(Berrens, R.P.(2009. Willingness to pay
for safe drinking water: Evidence from Parral, Mexico. Journal of Environmental
Management 90, 3391–3400.
Wei, Y.(Davidson, B.(Chen, D.(White, R.(Li, B.(Zhang, J.(2006. Can Contingent
Valuation be Used to Measure the in Situ Value of Groundwater on the North China Plain?
Water Resour Manage 21, 1735–1749.
Yang, D.T.(2002. What has caused regional inequality in China? China Economic Review
13, 331–334.
Yang, W.(Chang, J.(Xu, B.(Peng, C.(Ge, Y.(2008. Ecosystem service value assessment
for constructed wetlands: A case study in Hangzhou, China. Ecological Economics 68, 116–
125.
Zeidner, M.(Shechter, M.(1988. Psychological responses to air pollution: Some
personality and demographic correlates. Journal of Environmental Psychology 8, 191–208.
Zhang, X.(Kanbur, R.(2005. Spatial inequality in education and health care in China.
China Economic Review 16, 189–204.
Zhiqiang, Z.(Zhongmin, X.(Guodong, C.(Zhiyong, S.(2002. Contingent Valuation of the
Economic Benefits of Restoring Ecosystem Services of Zhangye Prefecture of Heihe River
Basin. Acta Ecologica Sinica 22, 885–893.
Zhongmin, X.(Guodong, C.(Zhiqiang, Z.(Zhiyong, S.(Loomis, J.(2003. Applying
contingent valuation in China to measure the total economic value of restoring ecosystem
services in Ejina region. Ecological Economics 44, 345–358.
Hills P. and P. Roverts (2011). Political Inergration, Transboudary Pollution and
Sustainability: Challenges for Environmental Policy in the Pearl River Delta Region. Journal
of Environmental Planning and Management. 44:4, 455-473.
Hills, P., L. Zhang, J. Liu (1998). Transboundary Pollution between Guangdong province
and Hong Kong: Threats to Water in the earl River Estuary and Their Implications for
Environmental Policy and Planning. Journal of Environmental Planning and Management.
41: 3, 375-396.
Zhao, L., Y. Qian, R. Huang, C. Li, J. Xue and Y. Hu (2012). Model of transfer tax on
transoundary water pollution in China's river basin. Operations Research Letter, 40(2012):
218-222.
Wang, S. (1999). Analysis and managemnet of industry pollution accident. Environment
Protect 19(9): 5-6. (in Chinese)
Zhao, L., X. Li, D.L. Zhu, H. Z. Li (2005). Study on model of macroscpical regulation
with emissions trading on transboudary water pollution disputes of river basin. Journal of
Systems Engineering. 20(4): 398-403. (in Chinese)
Yu, X. (2011). Transboundary water pollution management lessons learned from river
basin management in China, Europe and the Netherlands. Utrecht Law Review. 7(1):
188203.
Zhao, L. C. Li, R. Huang, S. Si, J. Xue, W. Huang and Y. Hu (2013). Harmonizing model
with transfer tax on water pollution across regional boundaries in a China's lake basin.
European Journal of Operational Research. 225(2013): 377-382.
Zhao, L. (2009). Model of collective coopration and reallocation of benefits related to
conflicts over water pollutoin acorss regional boundaries in a Chinese river basin.
Environmental Modelling & Software. 24(2009): 603-610.
Schaafsma, M., R. Brouwer, A. Gilbert, J. van den Bergh and A. Wagtendonk (2013).
Estimation of Distance-Decay Functions to account for Substitution and Spatial
Heterogeneity in Stated Preference Research. Land Economics. 89(3): 514-537.
Schaafsma, M. R. Brower and J. Rose. Directional heterogeneity in WTP models for
environmental valuation. Ecological Economics. 79(2012): 21-31.
Rolfe, J. and J. Windle (2010). Do values for protecting iconic assets vary across
populations? A Great Barrier Reef case study. Environmental Ecnomics Research Hub
Research report no. 65. ISSN 835-9728. May 2010.
Sutherland, R. J. and R. G. Walsh (1985). Effect of distance on the preservation value of
water quality. land Economics. 61(3): 281-291.
Hanley, N., F. Schläpfer and J. Spurgeon. (2003). Aggregating the benefits of
environmental improvements: distance-decay functions for use and non-use values. Journal
of Environmental Management, 68(2003): 297-304.
Pate, J. and J. Loomis (1997). The effect of distance on willingness to pay values: a case
study of wetland and salmon in California. Ecological Economics, 10(1997): 199-207.
Helland, E. and A. B. Whitford. (2003). Pollution incidence and political jurisdiction:
evidence from the TRI. Journal of Environmental Economics and management. 46(2003):
403-424.
Gray, W. B. and R. J. Shadbegian. (2004). 'Optimal' pollution abatement--whose benefits
matter, and how much. Journal of Environmental Economics and Management. 47(2004):
510-534.
Sigman, H. (2005). Transboundary spillovers and decentralization of environmental
policies. Journal of Environmental Economics and Management. 50(2005): 82-101.
Lipscomb, M. and A. M. Mobarak (2007). Decentralization and water pollution spillovers:
evidence from the redrawing of county boundaries in Brazil. International Policy Center
working paper series no. 67, University of Michigan.
Sigman, H. (2002). International spillovers and water quality in rivers: Does countries
free ride? American Economic Review. 92(4): 1152-1159.
Hein, L. K. van Koppen, R. S. de Groot and E. C. van Ierland (2006). Spatial scales,
stakholders and the valuation of ecosystem services. Ecological Economics, 57(2006): 209228.
Lanz, B. and A. Provins (2013). Valuing local environmental amenity with discrete choice
experiments: Spatial scope sensitivity and heterogeneous marginal utility of income.
Environmental and Resource economics , 56(2013): 105-130.
Moore, R., B. Provencher and R. C. Bishop (2011). Valuing a spatially variable
environmental resource: Reducing non-point source pollution in Green Bay, Wisconsin.
Land Economics. 87(1): 45-59.
Broch, S. W., N. Strange, J. B. Jacobsen, K. A. Wilson (2013). Famers' willingness to
provide ecosystem services and effects of their spatial distribution. Ecological Economics,
92(2013): 78-86.
Tait, P., R. Baskaran, R. Cullen and K. Bicknell (2012). Nonmarket valuation of water
quality: Addressing spatially heterogeneous preferences using GIS and a random parameter
logit model. Ecological Economics, 75(2012): 15-21.
Hime, S., I. Bateman, P. Posen and M. Hutchins (2009). A transferable water quality
ladder for conveying use and ecological information within public surveys. CSERGE
working paper EDM, No. 09-01.
Fig 1 Map of Xijiang River Basin
Table 1 Alternative Bids for the CVM survey
B
Bl
Bu
3
0
5
5
3
10
10
5
15
15
10
20
20
15
30
30
20
50
50
30
80
80
50
150
150
80
200
200
150
300
300
200
500
500
300
1000
Note: B is the initial bid(monthly voluntary donate, in Chinese
Yuan);Bl is second bid if response to first bid was “no”; B u is
second bid if response to first bid was “yes”
Table 2. Matrix used in MBDC WTP question
Price (Yuan)
Free
3
5
10
15
20
30
50
80
150
200
300
500
1000
Definitely not
Probably not
Not sure
Probably yes
Definitely yes
Table 3 City Level Sampling Detail
City/region
Guangzhou
Shengzhen
Dongguan
Fuoshan
Guiling
Nanning
Kunming
Qujing
Guiyang
Guigang
Zhaoqing
Liuzhou
Baise
Qiannan
Zhongshan
Wuzhou
Qianxinan
Yunfu
Yuxi
Zhuhai
Total
Population (million)
12.7
10.4
8.2
7.2
7.0
6.7
6.4
5.9
4.3
4.1
3.9
3.8
3.5
3.2
3.1
2.9
2.8
2.4
2.3
1.6
108.7
Sample Size (DC)
132
120
84
72
72
72
72
48
48
48
48
36
36
36
36
36
36
24
24
24
1104
Sample size (MBDC)
140
114
90
79
76
73
71
64
47
45
43
41
38
35
34
31
30
26
25
17
1119
Table 4 Respondents information of MBDC and DC questionnaires
MBDC
Variable
Description
mean
age
The age of respondents
33.27
edu
The years of education the respondents has acquired
14.57
income_level The monthly income of respondents(Yuan)
4101.88
gender
Gender of respondents(1=male,0=female)
0.54
N
1043
***p<0.01,**p<0.05,*p<0.1,Kruskal-Wallis Test's chi-squared Value
DC
mean KW-test
33.96 1.839
13.78 21.579***
4585.44 1.958
0.52 0.459
1090
Table 5 Comparisons of socio-economic statistics of respondent across provinces
Guangdong
Sample
Census
Guangxi
Sample
Census
Gender
Female
49.03%
47.38%
49.60%
48.90%
Male
50.97%
52.62%
50.40%
51.10%
Education years
0
0.90%
1.38%
1.00%
1.56%
6
0.00%
16.38%
0.00%
17.08%
9
8.70%
42.64%
14.50%
35.68%
12
23.30%
24.78%
25.60%
25.41%
16+
67.00%
14.81%
59.00%
20.27%
Age
20-29
45.86%
34.20%
55.87%
27.15%
30-39
29.95%
27.07%
22.50%
25.22%
40-49
13.82%
19.72%
12.81%
21.36%
50-59
8.38%
9.67%
6.14%
13.37%
60+
1.99%
9.17%
2.69%
12.88%
Data source: Tabulation On The 2010 Population Census
Guangxi Province, Yunnan Province
Guizhou
Sample
Census
Yunnan
Sample
Census
42.92%
57.08%
49.10%
50.90%
37.63%
62.37%
49.15%
50.85%
0.90%
0.00%
8.40%
31.90%
58.80%
3.32%
22.56%
34.71%
19.65%
19.75%
0.00%
20.20%
8.00%
29.60%
42.10%
2.99%
22.83%
32.18%
20.90%
21.11%
40.49%
30.94%
20.04%
7.30%
1.34%
Of Guizhou
26.31%
23.48%
24.42%
25.91%
31.27%
25.83%
21.05%
31.27%
22.90%
12.55%
12.59%
12.37%
14.10%
1.49%
14.45%
Province, Guangdong Province,
Table 6. WTP answer data cleaning details
Category Definition
1
2
3
4
5
6
WTP value covered by bid price range from 0 to
1000yuan
Negative response at zero bid
Positive response at highest bid (1000 yuan per
month)
Missing values
Disordered answers
Total number
Number of respondents
DC
MBDC
684
847
119
143
99
37
120
1
1023
63
non-applicable
1090
Table 7 Statistics of No demand
MBDC
Number of
respondents
Reasonable answers
It is not important to improve the water
quality and ecological environment
This kind of donation will put a heavy
burden to my family
Others: Have little chance to get in touch
with the river
Protest answers
I don't believe this hypothetic fund will be
used properly
Others: Government should take this
responsibility
DC
%
Number of
respondents
%
1
0.096
4
0.4
14
1.34
56
5.1
2
0.19
1
0.1
87
8.3
76
6.97
15
1.44
6
0.55
Table 8 Statistics of Extremely High Responses
MBDC
Number of
respondents
Reasonable high demand
I can afford 1000yuan per month for
environmental protection
It is very important to protect water
resource in this town
I want to make some contribution to
environmental protection
Unreliable high demand
I am kidding
Other(1000>one-tens of her or his
monthly income)
DC
%
Number of
respondents
%
5
0.479
0
0
8
0.77
0
0
13
1.25
2
0.183
31
3
9
0.826
42
4.03
26
2.385
Table 9 The distribution of answers in MBDC questionnaires
Bid Price
0
3
5
10
15
20
30
50
80
150
200
300
500
1000
Def yes
92.6%
78.6%
70.8%
56.5%
38.7%
32.5%
26.8%
17.1%
9.2%
3.3%
1.2%
0.3%
0.1%
0.0%
Prob yes
4.8%
14.7%
17.3%
22.2%
23.2%
18.1%
14.7%
15.0%
12.4%
6.2%
2.7%
1.1%
0.3%
0.0%
Not sure
2.5%
1.9%
4.7%
7.5%
13.7%
15.8%
13.0%
13.0%
15.0%
11.1%
7.9%
4.7%
2.4%
0.4%
Prob not
0.0%
1.6%
2.5%
5.2%
6.8%
9.8%
12.0%
11.4%
11.6%
13.1%
13.0%
11.9%
8.2%
4.4%
Def not
0.0%
3.2%
4.7%
8.6%
17.7%
23.8%
33.5%
43.5%
51.7%
66.3%
75.2%
82.1%
89.0%
95.2%
Table 10. Cumulative bid acceptance response rate for the 12 initial bid amounts
(shown in bold type) and corresponding double-bound trees in DC2 WTP question
→
3
5
10
15
20
30
50
80
150
200
300
500
0
3
5
10
15
20
30
50
80
150
200
300
500
1000
95.7%
88.2% 80.7%
80.6% 78.4% 87.4%
64.8% 81.1% 77.5%
70.5% 67.4% 75.9%
58.4% 65.5% 70.8%
51.7% 59.6% 59.1%
40.4% 48.4% 52.2%
33.3% 34.8% 43.3%
19.6% 18.9% 41.8%
13.3% 28.6% 32.3%
13.2% 23.7% 27.3%
14.0% 15.9%
8.0%
First row of this table gives the initial bid prices, and the first column gives the corresponding bid price.
Table 11. Descriptive statistics
Varibles
Description
1. Individual level variables
Environmental perception
Eco_imp
Economics development is more important than Environmental protection(0=no,1=yes)
rep_gov
Government should take the responsibility in environmental protection(0=no,1=yes)
water_problem
Water pollution is very serious in this region (0=no,1=yes)
Willingness to services for environmental protection program:5=definitely yes,4=probably yes,3=not
will_service
sure,2=probably not,1=definitely no
quality_imp
quality of xijiang improved in the last three years
quality_deg
quality of xijiang degraded in the last three years
Confidence of quality impovement:4=Fully achieved,3=have a high probability of being achieved,2=probably
confidence
cannot be achieved,1= cannot be achieved
Socio-economics factors
Income changing with water quality improvement:5=Increasing greatly,4=Increasing,3=no
income_willchange
change,2=decreasing,1=Decreasing greatly
age
age
edu
0=no,6=primary school,9=middle school,12=high school,16=college,19=graduate school+
income_level
Income, the income level has been subsituted with the local average wages of staff and workers(1000 Yuan)
mdaily_2
Respondents' income can only meet the needs of food(1=yes,0=no)
mdaily_3
Respondents' income can can cover the basic neccesity?(1=yes,0=no)
mdaily_4
Respondents' income can meet the needs of their daily life?(1=yes,0=no)
gender
Dummy for Male(0=female,1=male)
d_add2
Near the river but cannot see it directly
d_add3
Far away from the river
d_fish
The frequency of eating freshwater fishes (1=a least once a month,0=other wise)
2. City level macro-variables
share2nd_mean
The share of secondary sector in gross regional output
pop_density
Population density of sampled city
gdp_growth_mean Gdp growth rate of sampled city
3. Spatial related variables
degree
The level of water quality
degree_upper
The level of water quality of the upper stream city
stage
A variable that record how many cities locate a city’s upstream
distance
Distance between sample city and its upper stream city
gdppc_vs
The comparison of gdp per capita between sampled city and its upstream city, which measured as GDPPC of
upstream city/GDPPC of sampled city
pop_vs
The comparison of population between sampled city and its upstream city, which measured as Population of
upstream city/Population of sampled city
gdpgwru_vs
The comparison of gdp growth rate between sampled city and its upstream city, which measured as gdp
growth rate of upstream city/ gdp growth rate of sampled city
dum_ups
This city located on the upstream of Xijiang river(0=no,1=yes)
dum_mds
This city located on the mid-stream of Xijiang river(0=no,1=yes)
Ind_output
Aggregate industrial output of all upstream cities (weithed by distance) Variable 1 for Robustness check
Stage
Number of cities in the upstream part of the river: Variable 2 for Robustness check
N
MBDC
DC2
mean
sd
mean
sd
0.60
0.73
0.60
0.49
0.44
0.49
0.59
0.72
0.67
0.49
0.45
0.47
3.90
0.94
3.93
0.99
0.23
0.36
0.42
0.48
0.23
0.43
0.42
0.50
2.19
1.09
2.06
1.06
2.68
1.33
2.73
1.32
34.09
14.56
4.44
0.07
0.62
0.25
0.51
0.32
0.57
0.79
10.94
3.13
4.10
0.26
0.48
0.43
0.50
0.47
0.50
0.41
34.64
13.73
4.71
0.08
0.62
0.23
0.52
0.31
0.58
0.77
11.03
3.53
4.87
0.26
0.49
0.42
0.50
0.46
0.49
0.42
48.15
17.53
14.06
9.51
13.10
1.62
48.62
16.19
14.02
9.44
12.97
1.67
3.80
2.17
2.57
1.06
0.85
1.51
1.86
2.71
1.64
1.57
3.73
2.00
2.35
1.04
0.81
1.57
1.86
2.63
1.62
1.60
0.53
0.71
0.53
0.76
0.68
0.54
0.63
0.54
0.36
0.18
83.58
2.57
0.48
0.39
146.89
2.71
0.38
0.22
67.59
2.35
0.48
0.42
131.26
2.63
727
910
Table 12.1 Estimation results with spatial factors, MBDC data, Welsh and Poe (1998) approach, “probably yes” model
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
share2nd
pop_density
gdp_growth
degree
H1
-0.016
(36.10)***
-0.310
(3.45)***
0.027
(0.34)
0.169
(4.09)***
0.305
(3.77)***
-0.003
(0.95)
0.002
(0.12)
0.039
(3.62)***
0.371
(3.79)***
0.026
(0.34)
-0.163
(1.22)
-0.191
(1.51)
0.180
(1.89)*
-0.005
(1.21)
-0.022
(5.59)***
0.057
(2.31)**
0.137
(4.11)***
degree_upper
H2
-0.017
(36.10)***
-0.321
(3.57)***
0.083
(1.03)
0.143
(3.42)***
0.309
(3.82)***
-0.001
(0.40)
0.000
(0.02)
0.036
(3.29)***
0.389
(3.96)***
0.025
(0.33)
-0.153
(1.15)
-0.194
(1.53)
0.161
(1.69)*
-0.015
(3.20)***
-0.012
(2.84)***
0.084
(3.30)***
0.142
(4.27)***
-0.123
(4.53)***
H3
-0.017
(36.09)***
-0.314
(3.48)***
0.093
(1.17)
0.139
(3.32)***
0.275
(3.39)***
-0.002
(0.52)
0.002
(0.16)
0.040
(3.65)***
0.378
(3.85)***
0.024
(0.31)
-0.143
(1.07)
-0.175
(1.38)
0.163
(1.71)*
-0.019
(3.75)***
-0.007
(1.59)
0.062
(2.52)**
0.156
(4.65)***
-0.239
(5.30)***
degu_pop
H4
-0.017
(36.10)**
-0.313
(3.48)**
0.034
(0.43)
0.154
(3.69)**
0.296
(3.65)**
-0.003
(0.81)
-0.001
(0.10)
0.039
(3.57)**
0.387
(3.95)**
0.010
(0.14)
-0.148
(1.11)
-0.199
(1.57)
0.167
(1.75)
-0.006
(1.31)
-0.019
(4.75)**
0.044
(1.76)
0.133
(3.99)**
Prob Yes (727 obs)
H4
-0.017
(36.11)**
-0.278
(3.08)**
0.059
(0.75)
0.153
(3.68)**
0.276
(3.40)**
-0.001
(0.17)
0.000
(0.02)
0.039
(3.63)**
0.374
(3.81)**
0.025
(0.32)
-0.089
(0.67)
-0.151
(1.19)
0.199
(2.08)*
-0.003
(0.63)
-0.008
(1.67)
0.064
(2.59)**
0.101
(2.99)**
-0.065
(2.79)**
degu_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.017
(36.10)**
-0.316
(3.51)**
0.052
(0.65)
0.146
(3.48)**
0.293
(3.62)**
-0.003
(0.81)
-0.002
(0.15)
0.039
(3.58)**
0.390
(3.97)**
0.010
(0.13)
-0.154
(1.16)
-0.198
(1.56)
0.153
(1.60)
-0.011
(2.42)*
-0.015
(3.65)**
0.034
(1.34)
0.153
(4.56)**
-0.072
(5.43)**
-0.557
(1.07)
-2202.27
52.68
[48.00 57.48]
0.18
-0.292
(0.56)
-2192.02
20.50
(0.000)***
52.69
[48.56 57.02]
0.16
0.023
(0.04)
-2188.23
28.08
(0.000)***
52.71
[48.60 57.03]
0.16
-0.017
(36.10)**
-2198.39
7.77
(0.01)**
52.67
[49.57 57.43]
0.15
H5
-0.017
(36.09)**
-0.323
(3.58)**
0.092
(1.15)
0.137
(3.26)**
0.295
(3.64)**
-0.002
(0.44)
0.001
(0.08)
0.039
(3.57)**
0.384
(3.91)**
0.017
(0.23)
-0.151
(1.13)
-0.193
(1.52)
0.166
(1.75)
-0.018
(3.75)**
-0.010
(2.24)*
0.072
(2.91)**
0.147
(4.41)**
H5
-0.017
(36.09)**
-0.316
(3.51)**
0.099
(1.23)
0.135
(3.22)**
0.266
(3.28)**
-0.002
(0.62)
0.003
(0.27)
0.043
(3.93)**
0.371
(3.79)**
0.016
(0.21)
-0.146
(1.09)
-0.177
(1.40)
0.176
(1.85)
-0.021
(4.05)**
-0.007
(1.56)
0.051
(2.05)*
0.160
(4.78)**
-0.198
(3.90)**
degu_gdpgwr
Constant
H4
-0.017
(36.09)**
-0.323
(3.58)**
0.092
(1.15)
0.137
(3.26)**
0.295
(3.64)**
-0.002
(0.44)
0.001
(0.08)
0.039
(3.57)**
0.384
(3.91)**
0.017
(0.23)
-0.151
(1.13)
-0.193
(1.52)
0.166
(1.75)
-0.018
(3.75)**
-0.010
(2.24)*
0.072
(2.91)**
0.147
(4.41)**
-0.730
(1.40)
-2187.5
29.55
(0.006 )**
52.7
[49.66 57.37]
0.15
-0.131
(5.39)**
-0.131
(5.39)**
-0.017
(36.09)**
-2187.71
29.12
(0.007)**
52.7
[49.62 57.37]
0.15
0.194
(0.35)
-2194.67
15.21
(0.000)***
52.69
[49.61 57.41]
0.15
0.013
(0.03)
-2184.81
34.92
(0.000)***
52.72
[49.66 57.36]
0.15
-0.211
(5.58)**
0.253
(0.47)
-2186.68
31.19
(0.000)***
52.71
[49.65 57.37]
0.15
Table 12.2 Estimation results with spatial factors, MBDC data, approach Wang and He (2011)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
share2nd
pop_density
gdp_growth
degree
H1
-0.011
(38.13)***
-18.436
(2.19)**
7.080
(0.95)
15.093
(3.87)***
20.245
(2.67)***
-0.271
(0.79)
-0.554
(0.46)
4.019
(3.96)***
27.176
(2.96)***
-5.829
(0.81)
-25.250
(2.02)**
-28.295
(2.38)**
10.770
(1.20)
-0.917
(2.29)**
-1.011
(2.83)***
9.605
(4.16)***
5.134
(1.65)*
degree_upper
H2
-0.011
(38.13)***
-18.666
(2.22)**
9.700
(1.29)
13.791
(3.51)***
20.217
(2.68)***
-0.177
(0.51)
-0.614
(0.51)
3.825
(3.76)***
27.677
(3.03)***
-5.859
(0.81)
-24.755
(1.98)**
-28.315
(2.39)**
9.815
(1.10)
-1.383
(3.06)***
-0.580
(1.43)
10.790
(4.57)***
5.264
(1.69)*
-5.574
(2.22)**
H3
-0.011
(38.13)***
-18.159
(2.17)**
10.930
(1.46)
13.203
(3.37)***
18.151
(2.41)**
-0.178
(0.52)
-0.534
(0.44)
3.988
(3.95)***
27.058
(2.97)***
-5.975
(0.83)
-24.015
(1.93)*
-27.186
(2.30)**
9.611
(1.08)
-1.670
(3.63)***
-0.186
(0.43)
9.831
(4.29)***
5.990
(1.93)*
-13.564
(3.24)***
degu_pop
H4
-0.011
(38.13)**
-18.419
(2.19)*
7.360
(0.99)
14.454
(3.68)**
19.812
(2.62)**
-0.250
(0.73)
-0.674
(0.56)
3.986
(3.93)**
27.725
(3.02)**
-6.473
(0.89)
-24.656
(1.97)*
-28.562
(2.40)*
10.217
(1.14)
-0.933
(2.33)*
-0.899
(2.43)*
9.077
(3.86)**
4.926
(1.58)
Wang &He, 727 obs.
H4
H4
-0.011
-0.011
(38.13)**
(38.13)**
-16.188
-18.709
(1.93)
(2.24)*
9.038
11.104
(1.22)
(1.49)
14.014
12.935
(3.61)**
(3.31)**
17.902
19.221
(2.37)*
(2.56)*
-0.107
-0.154
(0.31)
(0.45)
-0.618
-0.591
(0.51)
(0.49)
3.977
3.928
(3.94)**
(3.90)**
26.846
27.369
(2.95)**
(3.01)**
-5.859
-6.354
(0.82)
(0.89)
-20.947
-24.429
(1.68)
(1.97)*
-25.531
-28.166
(2.16)*
(2.39)*
11.561
9.763
(1.30)
(1.10)
-0.831
-1.710
(2.09)*
(3.74)**
-0.207
-0.285
(0.48)
(0.69)
10.168
10.452
(4.42)**
(4.54)**
2.919
5.525
(0.92)
(1.79)
-2.623
(1.20)
degu_gdppc
-3.485
(3.36)**
Log likelihood
LR
Mean WTP
KR CI
Ratio
-62.418
(1.27)
-4344.03
75.99
[68.89 83.48]
0.19
-49.903
(1.01)
-4341.58
4.09
(0.0268)**
75.99
[69.54 82.95]
0.18
-28.804
(0.58)
-4338.81
10.45
(0.0012)***
75.99
[69.56 82.93]
0.18
-46.849
(0.92)
-4343.31
1.44
(0.230)
75.99
[71.13 83.46]
0.16
H5
-0.011
(38.13)**
-17.117
(2.04)*
10.162
(1.36)
13.821
(3.55)**
17.539
(2.32)*
-0.149
(0.44)
-0.539
(0.45)
3.925
(3.89)**
27.116
(2.98)**
-5.643
(0.78)
-22.824
(1.83)
-26.206
(2.21)*
9.805
(1.10)
-1.362
(3.22)**
-0.033
(0.07)
10.175
(4.42)**
5.080
(1.64)
H5
-0.011
(38.13)**
-18.206
(2.19)*
12.074
(1.62)
12.520
(3.21)**
17.078
(2.27)*
-0.182
(0.54)
-0.440
(0.37)
4.186
(4.17)**
26.535
(2.93)**
-6.558
(0.92)
-23.950
(1.93)
-27.104
(2.30)*
10.231
(1.15)
-1.955
(4.17)**
0.002
(0.00)
9.086
(3.97)**
6.469
(2.09)*
-9.626
(2.03)*
degu_gdpgwr
Constant
H5
-0.011
(38.13)**
-18.459
(2.20)*
8.299
(1.11)
13.867
(3.53)**
19.499
(2.58)**
-0.244
(0.71)
-0.728
(0.60)
3.976
(3.92)**
27.845
(3.04)**
-6.636
(0.92)
-24.793
(1.98)*
-28.521
(2.40)*
9.368
(1.04)
-1.194
(2.83)**
-0.693
(1.78)
8.467
(3.57)**
5.814
(1.86)
-71.629
(1.47)
-4338.43
11.2
(0.000) ***
75.99
[71.16 83.41]
0.16
-7.204
(3.11)**
-7.859
(3.51)**
-27.422
(0.55)
-4337.92
12.22
(0.000)***
75.99
[71.16 83.41]
0.16
-25.586
(0.49)
-4341.97
4.12
(0.000)***
75.99
[71.14 83.45]
0.16
-52.718
(1.08)
-4339.21
9.64
(0.000)**
75.99
[71.15 83.42]
0.16
-14.444
(4.15)**
-6.201
(0.12)
-4335.54
16.99
(0.000)***
75.99
[71.18 83.38]
0.16
Table 12.3 Estimation results with spatial factors, DC2 data, Double-bound methods (Hanemann 1991 approach)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
share2nd
pop_density
gdp_growth
degree
H1
-0.008
(23.33)***
-0.236
(2.34)**
-0.014
(0.15)
0.355
(7.57)***
0.235
(2.55)**
-0.003
(0.65)
-0.009
(0.70)
0.059
(4.98)***
0.202
(1.82)*
-0.060
(0.69)
-0.239
(1.53)
-0.169
(1.17)
0.187
(1.76)*
-0.004
(0.89)
-0.002
(0.48)
-0.019
(0.70)
0.137
(3.85)***
degree_upper
H2
-0.008
(23.39)***
-0.236
(2.35)**
-0.013
(0.14)
0.354
(7.50)***
0.236
(2.55)**
-0.003
(0.64)
-0.009
(0.69)
0.059
(4.98)***
0.202
(1.83)*
-0.060
(0.68)
-0.238
(1.52)
-0.166
(1.12)
0.187
(1.75)*
-0.005
(0.86)
-0.002
(0.38)
-0.019
(0.66)
0.137
(3.85)***
-0.003
(0.10)
H3
-0.008
(23.42)***
-0.239
(2.38)**
-0.008
(0.09)
0.350
(7.38)***
0.233
(2.53)**
-0.003
(0.64)
-0.008
(0.63)
0.060
(5.01)***
0.208
(1.88)*
-0.059
(0.67)
-0.227
(1.45)
-0.141
(0.94)
0.179
(1.67)*
-0.006
(1.13)
0.000
(0.03)
-0.018
(0.66)
0.141
(3.90)***
-0.041
(0.76)
degu_pop
H4
-0.008
(23.34)**
-0.232
(2.29)*
-0.013
(0.14)
0.357
(7.55)**
0.234
(2.54)*
-0.003
(0.62)
-0.009
(0.72)
0.059
(4.97)**
0.201
(1.82)
-0.060
(0.68)
-0.244
(1.56)
-0.176
(1.21)
0.191
(1.78)
-0.004
(0.85)
-0.002
(0.53)
-0.018
(0.67)
0.138
(3.86)**
Double Bound (910 obs.)
H4
H4
-0.008
-0.008
(23.33)**
(23.52)**
-0.236
-0.238
(2.35)*
(2.36)*
-0.012
-0.012
(0.13)
(0.12)
0.351
0.352
(7.41)**
(7.46)**
0.240
0.236
(2.59)**
(2.56)*
-0.003
-0.003
(0.64)
(0.65)
-0.008
-0.009
(0.64)
(0.68)
0.059
0.060
(4.97)**
(5.00)**
0.206
0.204
(1.86)
(1.84)
-0.061
-0.059
(0.70)
(0.67)
-0.229
-0.234
(1.46)
(1.49)
-0.152
-0.156
(1.03)
(1.05)
0.185
0.184
(1.74)
(1.73)
-0.004
-0.005
(0.84)
(0.96)
-0.000
-0.001
(0.06)
(0.25)
-0.018
-0.018
(0.66)
(0.65)
0.135
0.138
(3.74)**
(3.86)**
0.010
(0.41)
degu_gdppc
-0.009
(0.67)
Log likelihood
LR
Mean WTP
KR CI
Ratio
-0.516
(0.89)
-1045.41
103.43
[92.41 114.66]
0.22
-0.513
(0.89)
-1045.40
0.01
(0.9170)
103.42
[93.64 113.89]
0.20
-0.457
(0.79)
-1045.11
0.58
(0.4450)
103.42
[93.65 113.83]
0.20
-0.557
(0.95)
-1045.32
0.17
(0.681)
103.44
[95.09,114.92]
0.19
H5
-0.008
(23.27)**
-0.237
(2.36)*
-0.012
(0.13)
0.352
(7.41)**
0.235
(2.55)*
-0.003
(0.64)
-0.009
(0.66)
0.060
(4.99)**
0.205
(1.85)
-0.060
(0.69)
-0.232
(1.48)
-0.155
(1.04)
0.184
(1.72)
-0.005
(0.96)
-0.001
(0.12)
-0.019
(0.67)
0.138
(3.86)**
H5
-0.008
(23.30)**
-0.240
(2.39)*
-0.007
(0.08)
0.348
(7.35)**
0.229
(2.48)*
-0.003
(0.65)
-0.008
(0.60)
0.060
(5.03)**
0.211
(1.90)
-0.059
(0.67)
-0.225
(1.43)
-0.133
(0.88)
0.177
(1.65)
-0.007
(1.23)
0.001
(0.12)
-0.020
(0.72)
0.142
(3.93)**
-0.028
(0.48)
degu_gdpgwr
Constant
H5
-0.008
(23.52)**
-0.241
(2.38)*
-0.012
(0.13)
0.351
(7.40)**
0.233
(2.53)*
-0.003
(0.66)
-0.009
(0.66)
0.060
(5.00)**
0.204
(1.84)
-0.061
(0.69)
-0.232
(1.48)
-0.157
(1.06)
0.182
(1.70)
-0.005
(0.99)
-0.001
(0.29)
-0.021
(0.76)
0.140
(3.87)**
-0.539
(0.93)
-1045.18
0.45
(0.502)
103.4
[94.93,114.86]
0.19
-0.012
(0.40)
-0.011
(0.38)
-0.488
(0.84)
-1045.33
0.15
(0.701)
103.41
(94.93,114.81)
0.19
-0.445
(0.75)
-1045.29
0.23
(0.634)
103.44
(95.09,114.92)
0.19
-0.510
(0.88)
-1045.33
0.16
(0.692)
103.43
(94.97,114.85)
0.19
-0.044
(0.95)
-0.408
(0.69)
-1044.96
0.9
(0.343)
103.43
(95.01,114.87)
0.19
Table 13 Summary table for the five hypotheses related to the spatial aspect related to
transboundary pollution
Def Yes
+***
Prob
Yes
+***
Not
Sure
+
Wang
&He
+*
Single
Bound
+***
Double
Bound
+***
degree
Degree_upper
+***
-***
+***
-***
+
-*
+*
-**
+***
-
+***
-
H3
degree
Degree_upper_dist
+***
-***
+***
-***
+*
-***
+
-***
+***
-
+***
-
H4
degree
Degree_upper_pop
degree
Degree_upper_gdppc
degree
Degree_upper_gdpgwr
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+
+
-**
+
-**
+
+
-**
+*
-***
+**
+
+**
+**
-
+**
+
+**
+**
-
degree
Degree_upper_pop_dist
degree
Degree_upper_gdppc_dist
degree
Degree_upper_gdpgwr_dist
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+
+
-*
+
-**
+
-*
+
-**
+*
-**
+**
+**
+**
-
+**
+**
+**
-
Variables
degree
H2
H1
H4
H4
H5
H5
H5
Table 14. Robustness check 1: upstream city water quality replaced by the sum of weighted
industrial output of all upstream cities
Def Yes
Prob Yes
Not Sure
Variables
Degree
+***
+***
H2
Degree
Ind_out _upper
+***
-**
H3
Degree
Ind_out _upper_dist
H4
H1
H4
H4
H5
H5
H5
+
Wang
&Jie
+*
Single
Bound
+***
Double
Bound
+***
+***
-***
+*
-***
+*
-***
+***
-*
+***
-
+***
-**
+***
-***
+**
-***
+**
-***
+***
-**
+***
-**
Degree
Ind_out _upper_pop
Degree
Ind_out _upper_gdppc
Degree
Ind_out _upper_gdpgwr
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+
-*
+
-**
+*
-**
+*
-*
+
-**
+*
-**
+**
+
+**
+
+**
+
+**
+
+**
+
+**
+
Degree
Ind_out _upper_pop_dist
Degree
Ind_out _upper_gdppc_dist
Degree
Ind_out _upper_gdpgwr_dist
+**
-**
+**
-***
+**
-**
+**
-**
+**
-**
+**
-**
+*
-**
+*
-***
+*
-**
+*
-**
+*
-**
+*
-**
+**
+**
+**
-
+**
+**
+**
-
Table 15. Robustness check 2 : upstream water quality replaced by the sum of
weighted population of all upstream cities
Def Yes
Prob Yes Not Sure
+
Wang &
He
+*
Single
Bound
+***
Double
Bound
+***
+***
-***
+
-***
+
-***
+***
-**
+***
-
+***
-***
+***
-***
+*
-***
+**
-***
+***
-**
+***
-*
Degree
Tot_pop _upper_pop
Degree
Tot_pop _upper_gdppc
Degree
Tot_pop _upper_gdpgwr
+***
-***
+***
-***
+***
-***
+***
-***
+***
+***
-***
+
-**
+
+
-***
+
-**
+*
+
+
-***
+***
+***
+***
-**
+**
+**
+**
-
Degree
Tot_pop _upper_pop_dist
Degree
Tot_pop _upper_gdppc_dist
Degree
Tot_pop _upper_gdpgwr_dist
+***
-***
+***
-***
+***
-***
+***
-***
+***
-***
+***
-***
+*
-***
+***
-**
+
-***
+**
-***
+***
-***
+***
-***
+***
-**
+***
-*
+***
-**
+**
+**
+**
-*
Variables
Degree
+***
+***
H2
Degree
Tot_pop _upper
+***
-***
H3
Degree
Tot_pop _upper_dist
H4
H1
H4
H4
H5
H5
H5
Table 16. Robustness check 3 : upstream water quality replaced by number of
upstream cities
Def Yes
Prob Yes Not Sure
+
Wang &
He
+*
Single
Bound
+***
Double
Bound
+***
+**
-***
+
-***
+*
-***
+***
-**
+***
-*
+***
-***
+**
-***
+
-***
+***
-***
+***
-***
+***
-**
Degree
stage _upper_pop
Degree
stage _upper_gdppc
Degree
stage _upper_gdpgwr
+**
-**
+**
+**
-***
+**
-**
+**
+**
-**
+
-*
+
+
-**
+**
-**
+**
-***
+**
-***
+**
-*
+**
+**
-*
+**
+**
+**
-
Degree
stage _upper_pop_dist
Degree
stage _upper_gdppc_dist
Degree
stage _upper_gdpgwr_dist
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+**
-**
+
-*
+
-**
+
-**
+**
-***
+**
-***
+**
-***
+**
-**
+**
-*
+**
-**
+**
-*
+**
-*
+**
-*
Variables
Degree
+***
+***
H2
Degree
stage _upper
+***
-***
H3
Degree
stage _upper_dist
H4
H1
H4
H4
H5
H5
H5
Table 17. Estimated Mean WTP in Spatial model and Constrained Model (CNY per month)
Tributary
location
(up to
Def Yes
downstream) City/Region
BaU
Integrate
Prob Yes
BaU
Wang&He
Not Sure
Integrate
BaU
1
Qujing
37.7
37.7
55.2
55.2
70.8
1
Yuxi
38.9
38.9
53.9
53.9
91.3
96.7
Integrate
BaU
70.8
61.5
52.1
52.1
169.9
176.8
Single
Integrate
BaU
Double
Integrate
BaU
Integrate
61.5
133.9
133.9
116.1
116.1
45.1
45.1
179.9
179.9
130.6
130.6
136.4
142.6
128.1
132.2
82.9
85.2
2
Qianxinan
40.8
44.6
2
Guigang
22.7
29.1
51.5
60.7
98.4
110.1
78.1
88.7
101.5
108.6
74.7
78.6
52.4
60.1
83.6
93.4
70.0
78.9
61.1
67
60.3
63.6
3
Wuzhou
31.4
36.8
4
Yunfu
13.2
27.8
19.2
40
25.1
51.6
20.5
44.4
111.3
127.3
91.3
100.1
44.2
62.5
95.2
118.5
75.1
96.2
51.4
65.4
47.2
55
5
Zhaoqing
16.4
29.2
6
Foshan
13.1
21.6
38.2
50.2
81.4
96.7
64.9
78.8
79.3
88.6
53.9
59
36.7
86.4
78.1
141.5
63.6
120.9
139.0
177.2
109.6
130.9
27.8
42.2
29.8
48.1
29.3
45.9
72.1
83.2
69.5
75.6
42.8
63.7
36.3
55.1
108.2
120.7
86.1
93.1
78.1
103.6
64.1
87.1
109.8
125.2
87
95.5
7
Zhongshan
16.9
51.9
8
Guangzhou
16.4
26.5
Zhuhai
21.7
33.2
36.7
53.1
Average
WTP
22.7
36.7
44.5
64.5
9
Mean WTP
Table 18. Estimated Total WTP in Spatial model and Constrained Model (Million CNY per month)
Tributary
(up to
downstream)
1
1
2
2
3
4
5
6
7
8
9
Def Yes
Prob Yes
Not Sure
Wang&Jie
Single
Double
City/Region
BaU
Integrate
BaU
Integrate
BaU
Integrate
BaU
Integrate
BaU
Integrate
BaU
Integrate
Qujing
Yuxi
Qianxinan
Guigang
Wuzhou
Yunfu
Zhaoqing
Foshan
Zhongshan
Guangzhou
Zhuhai
241.28
241.28
353.28
353.28
453.12
453.12
393.6
393.6
856.96
856.96
743.04
743.04
89.47
89.47
123.97
123.97
119.83
119.83
103.73
103.73
413.77
413.77
300.38
300.38
114.24
124.88
255.64
270.76
475.72
495.04
381.92
399.28
358.68
370.16
232.12
238.56
86.26
110.58
195.7
230.66
373.92
418.38
296.78
337.06
385.7
412.68
283.86
298.68
91.06
106.72
151.96
174.29
242.44
270.86
203
228.81
177.19
194.3
174.87
184.44
34.72
53.12
58.72
84.96
68.48
101.92
58.08
88.16
173.12
193.12
137.76
148.96
Total WTP
1112.46
1803.45
2185.04
3167.21
3835.36
5087.54
3145.03
4277.46
5392.77
6147.85
4270.68
4690.39
31.68
66.72
46.08
96
60.24
123.84
49.2
106.56
267.12
305.52
219.12
240.24
63.96
113.88
172.38
243.75
371.28
462.15
292.89
375.18
200.46
255.06
184.08
214.5
94.32
155.52
275.04
361.44
586.08
696.24
467.28
567.36
570.96
637.92
388.08
424.8
214.63
659.13
466.09
1097.28
991.87
1797.05
807.72
1535.43
1765.3
2250.44
1391.92
1662.43
50.84
82.15
86.18
130.82
92.38
149.11
90.83
142.29
223.51
257.92
215.45
234.36
Annex 1 Water Quality Table used in the survey
Annex 2. Detailed data source for River water pollution
Data Source
Cities
Data Description
1. The weekly report of automatic water quality testing system Guigang City, Wuzhou
Transferring from
for main rivers of China
City, Guangzhou City,
Weekly Average to
Zhongshan City, Nanning Monthly Average
City, Guiyang City,
Kunming City
2. Environmental Information system of the environmental
protection bureau of Guangdong Province
Shenzhen City, Gongduan Monthly Average
City, Yunfu City,
Zhaoqing City， Zhuhai
City, Foshan City
3. Monthly report for the water quality of the main rivers in
Guizhou province.
Qiannan City, Qianxinan
City, Guiyang City,
Monthly Average
4. Monthly report for the water quality of the main rivers in
Guangxi province.
Baise City, Liuzhou City
Monthly Average
5. Monthly environmental quality reportof the main rivers in
Yunnan province.
Qujing City, Yuxi City,
Monthly Average
Annex 3. Distribution of results of two-stage approach (Wang & He, 2011)
WTP Mean
Distribution of 
n. obs
727
Sample mean of
: 75.99
Sample Std.
Deviation of :
103.33
WTP Std. Variance
Distribution of 
Sample mean of
: 43.29
Sample Std.
Deviation of :
72.65
Obs
727
Percentile
0
10
20
30
40
50
60
70
80
90
100
Percentile
0
10
20
30
40
50
60
70
80
90
100
Centile
0.1
6.1
11.2
16.0
21.7
34.4
55.5
85.6
129.5
129.5
999.7
Centile
0.3
1.1
2.3
5.6
8.2
12.1
23.2
482
72.6
108.9
567.9
[95% Conf.
0.1
5.9
11.0
15.2
20.7
29.2
55.5
78.2
106.3
175.0
999.7
[95% Conf.
0.3
0.5
1.8
4.7
6.9
10.0
24.7
37.5
61.5
96.3
567.9
Interval]
0.1
8.3
13.2
19.1
24.2
44.3
65.7
92.8
141.5
215.3
999.7
Interval]
0.3
1.4
3.6
6.4
9.1
14.1
30.7
52.1
80.0
127.4
567.9
Annex 4. Cross-table between realization probability and respondent’s location on the river
(stage)
MBDC
0
It has some possibility that
It has some possibility that the
the project can be realized on
Cannot be realize
project cannot be realized on time
time
3.20%
47.60%
31.90%
1
5.50%
31.50%
46.60%
2
7.50%
39.10%
21.30%
3
18.20%
36.40%
27.30%
4
0.00%
60.00%
13.30%
5
0.00%
40.00%
34.30%
6
0.00%
45.70%
42.90%
7
5.00%
51.20%
24.80%
8
0.00%
40.00%
26.70%
stage
stage
0
It will be realized on
time
DC
It has some possibility that
It will be realized on
It has some possibility that the
the project can be realized on
Cannot be realize
time
project cannot be realized on time
time
2.00%
40.90%
33.60%
1
3.00%
36.40%
33.30%
2
1.10%
49.50%
23.40%
3
2.50%
30.00%
27.50%
4
6.70%
53.30%
26.70%
5
0.00%
39.00%
34.10%
6
0.00%
35.90%
48.70%
7
3.90%
28.10%
45.30%
8
0.00%
38.90%
44.40%
Annex 5. Results of other three models
Table A5.1 Estimation results with spatial factors, MBDC data, Definitely Yes model, Welsh and Poe (1998)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
share2nd
pop_density
gdp_growth
degree
H1
-0.025
(34.88)***
-0.226
(2.50)**
0.038
(0.48)
0.165
(3.95)***
0.233
(2.86)***
-0.001
(0.39)
0.011
(0.85)
0.051
(4.63)***
0.300
(3.06)***
-0.042
(0.54)
-0.080
(0.59)
-0.100
(0.78)
0.121
(1.27)
-0.001
(0.23)
-0.027
(7.07)***
-0.013
(0.53)
0.184
(5.46)***
degree_upper
H2
-0.025
(34.90)***
-0.241
(2.67)***
0.117
(1.45)
0.129
(3.07)***
0.240
(2.95)***
0.001
(0.40)
0.009
(0.72)
0.046
(4.21)***
0.326
(3.31)***
-0.043
(0.56)
-0.067
(0.50)
-0.103
(0.80)
0.094
(0.98)
-0.016
(3.22)***
-0.015
(3.38)***
0.023
(0.92)
0.193
(5.70)***
-0.174
(6.36)***
H3
-0.025
(34.89)***
-0.228
(2.53)**
0.109
(1.36)
0.133
(3.15)***
0.199
(2.44)**
0.000
(0.07)
0.012
(0.89)
0.051
(4.69)***
0.305
(3.11)***
-0.044
(0.57)
-0.058
(0.43)
-0.080
(0.63)
0.101
(1.06)
-0.016
(3.12)***
-0.012
(2.63)***
-0.009
(0.38)
0.204
(6.02)***
-0.259
(5.69)***
degu_pop
H4
-0.025
(34.91)**
-0.231
(2.56)*
0.052
(0.65)
0.136
(3.23)**
0.215
(2.64)**
-0.000
(0.09)
0.005
(0.37)
0.050
(4.58)**
0.337
(3.42)**
-0.076
(0.97)
-0.052
(0.39)
-0.117
(0.91)
0.095
(0.99)
-0.002
(0.44)
-0.022
(5.55)**
-0.042
(1.65)
0.177
(5.24)**
Definitely Yes (727 obs.)
H4
H4
-0.025
-0.026
(34.91)**
(34.90)**
-0.191
-0.241
(2.11)*
(2.68)**
0.072
0.123
(0.90)
(1.53)
0.148
0.124
(3.53)**
(2.93)**
0.202
0.220
(2.48)*
(2.71)**
0.002
0.001
(0.42)
(0.29)
0.010
0.011
(0.76)
(0.82)
0.051
0.050
(4.66)**
(4.62)**
0.301
0.317
(3.07)**
(3.22)**
-0.044
-0.054
(0.57)
(0.69)
-0.003
-0.065
(0.02)
(0.48)
-0.056
-0.099
(0.43)
(0.77)
0.139
0.103
(1.46)
(1.07)
0.002
-0.018
(0.39)
(3.73)**
-0.013
-0.012
(2.86)**
(2.81)**
-0.008
0.004
(0.33)
(0.17)
0.147
0.198
(4.28)**
(5.86)**
-0.138
(5.82)**
degu_gdppc
-0.075
(5.56)**
Log likelihood
LR
Mean WTP
KR CI
Ratio
-0.306
(0.58)
-2184.30
27.21
[24.09 30.44]
0.23
0.087
(0.17)
-2164.04
40.53
(0.000)***
27.31
[24.54 30.19]
0.21
0.351
(0.66)
-2168.07
32.46
(0.000)***
27.31
[24.54 30.21]
0.21
0.518
(0.96)
-2167.33
33.94
(0.000)***
27.26
[25.24 30.39]
0.19
H5
-0.025
(34.91)**
-0.206
(2.29)*
0.098
(1.22)
0.142
(3.39)**
0.191
(2.34)*
0.001
(0.25)
0.011
(0.85)
0.050
(4.54)**
0.306
(3.12)**
-0.039
(0.50)
-0.036
(0.27)
-0.066
(0.51)
0.099
(1.04)
-0.010
(2.13)*
-0.008
(1.45)
-0.006
(0.24)
0.190
(5.63)**
H5
-0.025
(34.88)**
-0.230
(2.55)*
0.113
(1.41)
0.129
(3.06)**
0.190
(2.33)*
-0.000
(0.04)
0.013
(1.01)
0.055
(4.98)**
0.298
(3.04)**
-0.052
(0.68)
-0.062
(0.46)
-0.083
(0.65)
0.115
(1.21)
-0.017
(3.38)**
-0.013
(2.70)**
-0.022
(0.89)
0.209
(6.14)**
-0.313
(6.08)**
degu_gdpgwr
Constant
H5
-0.025
(34.90)**
-0.234
(2.59)**
0.076
(0.95)
0.129
(3.07)**
0.214
(2.64)**
-0.001
(0.17)
0.005
(0.42)
0.051
(4.63)**
0.330
(3.35)**
-0.067
(0.87)
-0.069
(0.51)
-0.111
(0.86)
0.078
(0.81)
-0.010
(2.24)*
-0.018
(4.25)**
-0.051
(2.00)*
0.211
(6.19)**
-0.298
(0.57)
-2168.81
30.99
(0.000)***
27.26
[25.23 30.4]
0.19
-0.180
(5.70)**
-0.172
(7.01)**
-0.172
(7.01)**
-2159.7
49.2
(0.000)***
27.31
[25.29 30.4]
0.19
0.909
(1.62)
-2165.79
37.03
(0.000)***
27.28
[25.26 30.4]
0.19
-0.001
(0.00)
-2168.02
29.55
(0.000)***
27.3
[25.25 30.41]
0.19
-0.223
(5.84)**
0.579
(1.06)
-2167.21
34.18
(0.006)***
27.31
[25.27 30.42]
0.19
Table A5.2 Estimation results with spatial factors, MBDC2 data, Welsh and Poe (1998) Not sure model.
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
share2nd
pop_density
gdp_growth
degree
H1
-0.009
(36.11)***
-0.126
(1.41)
0.067
(0.86)
0.167
(4.04)***
0.222
(2.76)***
-0.003
(0.73)
-0.003
(0.27)
0.036
(3.28)***
0.291
(2.99)***
-0.054
(0.70)
-0.325
(2.43)**
-0.395
(3.11)***
0.057
(0.60)
-0.011
(2.71)***
-0.009
(2.30)**
0.117
(4.74)***
0.049
(1.49)
degree_upper
H2
-0.009
(36.11)***
-0.128
(1.44)
0.090
(1.12)
0.156
(3.74)***
0.223
(2.77)***
-0.002
(0.50)
-0.004
(0.31)
0.034
(3.12)***
0.297
(3.04)***
-0.054
(0.70)
-0.321
(2.41)**
-0.396
(3.12)***
0.048
(0.51)
-0.016
(3.24)***
-0.005
(1.17)
0.127
(5.03)***
0.051
(1.53)
-0.048
(1.81)*
H3
-0.009
(36.12)***
-0.125
(1.39)
0.103
(1.29)
0.150
(3.60)***
0.205
(2.54)**
-0.002
(0.49)
-0.003
(0.25)
0.035
(3.27)***
0.293
(3.00)***
-0.055
(0.72)
-0.315
(2.36)**
-0.387
(3.05)***
0.046
(0.48)
-0.019
(3.77)***
-0.001
(0.24)
0.120
(4.85)***
0.058
(1.75)*
-0.127
(2.84)***
degu_pop
H4
-0.009
(36.11)**
-0.126
(1.41)
0.070
(0.89)
0.161
(3.85)**
0.218
(2.71)**
-0.002
(0.67)
-0.005
(0.36)
0.035
(3.25)**
0.298
(3.05)**
-0.061
(0.79)
-0.319
(2.39)*
-0.398
(3.13)**
0.051
(0.53)
-0.012
(2.75)**
-0.008
(1.93)
0.111
(4.44)**
0.047
(1.43)
Not Sure (727 obs)
H4
-0.009
(36.12)**
-0.107
(1.19)
0.084
(1.06)
0.158
(3.81)**
0.206
(2.55)*
-0.001
(0.33)
-0.004
(0.32)
0.035
(3.26)**
0.290
(2.97)**
-0.055
(0.72)
-0.288
(2.15)*
-0.375
(2.95)**
0.064
(0.68)
-0.010
(2.41)*
-0.002
(0.36)
0.121
(4.88)**
0.030
(0.90)
-0.028
(1.19)
degu_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.009
(36.11)**
-0.127
(1.42)
0.079
(1.00)
0.155
(3.71)**
0.216
(2.68)**
-0.002
(0.66)
-0.005
(0.40)
0.035
(3.25)**
0.299
(3.06)**
-0.062
(0.81)
-0.321
(2.40)*
-0.398
(3.14)**
0.043
(0.45)
-0.014
(3.18)**
-0.006
(1.36)
0.106
(4.18)**
0.056
(1.70)
-0.036
(2.76)**
-0.749
(1.44)
-2312.00
92.56
[84.07 101.28]
0.19
-0.643
(1.23)
-2310.37
3.26
(0.0710)*
92.56
[84.95 100.55]
0.17
-0.441
(0.83)
-2307.98
8.04
(0.0046)***
92.57
[84.98 100.53]
0.17
-0.586
(1.09)
-2311.29
1.42
(0.233)
92.56
[86.89 101.18]
0.15
H5
-0.009
(36.12)**
-0.115
(1.29)
0.094
(1.18)
0.156
(3.76)**
0.203
(2.51)*
-0.002
(0.43)
-0.004
(0.27)
0.035
(3.20)**
0.293
(3.00)**
-0.053
(0.69)
-0.306
(2.29)*
-0.381
(3.00)**
0.046
(0.49)
-0.015
(3.40)**
0.000
(0.03)
0.121
(4.90)**
0.051
(1.54)
H5
-0.009
(36.11)**
-0.126
(1.41)
0.117
(1.47)
0.143
(3.41)**
0.194
(2.40)*
-0.002
(0.49)
-0.002
(0.18)
0.038
(3.47)**
0.289
(2.96)**
-0.062
(0.80)
-0.315
(2.36)*
-0.388
(3.05)**
0.051
(0.54)
-0.022
(4.38)**
0.001
(0.30)
0.113
(4.58)**
0.064
(1.92)
-0.095
(1.89)
degu_gdpgwr
Constant
H4
-0.009
(36.11)**
-0.130
(1.45)
0.107
(1.34)
0.147
(3.51)**
0.214
(2.66)**
-0.002
(0.41)
-0.004
(0.30)
0.035
(3.22)**
0.296
(3.04)**
-0.059
(0.77)
-0.319
(2.39)*
-0.397
(3.12)**
0.046
(0.49)
-0.019
(3.96)**
-0.002
(0.37)
0.126
(5.08)**
0.054
(1.63)
-0.833
(1.60)
-2308.18
7.64
(0.004) ***
92.57
[86.92 101.15]
0.15
-0.078
(2.53)*
-0.078
(3.24)**
-0.409
(0.77)
-2306.75
10.5
(0.001)***
92.57
[86.9 101.13]
0.15
-0.386
(0.70)
-2310.21
3.59
(0.058)*
92.56
[86.89 101.17]
0.15
-0.658
(1.27)
-2308.81
6.39
(0.001) ***
92.58
[86.9 101.15]
0.15
-0.146
(3.89)**
-0.190
(0.35)
-2304.43
15.13
(0.006)***
92.57
[86.93 101.1]
0.15
Table A5.3 Estimation results with spatial factors, DC2 data, Single bound, simple probit model
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
share2nd
pop_density
gdp_growth
degree
H1
-0.005
(13.01)***
-0.265
(2.37)**
0.087
(0.84)
0.352
(6.69)***
0.370
(3.60)***
-0.002
(0.40)
0.000
(0.02)
0.055
(4.09)***
0.187
(1.49)
0.031
(0.32)
-0.079
(0.45)
-0.020
(0.12)
0.174
(1.47)
-0.003
(0.58)
-0.002
(0.38)
-0.017
(0.54)
0.137
(3.45)***
degree_upper
H2
-0.005
(13.01)***
-0.266
(2.38)**
0.088
(0.85)
0.350
(6.62)***
0.372
(3.60)***
-0.002
(0.40)
0.000
(0.02)
0.055
(4.09)***
0.188
(1.50)
0.032
(0.33)
-0.076
(0.43)
-0.015
(0.09)
0.173
(1.46)
-0.004
(0.60)
-0.002
(0.28)
-0.015
(0.49)
0.138
(3.45)***
-0.005
(0.15)
H3
-0.005
(13.02)***
-0.271
(2.42)**
0.095
(0.92)
0.344
(6.48)***
0.369
(3.59)***
-0.002
(0.39)
0.001
(0.08)
0.056
(4.13)***
0.196
(1.56)
0.032
(0.33)
-0.063
(0.36)
0.014
(0.08)
0.163
(1.37)
-0.006
(0.91)
0.001
(0.16)
-0.015
(0.50)
0.142
(3.52)***
-0.052
(0.87)
degu_pop
H4
-0.005
(13.00)**
-0.262
(2.32)*
0.087
(0.84)
0.354
(6.67)**
0.369
(3.59)**
-0.002
(0.38)
0.000
(0.00)
0.055
(4.08)**
0.185
(1.48)
0.031
(0.32)
-0.083
(0.47)
-0.026
(0.16)
0.177
(1.49)
-0.003
(0.57)
-0.002
(0.43)
-0.016
(0.50)
0.138
(3.46)**
Single Bound (910 obs.)
H4
-0.005
(13.01)**
-0.269
(2.40)*
0.091
(0.88)
0.343
(6.47)**
0.381
(3.69)**
-0.002
(0.37)
0.002
(0.11)
0.055
(4.08)**
0.197
(1.57)
0.029
(0.30)
-0.056
(0.32)
0.014
(0.08)
0.168
(1.41)
-0.003
(0.47)
0.002
(0.30)
-0.015
(0.49)
0.132
(3.27)**
0.009
(0.33)
degu_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.005
(13.02)**
-0.273
(2.42)*
0.089
(0.86)
0.346
(6.51)**
0.369
(3.59)**
-0.002
(0.42)
0.001
(0.07)
0.055
(4.11)**
0.191
(1.52)
0.030
(0.31)
-0.069
(0.39)
-0.004
(0.02)
0.167
(1.40)
-0.004
(0.73)
-0.001
(0.17)
-0.020
(0.63)
0.141
(3.49)**
-0.018
(1.21)
-1.130
(1.73)*
-457.41
134.23
[115.55 154.06]
0.29
-1.125
(1.72)*
-457.40
0.02
(0.8793)
134.22
[117.46 153.67]
0.27
-1.043
(1.58)
-457.03
0.76
(0.3827)
135.01
[117.53 153.61]
0.27
-1.170
(1.76)
-457.36
0.11
(0.741)
134.23
[120.96 154.98]
0.25
H5
-0.005
(13.01)**
-0.270
(2.41)*
0.091
(0.88)
0.345
(6.48)**
0.371
(3.61)**
-0.002
(0.38)
0.001
(0.08)
0.055
(4.12)**
0.194
(1.55)
0.030
(0.31)
-0.064
(0.36)
0.008
(0.05)
0.165
(1.38)
-0.004
(0.76)
0.001
(0.17)
-0.015
(0.49)
0.139
(3.48)**
H5
-0.005
(13.01)**
-0.272
(2.42)*
0.096
(0.93)
0.343
(6.46)**
0.364
(3.54)**
-0.002
(0.40)
0.002
(0.11)
0.056
(4.15)**
0.199
(1.58)
0.033
(0.34)
-0.061
(0.35)
0.021
(0.12)
0.161
(1.35)
-0.006
(1.00)
0.001
(0.24)
-0.018
(0.57)
0.143
(3.54)**
-0.038
(0.60)
degu_gdpgwr
Constant
H4
-0.005
(13.01)**
-0.268
(2.39)*
0.090
(0.87)
0.349
(6.58)**
0.372
(3.62)**
-0.002
(0.40)
0.000
(0.03)
0.055
(4.10)**
0.190
(1.52)
0.033
(0.34)
-0.072
(0.41)
-0.007
(0.04)
0.171
(1.44)
-0.004
(0.69)
-0.001
(0.17)
-0.015
(0.49)
0.138
(3.46)**
-1.173
(1.79)
-456.68
1.47
(0.225)
134.2
[120.91 154.78]
0.25
-0.025
(0.75)
-0.012
(0.38)
-0.172
(7.01)**
-457.34
0.15
(0.702)
134.22
[120.93 154.83]
0.25
-1.025
(1.51)
-457.24
0.36
(0.551)
134.29
[121.14 155.04]
0.25
-1.109
(1.70)
-457.13
0.57
(0.927)
134.22
[120.94 154.79]
0.25
-0.052
(1.03)
-0.987
(1.48)
-456.88
1.06
(0.302)
134.3
[121.1 154.96]
0.25
Annex 6. Detailed of the model estimation results related to the two robustness check
Tables A6.1 Robustness check: substitute variable: distance weighted aggregate Industrial Output
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.025
(34.88)***
-0.226
(2.50)**
0.038
(0.48)
0.165
(3.95)***
0.233
(2.86)***
-0.001
(0.39)
0.011
(0.85)
0.051
(4.63)***
0.300
(3.06)***
-0.042
(0.54)
-0.080
(0.59)
-0.100
(0.78)
0.121
(1.27)
0.184
(5.46)***
-0.001
(0.23)
-0.027
(7.07)***
-0.013
(0.53)
Total Output
H2
-0.025
(34.87)***
-0.238
(2.64)***
0.087
(1.08)
0.130
(3.07)***
0.225
(2.77)***
-0.002
(0.43)
0.012
(0.90)
0.056
(5.11)***
0.307
(3.12)***
-0.052
(0.67)
-0.068
(0.51)
-0.077
(0.60)
0.151
(1.57)
0.194
(5.75)***
-0.008
(1.79)*
-0.023
(5.73)***
-0.045
(1.76)*
-3.3510-7
(4.60)***
H3
-0.025
(34.86)***
-0.226
(2.51)**
0.079
(0.99)
0.139
(3.27)***
0.205
(2.51)**
-0.002
(0.41)
0.014
(1.08)
0.058
(5.25)***
0.285
(2.91)***
-0.056
(0.73)
-0.067
(0.50)
-0.086
(0.67)
0.165
(1.71)*
0.198
(5.83)***
-0.011
(2.12)**
-0.024
(5.92)***
-0.035
(1.39)
-0.000017
(3.77)***
output_pop
H4
-0.025
(34.91)**
-0.207
(2.30)*
0.110
(1.37)
0.136
(3.24)**
0.236
(2.90)**
0.001
(0.16)
0.009
(0.70)
0.050
(4.55)**
0.288
(2.93)**
-0.042
(0.55)
-0.062
(0.46)
-0.099
(0.78)
0.096
(1.01)
0.226
(6.55)**
-0.012
(2.55)*
-0.002
(0.30)
-0.029
(1.17)
-0.001
(5.89)**
output _gdppc
Def Yes
H4
-0.025
(34.91)**
-0.183
(2.02)*
0.137
(1.69)
0.136
(3.23)**
0.214
(2.63)**
-0.000
(0.07)
0.012
(0.95)
0.051
(4.63)**
0.278
(2.83)**
-0.032
(0.42)
-0.061
(0.46)
-0.052
(0.40)
0.077
(0.80)
0.199
(5.89)**
-0.001
(0.32)
-0.001
(0.25)
0.024
(0.96)
-0.000
(6.01)**
output _gdpgwr
Constant
Log likelihood
LR
Mean WTP
KR CI
Ratio
-0.306
(0.58)
-2184.30
27.21
[24.09 30.44]
0.23
0.586
(1.05)
-2173.72
21.18
(0.000)***
27.27
[24.47 30.21]
0.21
0.439
(0.79)
-2177.21
14.18
(0.0002)***
27.26
[24.44 30.21]
0.21
0.200
(0.38)
-2166.93
34.75
(0.000)***
27.28
[25.28 30.42]
0.19
-1.025
(1.91)
-2166.24
36.13
(0.000)***
27.28
[25.27 30.42]
0.19
H4
-0.025
(34.89)**
-0.209
(2.32)*
0.133
(1.63)
0.132
(3.12)**
0.203
(2.49)*
-0.001
(0.27)
0.014
(1.12)
0.054
(4.97)**
0.282
(2.87)**
-0.045
(0.58)
-0.080
(0.60)
-0.071
(0.55)
0.099
(1.03)
0.217
(6.34)**
-0.011
(2.35)*
-0.006
(1.22)
0.000
(0.01)
H5
-0.025
(34.88)**
-0.220
(2.44)*
0.101
(1.26)
0.133
(3.15)**
0.212
(2.61)**
-0.001
(0.24)
0.012
(0.97)
0.056
(5.11)**
0.283
(2.88)**
-0.054
(0.69)
-0.073
(0.54)
-0.096
(0.75)
0.137
(1.43)
0.223
(6.44)**
-0.015
(3.00)**
-0.014
(2.88)**
-0.036
(1.43)
-0.000
(5.00)**
-0.001
(5.73)**
-0.216
(0.41)
-2167.87
32.87
(0.000)***
27.29
[25.31 30.45]
0.19
0.565
(1.02)
-2171.8
25
(0.000)***
27.28
[25.3 30.44]
0.19
H5
-0.025
(34.88)**
-0.214
(2.37)*
0.121
(1.49)
0.130
(3.08)**
0.199
(2.45)*
-0.001
(0.37)
0.015
(1.14)
0.057
(5.16)**
0.279
(2.84)**
-0.050
(0.65)
-0.078
(0.58)
-0.075
(0.58)
0.123
(1.29)
0.214
(6.24)**
-0.012
(2.52)*
-0.013
(2.73)**
-0.007
(0.30)
-0.001
(5.17)**
0.017
(0.03)
-2170.92
26.77
(0.000)***
27.29
[25.31 30.45]
0.19
H5
-0.025
(34.87)**
-0.218
(2.42)*
0.104
(1.29)
0.134
(3.18)**
0.193
(2.36)*
-0.001
(0.36)
0.015
(1.18)
0.058
(5.26)**
0.278
(2.83)**
-0.055
(0.71)
-0.072
(0.53)
-0.082
(0.64)
0.145
(1.51)
0.209
(6.13)**
-0.013
(2.62)**
-0.016
(3.60)**
-0.022
(0.90)
-0.000
(4.65)**
0.276
(0.51)
-2173.49
21.62
(0.000)***
27.28
[25.3 30.45]
0.19
Tables A6.2 Robustness check: substitute variable: distance weighted aggregate Industrial Output(continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.016
(36.10)***
-0.310
(3.45)***
0.027
(0.34)
0.169
(4.09)***
0.305
(3.77)***
-0.003
(0.95)
0.002
(0.12)
0.039
(3.62)***
0.371
(3.79)***
0.026
(0.34)
-0.163
(1.22)
-0.191
(1.51)
0.180
(1.89)*
0.137
(4.11)***
-0.005
(1.21)
-0.022
(5.59)***
0.057
(2.31)**
Total Output
H2
-0.017
(36.08)***
-0.323
(3.58)***
0.072
(0.90)
0.138
(3.26)***
0.299
(3.69)***
-0.004
(0.98)
0.002
(0.16)
0.044
(4.05)***
0.379
(3.87)***
0.017
(0.22)
-0.151
(1.13)
-0.171
(1.35)
0.208
(2.17)**
0.146
(4.39)***
-0.012
(2.59)***
-0.017
(4.36)***
0.029
(1.13)
-3.0710-7
(4.25)***
H3
-0.017
(36.09)***
-0.312
(3.47)***
0.075
(0.94)
0.139
(3.31)***
0.274
(3.38)***
-0.004
(0.98)
0.005
(0.38)
0.048
(4.37)***
0.356
(3.63)***
0.009
(0.12)
-0.147
(1.10)
-0.177
(1.39)
0.231
(2.41)**
0.153
(4.57)***
-0.016
(3.29)***
-0.017
(4.32)***
0.032
(1.28)
-0.00020
(4.41)***
output_pop
H4
-0.017
(36.09)**
-0.297
(3.29)**
0.084
(1.05)
0.146
(3.50)**
0.307
(3.79)**
-0.002
(0.51)
-0.000
(0.00)
0.038
(3.52)**
0.362
(3.69)**
0.026
(0.33)
-0.149
(1.12)
-0.193
(1.52)
0.161
(1.69)
0.169
(4.98)**
-0.014
(2.96)**
-0.001
(0.19)
0.046
(1.86)
-0.001
(4.67)**
output _gdppc
Prob Yes
H4
-0.017
(36.09)**
-0.270
(2.99)**
0.125
(1.54)
0.141
(3.37)**
0.287
(3.54)**
-0.002
(0.64)
0.003
(0.21)
0.039
(3.58)**
0.351
(3.58)**
0.036
(0.46)
-0.147
(1.10)
-0.148
(1.17)
0.138
(1.45)
0.152
(4.55)**
-0.006
(1.32)
0.004
(0.75)
0.096
(3.77)**
-0.000
(5.94)**
output _gdpgwr
Constant
Log likelihood
LR
Mean WTP
KR CI
Ratio
-0.557
(1.07)
-2202.27
52.68
[48.00 57.48]
0.18
0.239
(0.43)
-2193.23
18.08
(0.0002)***
52.69
[48.56 57.02]
0.16
0.292
(0.53)
-2192.54
19.46
(0.000)***
52.69
[48.56 57.02]
0.16
-0.178
(0.34)
-2191.36
21.83
(0.000)***
52.69
[49.63 57.39]
0.15
-1.288
(2.41)*
-2184.59
35.36
(0.000)***
52.71
[49.69 57.37]
0.15
H4
-0.017
(36.08)**
-0.296
(3.28)**
0.125
(1.54)
0.135
(3.23)**
0.275
(3.39)**
-0.003
(0.83)
0.005
(0.39)
0.043
(3.94)**
0.355
(3.61)**
0.023
(0.30)
-0.165
(1.23)
-0.165
(1.30)
0.159
(1.66)
0.171
(5.06)**
-0.015
(3.34)**
0.000
(0.04)
0.073
(2.94)**
H5
-0.017
(36.09)**
-0.307
(3.41)**
0.088
(1.09)
0.139
(3.30)**
0.287
(3.54)**
-0.003
(0.81)
0.003
(0.23)
0.044
(4.07)**
0.356
(3.63)**
0.015
(0.19)
-0.156
(1.17)
-0.189
(1.49)
0.196
(2.05)*
0.174
(5.09)**
-0.019
(3.70)**
-0.008
(1.74)
0.037
(1.47)
-0.000
(4.82)**
-0.001
(5.92)**
-0.491
(0.94)
-2184.75
35.04
(0.000)***
52.71
[49.68 57.37]
0.15
0.253
(0.46)
-2190.66
23.22
(0.000)***
52.7
[49.63 57.38]
0.15
H5
-0.017
(36.09)**
-0.300
(3.33)**
0.116
(1.44)
0.132
(3.15)**
0.271
(3.33)**
-0.003
(0.94)
0.006
(0.43)
0.046
(4.18)**
0.351
(3.58)**
0.017
(0.22)
-0.162
(1.21)
-0.168
(1.32)
0.183
(1.92)
0.169
(4.99)**
-0.017
(3.57)**
-0.006
(1.24)
0.065
(2.64)**
-0.001
(5.56)**
-0.235
(0.45)
-2190.66
30.99
(0.000)***
52.71
[49.66 57.38]
0.15
H5
-0.017
(36.09)**
-0.304
(3.38)**
0.103
(1.28)
0.135
(3.22)**
0.261
(3.21)**
-0.003
(0.92)
0.006
(0.49)
0.048
(4.36)**
0.349
(3.56)**
0.011
(0.14)
-0.154
(1.15)
-0.174
(1.37)
0.208
(2.18)*
0.166
(4.92)**
-0.019
(3.81)**
-0.009
(1.95)
0.048
(1.96)
-0.000
(5.33)**
0.083
(0.16)
-2188.04
28.46
(0.000)***
52.7
[49.65 57.37]
0.15
Tables A6.3 Robustness check: substitute variable: distance weighted aggregate Industrial Output(continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.009
(36.11)***
-0.126
(1.41)
0.067
(0.86)
0.167
(4.04)***
0.222
(2.76)***
-0.003
(0.73)
-0.003
(0.27)
0.036
(3.28)***
0.291
(2.99)***
-0.054
(0.70)
-0.325
(2.43)**
-0.395
(3.11)***
0.057
(0.60)
0.049
(1.49)
-0.011
(2.71)***
-0.009
(2.30)**
0.117
(4.74)***
Total Output
H2
-0.009
(36.10)***
-0.134
(1.49)
0.105
(1.32)
0.140
(3.34)***
0.216
(2.69)***
-0.003
(0.74)
-0.003
(0.24)
0.039
(3.62)***
0.297
(3.04)***
-0.062
(0.81)
-0.317
(2.37)**
-0.380
(2.99)***
0.078
(0.82)
0.056
(1.69)*
-0.017
(3.75)***
-0.005
(1.29)
0.094
(3.69)***
-2.5410-7
(3.53)***
H3
-0.009
(36.10)***
-0.125
(1.40)
0.125
(1.57)
0.132
(3.15)***
0.186
(2.30)**
-0.003
(0.75)
0.000
(0.03)
0.046
(4.17)***
0.274
(2.80)***
-0.075
(0.98)
-0.312
(2.33)**
-0.384
(3.02)***
0.114
(1.19)
0.067
(2.01)**
-0.024
(4.96)***
-0.003
(0.87)
0.089
(3.55)***
-0.0000231
(5.14)***
output_pop
H4
-0.009
(36.11)**
-0.116
(1.30)
0.097
(1.21)
0.154
(3.71)**
0.222
(2.76)**
-0.002
(0.49)
-0.004
(0.34)
0.035
(3.21)**
0.285
(2.92)**
-0.055
(0.71)
-0.318
(2.38)*
-0.396
(3.12)**
0.045
(0.48)
0.065
(1.93)
-0.016
(3.45)**
0.002
(0.33)
0.111
(4.50)**
-0.000
(2.44)*
output _gdppc
Not Sure
H4
-0.009
(36.12)**
-0.094
(1.05)
0.134
(1.66)
0.147
(3.53)**
0.208
(2.58)**
-0.002
(0.51)
-0.003
(0.22)
0.035
(3.23)**
0.275
(2.82)**
-0.048
(0.63)
-0.315
(2.36)*
-0.367
(2.89)**
0.026
(0.27)
0.057
(1.74)
-0.012
(2.79)**
0.009
(1.54)
0.144
(5.63)**
-0.000
(4.05)**
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.009
(36.11)**
-0.119
(1.34)
0.120
(1.51)
0.140
(3.35)**
0.205
(2.54)*
-0.002
(0.59)
-0.002
(0.19)
0.040
(3.65)**
0.277
(2.84)**
-0.065
(0.85)
-0.322
(2.41)*
-0.397
(3.12)**
0.068
(0.72)
0.080
(2.37)*
-0.023
(4.59)**
0.003
(0.63)
0.100
(4.01)**
-0.000
(4.19)**
-0.749
(1.44)
-2312.00
-0.092
(0.17)
-2305.76
12.48
(0.0004)***
0.235
(0.42)
-2298.76
26.48
(0.000)***
-0.552
(1.05)
-2309.03
5.94
(0.01)**
-1.240
(2.33)*
-2303.8
16.4
(0.000)***
-0.000
(4.75)**
-0.698
(1.34)
-2300.72
22.56
(0.000)***
92.56
[84.07 101.28]
0.19
92.57
[85.02 100.50]
0.17
92.57
[85.09 100.42]
0.17
92.57
[86.89
0.15
101.19]
92.58
[86.96 101.15]
0.15
92.58
[86.98 101.12]
0.15
output _gdpgwr
Constant
H4
-0.009
(36.11)**
-0.109
(1.22)
0.146
(1.81)
0.139
(3.34)**
0.196
(2.43)*
-0.002
(0.62)
-0.001
(0.07)
0.038
(3.52)**
0.275
(2.82)**
-0.057
(0.75)
-0.329
(2.46)*
-0.378
(2.97)**
0.037
(0.39)
0.074
(2.22)*
-0.020
(4.29)**
0.009
(1.67)
0.130
(5.25)**
H5
-0.009
(36.11)**
-0.111
(1.24)
0.151
(1.88)
0.133
(3.16)**
0.189
(2.34)*
-0.003
(0.70)
0.000
(0.00)
0.041
(3.78)**
0.271
(2.77)**
-0.063
(0.83)
-0.328
(2.46)*
-0.378
(2.97)**
0.057
(0.60)
0.077
(2.29)*
-0.022
(4.74)**
0.006
(1.26)
0.125
(5.07)**
-0.001
(5.17)**
H5
-0.009
(36.11)**
-0.115
(1.29)
0.146
(1.82)
0.132
(3.16)**
0.176
(2.18)*
-0.002
(0.69)
0.001
(0.09)
0.044
(4.03)**
0.267
(2.73)**
-0.071
(0.93)
-0.321
(2.40)*
-0.383
(3.01)**
0.082
(0.87)
0.077
(2.31)*
-0.026
(5.15)**
0.004
(1.00)
0.109
(4.43)**
-0.051
(0.09)
-2303.22
17.55
(0.000)***
-0.455
(0.87)
-2298.61
26.77
(0.000)***
-0.000
(5.45)**
-0.102
(0.19)
-2297.14
29.72
(0.000)***
92.57
[86.94 101.12]
0.15
92.58
[86.99 101.09]
0.15
92.58
[87 101.07]
0.15
Tables A6.4 Robustness check: substitute variable: distance weighted aggregate Industrial Output(continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.011
(38.13)***
-18.436
(2.19)**
7.080
(0.95)
15.093
(3.87)***
20.245
(2.67)***
-0.271
(0.79)
-0.554
(0.46)
4.019
(3.96)***
27.176
(2.96)***
-5.829
(0.81)
-25.250
(2.02)**
-28.295
(2.38)**
10.770
(1.20)
5.134
(1.65)*
-0.917
(2.29)**
-1.011
(2.83)***
9.605
(4.16)***
Total Output
H2
-0.011
(38.13)***
-18.993
(2.28)**
10.728
(1.44)
12.360
(3.14)***
19.431
(2.59)***
-0.275
(0.81)
-0.521
(0.43)
4.373
(4.32)***
27.382
(3.01)***
-6.605
(0.92)
-24.270
(1.95)*
-26.521
(2.25)**
12.769
(1.43)
5.708
(1.85)*
-1.433
(3.40)***
-0.646
(1.76)*
7.284
(3.07)***
-0.000025
(3.69)***
H3
-0.011
(38.13)***
-18.096
(2.19)**
12.229
(1.66)*
11.653
(3.00)***
16.507
(2.21)**
-0.275
(0.82)
-0.212
(0.18)
4.911
(4.84)***
25.036
(2.77)***
-7.636
(1.07)
-23.611
(1.92)*
-26.661
(2.28)**
15.865
(1.79)*
6.595
(2.14)**
-2.079
(4.57)***
-0.510
(1.40)
6.921
(2.97)***
-0.002
(5.09)***
output_pop
H4
-0.011
(38.13)**
-17.408
(2.07)*
10.016
(1.33)
13.806
(3.53)**
20.065
(2.66)**
-0.187
(0.55)
-0.643
(0.53)
3.930
(3.88)**
26.356
(2.88)**
-5.890
(0.82)
-24.500
(1.96)*
-28.230
(2.38)*
9.667
(1.08)
6.623
(2.10)*
-1.342
(3.10)**
0.026
(0.05)
9.022
(3.90)**
-0.041
(2.52)*
output _gdppc
Wang and He
H4
-0.011
(38.13)**
-14.975
(1.79)
14.097
(1.88)
12.773
(3.29)**
18.140
(2.42)*
-0.204
(0.60)
-0.450
(0.38)
3.977
(3.97)**
24.972
(2.76)**
-5.283
(0.74)
-24.055
(1.95)
-24.743
(2.10)*
7.806
(0.88)
5.807
(1.89)
-0.937
(2.37)*
0.700
(1.34)
12.381
(5.24)**
-0.022
(4.44)**
output _gdpgwr
Constant
Log likelihood
LR
Mean WTP
KR CI
Ratio
-62.418
(1.27)
-4344.03
75.99
[68.89 83.48]
0.19
2.316
(0.04)
-4337.28
13.51
(0.0002)***
75.99
[68.58 82.91]
0.18
28.273
(0.55)
-4331.31
25.45
(0.000)***
75.99
[68.63 82.86]
0.17
-42.992
(0.87)
-4340.87
6.32
(0.01)**
75.99
[71.14 83.44]
0.16
-112.928
(2.27)*
-4334.29
19.48
(0.000)***
75.99
[71.19 83.37]
0.16
H4
-0.011
(38.13)**
-16.569
(2.00)*
14.470
(1.93)
12.276
(3.16)**
17.406
(2.33)*
-0.230
(0.68)
-0.314
(0.26)
4.206
(4.20)**
25.093
(2.78)**
-6.109
(0.86)
-25.168
(2.04)*
-26.076
(2.22)*
8.868
(1.00)
7.346
(2.37)*
-1.660
(3.92)**
0.642
(1.30)
10.666
(4.67)**
H5
-0.011
(38.13)**
-17.602
(2.11)*
12.009
(1.61)
12.437
(3.19)**
18.314
(2.44)*
-0.223
(0.66)
-0.462
(0.39)
4.370
(4.34)**
25.434
(2.80)**
-6.783
(0.95)
-24.642
(1.99)*
-27.986
(2.38)*
11.760
(1.33)
7.887
(2.50)*
-2.011
(4.24)**
0.083
(0.19)
7.931
(3.42)**
-0.025
(4.18)**
-0.040
(4.80)**
-56.131
(1.16)
-4332.7
22.65
(0.000)***
75.99
[71.2 83.36]
0.16
2.938
(0.06)
-4335.41
17.23
(0.000)***
75.99
[71.18 83.38]
0.16
H5
-0.011
(38.13)**
-16.923
(2.04)*
14.480
(1.94)
11.807
(3.04)**
17.114
(2.29)*
-0.247
(0.73)
-0.276
(0.23)
4.407
(4.40)**
24.789
(2.74)**
-6.570
(0.92)
-25.101
(2.04)*
-26.431
(2.26)*
10.424
(1.18)
7.647
(2.46)*
-1.916
(4.33)**
0.368
(0.82)
10.052
(4.42)**
-0.118
(4.97)**
-32.159
(0.66)
-4331.89
24.29
(0.000)***
75.99
[71.2 83.35]
0.16
H5
-0.011
(38.13)**
-17.113
(2.07)*
14.212
(1.92)
11.624
(3.00)**
15.555
(2.08)*
-0.251
(0.75)
-0.133
(0.11)
4.732
(4.71)**
24.340
(2.70)**
-7.294
(1.03)
-24.316
(1.98)*
-26.468
(2.27)*
13.003
(1.48)
7.531
(2.44)*
-2.182
(4.77)**
0.213
(0.51)
8.699
(3.83)**
-0.016
(5.41)**
-2.057
(0.04)
-4329.68
28.7
(0.000)***
75.99
[71.22 83.33]
0.16
Tables A6.5 Robustness check: substitute variable: distance weighted aggregate Industrial Output (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.005
(13.01)***
-0.265
(2.37)**
0.087
(0.84)
0.352
(6.69)***
0.370
(3.60)***
-0.002
(0.40)
0.000
(0.02)
0.055
(4.09)***
0.187
(1.49)
0.031
(0.32)
-0.079
(0.45)
-0.020
(0.12)
0.174
(1.47)
0.137
(3.45)***
-0.003
(0.58)
-0.002
(0.38)
-0.017
(0.54)
Total Output
H2
-0.005
(13.02)***
-0.280
(2.50)**
0.098
(0.94)
0.338
(6.37)***
0.354
(3.43)***
-0.002
(0.44)
0.004
(0.27)
0.056
(4.14)***
0.206
(1.63)
0.029
(0.30)
-0.052
(0.29)
0.034
(0.20)
0.160
(1.35)
0.141
(3.51)***
-0.006
(0.97)
0.000
(0.04)
-0.033
(1.03)
-1.7210-7
(1.90)*
H3
-0.005
(13.00)***
-0.268
(2.38)**
0.109
(1.05)
0.338
(6.39)***
0.329
(3.16)***
-0.002
(0.41)
0.005
(0.36)
0.058
(4.21)***
0.219
(1.73)*
0.028
(0.29)
-0.045
(0.25)
0.064
(0.38)
0.157
(1.32)
0.154
(3.77)***
-0.011
(1.68)*
0.001
(0.21)
-0.032
(1.01)
-0.000015
(2.47)**
output_pop
H4
-0.005
(13.01)**
-0.265
(2.35)*
0.086
(0.83)
0.352
(6.63)**
0.370
(3.60)**
-0.002
(0.40)
0.000
(0.02)
0.055
(4.09)**
0.186
(1.47)
0.031
(0.32)
-0.080
(0.45)
-0.022
(0.13)
0.175
(1.46)
0.137
(3.35)**
-0.003
(0.53)
-0.002
(0.31)
-0.016
(0.53)
0.000
(0.06)
output _gdppc
Single Bound
H4
-0.005
(13.01)**
-0.261
(2.33)*
0.076
(0.73)
0.358
(6.75)**
0.358
(3.46)**
-0.002
(0.40)
-0.000
(0.03)
0.055
(4.10)**
0.176
(1.40)
0.031
(0.32)
-0.088
(0.50)
-0.050
(0.30)
0.189
(1.58)
0.135
(3.37)**
-0.004
(0.63)
-0.007
(0.93)
-0.023
(0.74)
0.000
(0.90)
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.005
(13.02)**
-0.274
(2.44)*
0.100
(0.97)
0.343
(6.47)**
0.358
(3.47)**
-0.002
(0.38)
0.001
(0.09)
0.056
(4.14)**
0.205
(1.63)
0.033
(0.34)
-0.058
(0.33)
0.027
(0.16)
0.158
(1.32)
0.152
(3.64)**
-0.008
(1.16)
0.003
(0.42)
-0.022
(0.70)
-0.000
(1.28)
-1.130
(1.73)*
-457.41
-0.782
(1.15)
-455.61
3.61
(0.0575)*
-0.664
(0.98)
-454.36
6.11
(0.0135)**
-1.135
(1.73)
-457.41
0
(0.951)
-0.981
(1.46)
-457.01
0.82
(0.366)
0.000
(0.39)
-1.127
(1.73)
-457.34
0.15
(0.698)
134.23
[115.55 154.06]
0.29
134.64
[117.88 154.12]
0.27
134.89
[118.04 154.12]
0.27
134.23
[120.45
0.26
154.76]
134.32
[120.48 154.8]
0.26
134.24
[120.47 154.74]
0.26
output _gdpgwr
Constant
H4
-0.005
(13.01)**
-0.263
(2.34)*
0.083
(0.80)
0.355
(6.68)**
0.369
(3.59)**
-0.002
(0.40)
-0.000
(0.00)
0.055
(4.07)**
0.182
(1.45)
0.030
(0.31)
-0.083
(0.47)
-0.034
(0.21)
0.180
(1.51)
0.135
(3.34)**
-0.003
(0.45)
-0.004
(0.55)
-0.018
(0.57)
H5
-0.005
(13.01)**
-0.267
(2.38)*
0.091
(0.88)
0.349
(6.58)**
0.369
(3.59)**
-0.002
(0.40)
0.001
(0.05)
0.055
(4.10)**
0.191
(1.52)
0.031
(0.32)
-0.074
(0.42)
-0.007
(0.04)
0.169
(1.41)
0.140
(3.45)**
-0.004
(0.68)
-0.000
(0.08)
-0.016
(0.52)
-0.000
(0.36)
H5
-0.005
(13.00)**
-0.268
(2.39)*
0.099
(0.95)
0.344
(6.50)**
0.357
(3.46)**
-0.002
(0.40)
0.002
(0.15)
0.056
(4.15)**
0.203
(1.61)
0.032
(0.33)
-0.064
(0.36)
0.024
(0.14)
0.160
(1.34)
0.147
(3.60)**
-0.007
(1.06)
0.002
(0.28)
-0.018
(0.58)
-0.924
(1.37)
-456.6
1.63
(0.201)
-1.115
(1.71)
-457.35
0.13
(0.718)
-0.000
(1.14)
-1.009
(1.53)
-456.76
1.31
(0.252)
134.48
[121.06 154.95]
0.25
134.26
[120.43 154.74]
0.26
134.39
[120.88 154.89]
0.25
Tables A6.6 Robustness check: substitute variable: distance weighted aggregate Industrial Output(continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.008
(23.33)***
-0.236
(2.34)**
-0.014
(0.15)
0.355
(7.57)***
0.235
(2.55)**
-0.003
(0.65)
-0.009
(0.70)
0.059
(4.98)***
0.202
(1.82)*
-0.060
(0.69)
-0.239
(1.53)
-0.169
(1.17)
0.187
(1.76)*
0.137
(3.85)***
-0.004
(0.89)
-0.002
(0.48)
-0.019
(0.70)
Total Output
H2
-0.008
(23.28)***
-0.245
(2.43)**
-0.009
(0.09)
0.347
(7.32)***
0.224
(2.41)**
-0.003
(0.68)
-0.007
(0.52)
0.060
(5.02)***
0.212
(1.91)*
-0.061
(0.69)
-0.223
(1.42)
-0.132
(0.89)
0.180
(1.68)*
0.139
(3.89)***
-0.006
(1.13)
-0.001
(0.18)
-0.029
(1.02)
-1.0610-7
(1.28)
H3
-0.008
(23.43)***
-0.237
(2.35)**
0.002
(0.03)
0.345
(7.30)***
0.203
(2.17)**
-0.003
(0.66)
-0.005
(0.39)
0.061
(5.09)***
0.223
(2.01)**
-0.063
(0.71)
-0.214
(1.37)
-0.102
(0.68)
0.173
(1.62)
0.151
(4.13)***
-0.010
(1.75)*
0.000
(0.03)
-0.031
(1.09)
-0.000012
(2.05)**
output_pop
H4
-0.008
(23.21)**
-0.234
(2.32)*
-0.015
(0.16)
0.356
(7.52)**
0.235
(2.54)*
-0.003
(0.65)
-0.009
(0.69)
0.059
(4.98)**
0.200
(1.80)
-0.061
(0.69)
-0.241
(1.54)
-0.174
(1.17)
0.189
(1.76)
0.136
(3.71)**
-0.004
(0.79)
-0.003
(0.43)
-0.019
(0.69)
0.000
(0.15)
output _gdppc
Double Bound
H4
-0.008
(23.37)**
-0.234
(2.32)*
-0.019
(0.21)
0.359
(7.58)**
0.229
(2.47)*
-0.003
(0.66)
-0.010
(0.73)
0.060
(4.99)**
0.196
(1.76)
-0.060
(0.69)
-0.245
(1.56)
-0.187
(1.27)
0.196
(1.82)
0.136
(3.79)**
-0.005
(0.91)
-0.005
(0.76)
-0.023
(0.82)
0.000
(0.60)
output _gdpgwr
Constant
Log likelihood
LR
Mean WTP
KR CI
Ratio
-0.516
(0.89)
-1045.41
103.43
[92.41 114.66]
0.22
-0.307
(0.51)
-1044.59
1.64
(0.2008)
103.58
[93.90 113.93]
0.19
-0.172
(0.29)
-1043.30
4.21
(0.0420)**
103.64
[91.97 113.97]
0.19
-0.524
(0.90)
-1045.4
21.83
(0.884)
103.43
[94.95 114.65]
0.19
-0.427
(0.72)
-1045.23
28.59
(0.549)
103.51
[94.82 114.67]
0.19
H4
-0.008
(23.37)**
-0.235
(2.33)*
-0.015
(0.16)
0.356
(7.52)**
0.235
(2.55)*
-0.003
(0.65)
-0.009
(0.71)
0.059
(4.97)**
0.199
(1.80)
-0.061
(0.69)
-0.242
(1.54)
-0.177
(1.19)
0.190
(1.77)
0.136
(3.76)**
-0.004
(0.79)
-0.003
(0.49)
-0.020
(0.72)
H5
-0.008
(23.65)**
-0.241
(2.39)*
-0.004
(0.04)
0.349
(7.38)**
0.225
(2.43)*
-0.003
(0.63)
-0.008
(0.62)
0.060
(5.03)**
0.213
(1.92)
-0.059
(0.67)
-0.224
(1.43)
-0.131
(0.88)
0.175
(1.63)
0.148
(3.98)**
-0.008
(1.31)
0.001
(0.23)
-0.023
(0.83)
-0.000
(1.07)
0.000
(0.21)
-0.513
(0.89)
-1045.38
35.04
(0.833)
103.44
[94.91 114.61]
0.19
-0.369
(0.62)
-1044.83
23.22
(0.285)
103.49
[95.2 114.71]
0.19
H5
-0.008
(23.45)**
-0.237
(2.36)*
-0.009
(0.10)
0.352
(7.45)**
0.233
(2.53)*
-0.003
(0.64)
-0.009
(0.65)
0.060
(5.00)**
0.206
(1.86)
-0.060
(0.68)
-0.234
(1.50)
-0.154
(1.03)
0.181
(1.69)
0.141
(3.85)**
-0.005
(0.99)
-0.001
(0.11)
-0.019
(0.68)
-0.000
(0.43)
-0.502
(0.87)
-1045.31
27.71
(0.665)
103.41
[94.96 114.58]
0.19
H5
-0.008
(23.48)**
-0.237
(2.36)*
-0.004
(0.04)
0.349
(7.38)**
0.223
(2.41)*
-0.003
(0.65)
-0.007
(0.56)
0.060
(5.04)**
0.214
(1.92)
-0.060
(0.68)
-0.226
(1.45)
-0.130
(0.87)
0.175
(1.63)
0.146
(3.98)**
-0.007
(1.30)
0.001
(0.18)
-0.021
(0.74)
-0.000
(1.08)
-0.418
(0.72)
-1044.82
28.46
(0.279)
103.45
[95.13 114.66]
0.19
Tables A6.7 Robustness check: substitute variable: stage
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.025
(34.88)***
-0.226
(2.50)**
0.038
(0.48)
0.165
(3.95)***
0.233
(2.86)***
-0.001
(0.39)
0.011
(0.85)
0.051
(4.63)***
0.300
(3.06)***
-0.042
(0.54)
-0.080
(0.59)
-0.100
(0.78)
0.121
(1.27)
0.184
(5.46)***
-0.001
(0.23)
-0.027
(7.07)***
-0.013
(0.53)
Stage
H2
-0.025
(34.88)***
-0.226
(2.50)**
0.103
(1.28)
0.124
(2.92)***
0.191
(2.35)**
-0.001
(0.36)
0.013
(0.98)
0.058
(5.24)***
0.307
(3.12)***
-0.048
(0.62)
-0.041
(0.30)
-0.039
(0.31)
0.154
(1.61)
0.167
(4.92)***
-0.006
(1.34)
-0.025
(6.28)***
-0.013
(0.53)
-0.083
(5.45)***
H3
-0.025
(34.87)***
-0.222
(2.46)**
0.089
(1.10)
0.133
(3.13)***
0.197
(2.42)**
-0.001
(0.40)
0.013
(0.98)
0.057
(5.19)***
0.297
(3.02)***
-0.047
(0.61)
-0.051
(0.38)
-0.055
(0.43)
0.152
(1.58)
0.181
(5.37)***
-0.007
(1.57)
-0.025
(6.33)***
-0.018
(0.74)
-0.096
(4.50)***
stage_pop
H4
-0.025
(34.89)**
-0.227
(2.52)*
0.059
(0.74)
0.124
(2.92)**
0.227
(2.80)**
-0.001
(0.40)
0.004
(0.34)
0.055
(5.02)**
0.328
(3.33)**
-0.068
(0.88)
-0.047
(0.35)
-0.088
(0.69)
0.139
(1.45)
0.181
(5.35)**
-0.002
(0.53)
-0.029
(7.38)**
-0.060
(2.30)*
Def Yes
H4
-0.025
(34.87)**
-0.257
(2.84)**
0.098
(1.21)
0.123
(2.89)**
0.242
(2.98)**
-0.001
(0.27)
0.010
(0.80)
0.053
(4.83)**
0.313
(3.19)**
-0.047
(0.61)
-0.070
(0.52)
-0.094
(0.74)
0.143
(1.50)
0.177
(5.25)**
-0.009
(2.02)*
-0.029
(7.45)**
-0.005
(0.19)
-0.087
(5.56)**
stage_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.025
(34.89)**
-0.227
(2.52)*
0.064
(0.80)
0.132
(3.11)**
0.237
(2.92)**
-0.002
(0.45)
0.007
(0.50)
0.054
(4.96)**
0.313
(3.19)**
-0.055
(0.71)
-0.064
(0.48)
-0.088
(0.68)
0.137
(1.43)
0.196
(5.80)**
-0.006
(1.27)
-0.028
(7.20)**
-0.049
(1.90)
-0.079
(4.62)**
-0.306
(0.58)
-2184.30
27.21
[24.09 30.44]
0.23
0.179
(0.34)
-2169.14
29.73
(0.000)***
27.30
[24.50 30.22]
0.21
0.186
(0.35)
-2174.19
20.22
(0.000) ***
27.28
[24.47 30.22]
0.21
0.840
(1.50)
-2168.79
31.02
(0.000)***
27.26
[25.26 30.37]
0.19
H5
-0.025
(34.87)**
-0.236
(2.62)**
0.093
(1.16)
0.125
(2.95)**
0.212
(2.60)**
-0.001
(0.39)
0.012
(0.90)
0.057
(5.17)**
0.299
(3.05)**
-0.051
(0.66)
-0.057
(0.43)
-0.071
(0.56)
0.153
(1.60)
0.184
(5.45)**
-0.009
(1.97)*
-0.027
(6.83)**
-0.026
(1.04)
H5
-0.025
(34.87)**
-0.227
(2.52)*
0.091
(1.13)
0.131
(3.10)**
0.195
(2.39)*
-0.002
(0.42)
0.014
(1.07)
0.058
(5.28)**
0.296
(3.02)**
-0.053
(0.68)
-0.057
(0.42)
-0.062
(0.49)
0.160
(1.67)
0.184
(5.46)**
-0.008
(1.85)
-0.024
(6.20)**
-0.029
(1.18)
-0.108
(4.59)**
stage_gdpgwr
Constant
H4
-0.025
(34.87)**
-0.233
(2.59)**
0.104
(1.29)
0.123
(2.90)**
0.189
(2.32)*
-0.001
(0.40)
0.014
(1.09)
0.059
(5.38)**
0.306
(3.11)**
-0.055
(0.71)
-0.048
(0.35)
-0.048
(0.37)
0.163
(1.71)
0.171
(5.06)**
-0.008
(1.72)
-0.024
(6.13)**
-0.029
(1.17)
0.335
(0.62)
-2173.63
21.34
(0.000)***
27.27
[25.25 30.41]
0.19
-0.127
(4.64)**
-0.070
(5.41)**
0.441
(0.82)
-2169.64
29.32
(0.000)***
27.3
[25.32 30.43]
0.19
0.664
(1.18)
-2173.72
21.18
(0.000)***
27.25
[25.24 30.39]
0.19
0.476
(0.87)
-2173.55
21.50
(0.000)***
27.28
[25.28 30.42]
0.19
-0.080
(4.45)**
0.374
(0.69)
-2174.41
19.78
(0.000)***
27.28
[25.29 30.43]
0.19
Tables A6.8 Robustness check: substitute variable: stage (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.016
(36.10)***
-0.310
(3.45)***
0.027
(0.34)
0.169
(4.09)***
0.305
(3.77)***
-0.003
(0.95)
0.002
(0.12)
0.039
(3.62)***
0.371
(3.79)***
0.026
(0.34)
-0.163
(1.22)
-0.191
(1.51)
0.180
(1.89)*
0.137
(4.11)***
-0.005
(1.21)
-0.022
(5.59)***
0.057
(2.31)**
Stage
H2
-0.017
(36.08)***
-0.313
(3.47)***
0.096
(1.20)
0.127
(3.01)***
0.265
(3.26)***
-0.003
(0.91)
0.003
(0.26)
0.047
(4.26)***
0.382
(3.90)***
0.020
(0.26)
-0.121
(0.91)
-0.129
(1.02)
0.217
(2.27)**
0.119
(3.56)***
-0.010
(2.38)**
-0.018
(4.75)***
0.059
(2.40)**
-0.088
(5.78)***
H3
-0.017
(36.08)***
-0.308
(3.43)***
0.086
(1.08)
0.133
(3.16)***
0.267
(3.29)***
-0.003
(0.96)
0.004
(0.28)
0.047
(4.27)***
0.371
(3.79)***
0.020
(0.26)
-0.129
(0.97)
-0.141
(1.11)
0.217
(2.27)**
0.134
(4.03)***
-0.012
(2.73)***
-0.018
(4.74)***
0.053
(2.14)**
-0.111
(5.24)***
stage_pop
H4
-0.017
(36.10)**
-0.312
(3.46)**
0.041
(0.51)
0.143
(3.40)**
0.302
(3.73)**
-0.003
(0.95)
-0.002
(0.18)
0.042
(3.83)**
0.389
(3.96)**
0.010
(0.13)
-0.141
(1.06)
-0.183
(1.44)
0.191
(2.00)*
0.134
(4.03)**
-0.006
(1.39)
-0.022
(5.75)**
0.029
(1.11)
Prob Yes
H4
-0.017
(36.10)**
-0.322
(3.57)**
0.051
(0.64)
0.152
(3.58)**
0.308
(3.80)**
-0.003
(0.90)
0.001
(0.10)
0.040
(3.68)**
0.376
(3.83)**
0.024
(0.32)
-0.158
(1.19)
-0.189
(1.49)
0.188
(1.97)*
0.134
(4.01)**
-0.008
(1.83)
-0.022
(5.71)**
0.060
(2.44)*
-0.053
(3.47)**
stage_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.017
(36.09)**
-0.312
(3.47)**
0.046
(0.58)
0.145
(3.45)**
0.309
(3.82)**
-0.004
(0.99)
-0.002
(0.12)
0.042
(3.84)**
0.382
(3.89)**
0.017
(0.22)
-0.150
(1.13)
-0.182
(1.43)
0.192
(2.01)*
0.145
(4.36)**
-0.008
(1.93)
-0.022
(5.66)**
0.032
(1.23)
-0.031
(1.84)
-0.557
(1.07)
-2202.27
52.68
[48.00 57.48]
0.18
-0.078
(0.15)
-2185.53
33.49
(0.000) ***
52.72
[48.62 56.99]
0.16
-0.016
(0.03)
-2188.56
27.42
(0.000) ***
52.71
[48.60 57.01]
0.16
0.132
(0.24)
-2196.26
12.02
(0.000)***
52.69
[49.58 57.43]
0.15
H5
-0.017
(36.09)**
-0.320
(3.55)**
0.074
(0.93)
0.136
(3.21)**
0.288
(3.56)**
-0.003
(0.95)
0.002
(0.17)
0.045
(4.06)**
0.372
(3.79)**
0.019
(0.24)
-0.143
(1.07)
-0.168
(1.32)
0.207
(2.17)*
0.137
(4.10)**
-0.012
(2.61)**
-0.021
(5.36)**
0.047
(1.90)
H5
-0.017
(36.08)**
-0.314
(3.49)**
0.090
(1.12)
0.131
(3.10)**
0.264
(3.25)**
-0.004
(0.98)
0.005
(0.38)
0.048
(4.40)**
0.370
(3.78)**
0.013
(0.17)
-0.135
(1.01)
-0.149
(1.17)
0.227
(2.37)*
0.138
(4.13)**
-0.014
(3.05)**
-0.018
(4.60)**
0.040
(1.59)
-0.077
(3.34)**
stage_gdpgwr
Constant
H4
-0.017
(36.08)**
-0.320
(3.56)**
0.097
(1.22)
0.125
(2.97)**
0.262
(3.22)**
-0.003
(0.96)
0.005
(0.37)
0.049
(4.41)**
0.381
(3.88)**
0.012
(0.16)
-0.128
(0.96)
-0.138
(1.09)
0.227
(2.37)*
0.124
(3.71)**
-0.012
(2.77)**
-0.018
(4.59)**
0.042
(1.71)
-0.309
(0.58)
-2200.59
3.37
(0.001)***
52.68
[49.51 57.46]
0.15
-0.107
(3.94)**
-0.075
(5.77)**
0.204
(0.38)
-2185.63
33.29
(0.066)**
52.71
[49.67 57.36]
0.15
0.125
(0.22)
-2196.69
11.17
(0.000)***
52.69
[49.58 57.43]
0.15
0.085
(0.16)
-2194.49
15.56
(0.000)***
52.7
[49.6 57.42]
0.15
-0.094
(5.22)**
0.211
(0.39)
-2188.64
27.27
(0.000)***
52.71
[49.65 57.38]
0.15
Tables A6.9 Robustness check: substitute variable: stage (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.009
(36.11)***
-0.126
(1.41)
0.067
(0.86)
0.167
(4.04)***
0.222
(2.76)***
-0.003
(0.73)
-0.003
(0.27)
0.036
(3.28)***
0.291
(2.99)***
-0.054
(0.70)
-0.325
(2.43)**
-0.395
(3.11)***
0.057
(0.60)
0.049
(1.49)
-0.011
(2.71)***
-0.009
(2.30)**
0.117
(4.74)***
Stage
H2
-0.009
(36.11)***
-0.123
(1.38)
0.121
(1.52)
0.133
(3.17)***
0.189
(2.34)**
-0.003
(0.69)
-0.002
(0.17)
0.041
(3.75)***
0.297
(3.04)***
-0.060
(0.78)
-0.295
(2.21)**
-0.350
(2.75)***
0.082
(0.87)
0.034
(1.01)
-0.016
(3.58)***
-0.006
(1.58)
0.119
(4.83)***
-0.068
(4.50)***
H3
-0.009
(36.11)***
-0.121
(1.35)
0.117
(1.47)
0.136
(3.25)***
0.189
(2.34)**
-0.003
(0.73)
-0.002
(0.15)
0.042
(3.80)***
0.289
(2.97)***
-0.060
(0.78)
-0.300
(2.24)**
-0.357
(2.80)***
0.085
(0.89)
0.045
(1.38)
-0.017
(3.89)***
-0.006
(1.54)
0.114
(4.62)***
-0.092
(4.36)***
stage_pop
H4
-0.009
(36.11)**
-0.125
(1.40)
0.076
(0.97)
0.149
(3.54)**
0.220
(2.73)**
-0.003
(0.73)
-0.006
(0.48)
0.037
(3.41)**
0.302
(3.10)**
-0.065
(0.85)
-0.311
(2.33)*
-0.390
(3.07)**
0.063
(0.66)
0.047
(1.42)
-0.012
(2.83)**
-0.009
(2.39)*
0.098
(3.78)**
Not Sure
H4
-0.009
(36.11)**
-0.136
(1.51)
0.088
(1.11)
0.152
(3.58)**
0.225
(2.79)**
-0.003
(0.69)
-0.004
(0.29)
0.036
(3.33)**
0.295
(3.02)**
-0.056
(0.72)
-0.321
(2.41)*
-0.394
(3.10)**
0.063
(0.67)
0.046
(1.40)
-0.014
(3.12)**
-0.009
(2.41)*
0.120
(4.85)**
-0.036
(2.38)*
stage_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.009
(36.11)**
-0.126
(1.41)
0.081
(1.02)
0.150
(3.56)**
0.225
(2.79)**
-0.003
(0.75)
-0.006
(0.44)
0.037
(3.42)**
0.298
(3.05)**
-0.061
(0.80)
-0.317
(2.37)*
-0.389
(3.06)**
0.064
(0.67)
0.055
(1.66)
-0.014
(3.18)**
-0.009
(2.33)*
0.099
(3.84)**
-0.027
(1.62)
-0.749
(1.44)
-2312.00
92.56
[84.07 101.28]
0.19
-0.379
(0.72)
-2301.86
20.28
(0.000)***
92.58
[85.07 100.48]
0.17
-0.303
(0.57)
-2302.51
18.99
(0.000)***
92.58
[85.06 100.48]
0.17
-0.279
(0.50)
-2309.18
5.65
(0.018)**
92.56
[86.9 101.16]
0.15
H5
-0.009
(36.11)**
-0.132
(1.48)
0.112
(1.40)
0.135
(3.21)**
0.206
(2.56)*
-0.003
(0.73)
-0.003
(0.23)
0.040
(3.68)**
0.291
(2.98)**
-0.061
(0.80)
-0.309
(2.31)*
-0.376
(2.96)**
0.081
(0.85)
0.048
(1.45)
-0.018
(3.90)**
-0.008
(2.04)*
0.108
(4.37)**
H5
-0.009
(36.11)**
-0.126
(1.41)
0.128
(1.60)
0.130
(3.10)**
0.182
(2.25)*
-0.003
(0.75)
-0.000
(0.03)
0.044
(4.00)**
0.289
(2.96)**
-0.067
(0.88)
-0.303
(2.27)*
-0.360
(2.83)**
0.098
(1.03)
0.048
(1.46)
-0.020
(4.35)**
-0.005
(1.32)
0.101
(4.08)**
-0.055
(2.41)*
stage_gdpgwr
Constant
H4
-0.009
(36.11)**
-0.130
(1.46)
0.130
(1.63)
0.128
(3.04)**
0.182
(2.25)*
-0.003
(0.73)
-0.001
(0.06)
0.043
(3.96)**
0.297
(3.04)**
-0.067
(0.88)
-0.298
(2.23)*
-0.353
(2.77)**
0.095
(0.99)
0.036
(1.08)
-0.018
(4.02)**
-0.005
(1.36)
0.105
(4.24)**
-0.531
(0.99)
-2310.69
2.61
(0.106)
92.56
[86.83 101.18]
0.16
-0.100
(3.70)**
-0.066
(5.09)**
-0.083
(0.16)
-2299.03
25.95
(0.000)***
92.58
[87.01 101.08]
0.15
-0.260
(0.47)
-2309.10
5.80
(0.000)***
92.57
[86.9 101.17]
0.15
-0.150
(0.28)
-2305.16
13.68
(0.000)***
92.57
[86.93 101.13]
0.15
-0.089
(4.96)**
-0.025
(0.05)
-2299.70
24.60
(0.000)***
92.58
[86.99 101.09]
0.15
Tables A6.10 Robustness check: substitute variable: stage (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.005
(13.01)***
-0.265
(2.37)**
0.087
(0.84)
0.352
(6.69)***
0.370
(3.60)***
-0.002
(0.40)
0.000
(0.02)
0.055
(4.09)***
0.187
(1.49)
0.031
(0.32)
-0.079
(0.45)
-0.020
(0.12)
0.174
(1.47)
0.137
(3.45)***
-0.003
(0.58)
-0.002
(0.38)
-0.017
(0.54)
Stage
H2
-0.005
(13.03)***
-0.275
(2.45)**
0.104
(1.00)
0.333
(6.25)***
0.356
(3.45)***
-0.002
(0.47)
0.008
(0.50)
0.056
(4.16)***
0.209
(1.66)*
0.022
(0.23)
-0.044
(0.25)
0.061
(0.37)
0.154
(1.29)
0.129
(3.20)***
-0.005
(0.82)
-0.001
(0.12)
-0.018
(0.60)
-0.046
(2.34)**
H3
-0.005
(13.05)***
-0.275
(2.45)**
0.110
(1.06)
0.332
(6.23)***
0.342
(3.31)***
-0.002
(0.44)
0.008
(0.56)
0.057
(4.19)***
0.215
(1.70)*
0.019
(0.20)
-0.040
(0.23)
0.072
(0.43)
0.151
(1.27)
0.139
(3.45)***
-0.007
(1.15)
-0.000
(0.04)
-0.021
(0.68)
-0.077
(2.70)***
stage_pop
H4
-0.005
(13.07)**
-0.287
(2.55)*
0.093
(0.90)
0.336
(6.32)**
0.362
(3.51)**
-0.002
(0.49)
0.005
(0.31)
0.055
(4.12)**
0.200
(1.59)
0.020
(0.21)
-0.042
(0.24)
0.027
(0.16)
0.159
(1.34)
0.139
(3.45)**
-0.004
(0.67)
-0.003
(0.60)
-0.034
(1.05)
Single Bound
H4
-0.005
(13.01)**
-0.268
(2.40)*
0.097
(0.93)
0.348
(6.61)**
0.373
(3.63)**
-0.002
(0.42)
0.001
(0.06)
0.055
(4.08)**
0.195
(1.55)
0.034
(0.35)
-0.072
(0.41)
-0.002
(0.01)
0.172
(1.45)
0.135
(3.36)**
-0.005
(0.82)
-0.002
(0.47)
-0.013
(0.43)
-0.038
(2.05)*
stage_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.005
(13.08)**
-0.291
(2.58)**
0.106
(1.02)
0.333
(6.28)**
0.353
(3.41)**
-0.002
(0.45)
0.005
(0.37)
0.056
(4.12)**
0.206
(1.63)
0.018
(0.18)
-0.033
(0.19)
0.047
(0.29)
0.158
(1.33)
0.150
(3.68)**
-0.006
(1.04)
-0.003
(0.55)
-0.040
(1.24)
-0.019
(0.81)
-1.130
(1.73)*
-457.41
134.23
[115.55
0.29
154.06]
-0.976
(1.49)
-454.66
5.50
(0.0190)**
134.53
[117.82
0.27
153.62]
-0.905
(1.37)
-453.75
7.33
(0.0068)***
134.72
[118.01 153.94]
0.27
-0.773
(1.14)
-455.29
4.25
(0.039)**
134.68
[121.8
0.25
155.26]
H5
-0.005
(13.03)**
-0.273
(2.44)*
0.115
(1.10)
0.336
(6.35)**
0.353
(3.42)**
-0.002
(0.47)
0.006
(0.38)
0.056
(4.15)**
0.214
(1.69)
0.028
(0.28)
-0.049
(0.28)
0.055
(0.33)
0.156
(1.31)
0.139
(3.45)**
-0.008
(1.34)
-0.002
(0.33)
-0.023
(0.75)
H5
-0.005
(13.02)**
-0.270
(2.41)*
0.109
(1.05)
0.333
(6.27)**
0.338
(3.26)**
-0.002
(0.46)
0.008
(0.53)
0.057
(4.21)**
0.216
(1.71)
0.023
(0.23)
-0.040
(0.23)
0.074
(0.44)
0.154
(1.29)
0.140
(3.49)**
-0.008
(1.31)
0.000
(0.03)
-0.029
(0.92)
-0.079
(2.71)**
stage_gdpgwr
Constant
H4
-0.005
(13.01)**
-0.272
(2.43)*
0.101
(0.97)
0.334
(6.29)**
0.351
(3.40)**
-0.002
(0.49)
0.007
(0.46)
0.057
(4.18)**
0.209
(1.66)
0.025
(0.26)
-0.046
(0.26)
0.058
(0.35)
0.157
(1.32)
0.131
(3.25)**
-0.005
(0.94)
-0.000
(0.07)
-0.026
(0.84)
-1.053
(1.59)
-457.08
0.66
(0.415)
134.22
[120.76 154.76]
0.25
-0.105
(2.39)*
-0.037
(2.23)*
-0.855
(1.28)
-454.93
4.97
(0.026)**
134.57
[121.66 155.16]
0.25
-0.642
(0.94)
-453.67
7.49
(0.006)***
134.97
[122.03 155.48]
0.25
-0.774
(1.15)
-454.54
5.74
(0.017)**
134.63
[121.7 155.14]
0.25
-0.064
(2.61)**
-0.770
(1.15)
-453.99
6.85
(0.009)***
134.77
[121.81 155.34]
0.25
Tables A6.11 Robustness check: substitute variable: stage (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.008
(23.33)***
-0.236
(2.34)**
-0.014
(0.15)
0.355
(7.57)***
0.235
(2.55)**
-0.003
(0.65)
-0.009
(0.70)
0.059
(4.98)***
0.202
(1.82)*
-0.060
(0.69)
-0.239
(1.53)
-0.169
(1.17)
0.187
(1.76)*
0.137
(3.85)***
-0.004
(0.89)
-0.002
(0.48)
-0.019
(0.70)
Stage
H2
-0.008
(23.39)***
-0.241
(2.40)**
-0.005
(0.05)
0.342
(7.19)***
0.223
(2.41)**
-0.003
(0.71)
-0.004
(0.31)
0.060
(5.04)***
0.218
(1.96)*
-0.066
(0.75)
-0.215
(1.37)
-0.106
(0.71)
0.176
(1.64)*
0.131
(3.64)***
-0.005
(1.08)
-0.001
(0.26)
-0.020
(0.73)
-0.033
(1.84)*
H3
-0.008
(23.35)***
-0.240
(2.39)**
0.000
(0.00)
0.340
(7.18)***
0.213
(2.30)**
-0.003
(0.68)
-0.003
(0.26)
0.061
(5.06)***
0.222
(2.00)**
-0.068
(0.78)
-0.212
(1.35)
-0.098
(0.65)
0.173
(1.62)
0.138
(3.85)***
-0.007
(1.34)
-0.001
(0.19)
-0.022
(0.80)
-0.056
(2.16)**
stage_pop
H4
-0.008
(23.48)**
-0.249
(2.47)*
-0.009
(0.10)
0.344
(7.26)**
0.228
(2.47)*
-0.003
(0.72)
-0.007
(0.49)
0.060
(5.02)**
0.209
(1.89)
-0.067
(0.76)
-0.217
(1.39)
-0.135
(0.92)
0.177
(1.66)
0.138
(3.86)**
-0.005
(0.97)
-0.003
(0.63)
-0.030
(1.04)
Double Bound
H4
-0.008
(23.42)**
-0.239
(2.38)*
-0.005
(0.05)
0.352
(7.49)**
0.236
(2.56)*
-0.003
(0.66)
-0.009
(0.65)
0.060
(4.99)**
0.208
(1.88)
-0.057
(0.65)
-0.234
(1.50)
-0.149
(1.01)
0.184
(1.72)
0.135
(3.75)**
-0.006
(1.12)
-0.002
(0.56)
-0.016
(0.58)
-0.025
(1.50)
stage_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.008
(23.32)**
-0.252
(2.50)*
-0.001
(0.01)
0.342
(7.22)**
0.220
(2.38)*
-0.003
(0.69)
-0.006
(0.43)
0.060
(5.01)**
0.215
(1.94)
-0.069
(0.78)
-0.209
(1.33)
-0.117
(0.79)
0.176
(1.64)
0.146
(4.04)**
-0.007
(1.27)
-0.003
(0.59)
-0.035
(1.21)
-0.018
(0.86)
-0.516
(0.89)
-1045.41
103.43
[92.41 114.66]
0.22
-0.410
(0.71)
-1043.72
3.38
(0.0660)*
103.58
[93.88 113.91]
0.19
-0.360
(0.62)
-1043.08
4.65
(0.0311)**
103.65
[93.97 113.98]
0.19
-0.284
(0.48)
-1044.27
2.26
(0.132)
103.6
[95.91 115.14]
0.19
H5
-0.008
(23.33)**
-0.242
(2.40)*
0.007
(0.07)
0.343
(7.26)**
0.218
(2.36)*
-0.003
(0.70)
-0.005
(0.37)
0.061
(5.06)**
0.222
(2.00)*
-0.062
(0.71)
-0.216
(1.38)
-0.101
(0.68)
0.172
(1.61)
0.138
(3.86)**
-0.008
(1.55)
-0.002
(0.41)
-0.024
(0.87)
H5
-0.008
(23.33)**
-0.238
(2.37)*
-0.000
(0.00)
0.341
(7.20)**
0.210
(2.26)*
-0.003
(0.70)
-0.004
(0.27)
0.061
(5.08)**
0.223
(2.00)*
-0.066
(0.75)
-0.211
(1.35)
-0.094
(0.63)
0.174
(1.63)
0.140
(3.89)**
-0.008
(1.46)
-0.001
(0.12)
-0.028
(1.01)
-0.056
(2.13)*
stage_gdpgwr
Constant
H4
-0.008
(23.52)**
-0.241
(2.39)*
-0.006
(0.07)
0.343
(7.22)**
0.219
(2.37)*
-0.003
(0.72)
-0.005
(0.34)
0.061
(5.06)**
0.217
(1.95)
-0.064
(0.73)
-0.216
(1.38)
-0.108
(0.72)
0.177
(1.66)
0.132
(3.69)**
-0.006
(1.17)
-0.001
(0.22)
-0.026
(0.93)
-0.445
(0.76)
-1045.04
0.74
(0.391)
103.39
[95.35 114.71]
0.19
-0.086
(2.13)*
-0.027
(1.76)
-0.321
(0.55)
-1043.85
3.11
(0.078)*
103.58
[95.78 115.14]
0.19
-0.178
(0.30)
-1043.12
4.57
(0.032)**
103.72
[96 115.23]
0.19
-0.238
(0.40)
-1043.14
4.54
(0.033)**
103.55
[95.68 114.99]
0.19
-0.047
(2.12)*
-0.257
(0.44)
-1043.15
4.50
(0.034)**
103.64
[95.79 115.18]
0.19
Tables A6.12 Robustness check: substitute variable: stage (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.011
(38.13)***
-18.436
(2.19)**
7.080
(0.95)
15.093
(3.87)***
20.245
(2.67)***
-0.271
(0.79)
-0.554
(0.46)
4.019
(3.96)***
27.176
(2.96)***
-5.829
(0.81)
-25.250
(2.02)**
-28.295
(2.38)**
10.770
(1.20)
5.134
(1.65)*
-0.917
(2.29)**
-1.011
(2.83)***
9.605
(4.16)***
Stage
H2
-0.011
(38.13)***
-17.912
(2.16)**
12.432
(1.68)*
11.499
(2.94)***
16.514
(2.20)**
-0.255
(0.76)
-0.433
(0.36)
4.503
(4.48)***
27.268
(3.02)***
-6.344
(0.89)
-21.916
(1.77)*
-23.241
(1.98)**
13.263
(1.50)
3.462
(1.12)
-1.303
(3.24)***
-0.728
(2.04)**
9.641
(4.24)***
-6.757
(4.87)***
H3
-0.011
(38.13)***
-17.669
(2.13)**
11.950
(1.61)
11.844
(3.03)***
16.596
(2.21)**
-0.269
(0.80)
-0.406
(0.34)
4.565
(4.53)***
26.538
(2.94)***
-6.381
(0.89)
-22.413
(1.81)*
-24.003
(2.04)**
13.488
(1.52)
4.646
(1.51)
-1.477
(3.58)***
-0.720
(2.01)**
9.140
(4.01)***
-9.108
(4.66)***
stage_pop
H4
-0.011
(38.13)**
-18.284
(2.18)*
7.961
(1.07)
13.287
(3.36)**
19.901
(2.64)**
-0.270
(0.79)
-0.820
(0.68)
4.151
(4.10)**
28.107
(3.07)**
-6.902
(0.96)
-23.813
(1.91)
-27.675
(2.33)*
11.384
(1.27)
4.851
(1.56)
-0.963
(2.41)*
-1.038
(2.91)**
7.723
(3.18)**
Wang & He
H4
-0.011
(38.13)**
-19.178
(2.27)*
8.632
(1.14)
13.966
(3.49)**
20.698
(2.73)**
-0.253
(0.74)
-0.575
(0.48)
4.024
(3.97)**
27.422
(2.99)**
-5.939
(0.82)
-25.082
(2.00)*
-28.317
(2.38)*
11.185
(1.25)
4.986
(1.60)
-1.130
(2.59)**
-1.052
(2.93)**
9.847
(4.25)**
-3.510
(2.45)*
stage_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.011
(38.13)**
-18.323
(2.18)*
8.423
(1.13)
13.287
(3.37)**
20.356
(2.70)**
-0.279
(0.82)
-0.780
(0.65)
4.171
(4.12)**
27.701
(3.03)**
-6.505
(0.90)
-24.359
(1.95)
-27.554
(2.32)*
11.493
(1.29)
5.641
(1.81)
-1.142
(2.80)**
-1.016
(2.85)**
7.776
(3.23)**
-2.317
(1.23)
-62.418
(1.27)
-4344.03
75.99
[68.89 83.48]
0.19
-23.962
(0.49)
-4332.38
23.30
(0.000)***
75.99
[69.62 82.872]
0.17
-17.051
(0.35)
-4333.34
21.37
(0.000)***
75.99
[69.62 82.88]
0.17
-16.570
(0.32)
-4341.04
5.98
(0.014)**
75.99
[71.14 83.44]
0.16
H5
-0.011
(38.13)**
-18.976
(2.27)*
11.204
(1.50)
11.929
(3.02)**
18.963
(2.53)*
-0.261
(0.77)
-0.537
(0.45)
4.392
(4.34)**
26.840
(2.95)**
-6.504
(0.91)
-23.689
(1.91)
-26.435
(2.24)*
12.798
(1.44)
5.087
(1.65)
-1.516
(3.53)**
-0.928
(2.61)**
8.608
(3.74)**
H5
-0.011
(38.13)**
-18.149
(2.19)*
12.786
(1.73)
11.319
(2.91)**
15.976
(2.14)*
-0.276
(0.82)
-0.265
(0.22)
4.766
(4.73)**
26.389
(2.93)**
-7.023
(0.99)
-22.687
(1.84)
-24.280
(2.07)*
14.670
(1.66)
4.909
(1.60)
-1.697
(4.03)**
-0.644
(1.80)
7.913
(3.45)**
-5.500
(2.55)*
stage_gdpgwr
Constant
H4
-0.011
(38.13)**
-18.542
(2.24)*
13.078
(1.77)
11.041
(2.83)**
15.905
(2.13)*
-0.266
(0.79)
-0.294
(0.25)
4.708
(4.69)**
27.147
(3.02)**
-7.026
(0.99)
-22.213
(1.81)
-23.532
(2.01)*
14.344
(1.63)
3.683
(1.20)
-1.503
(3.68)**
-0.655
(1.83)
8.227
(3.61)**
-46.745
(0.92)
-4343.27
1.52
(0.217)
75.99
[71.13 83.46]
0.16
-11.542
(3.69)**
-6.352
(5.35)**
3.534
(0.07)
-4329.98
28.10
(0.000)***
75.99
[71.21 83.33]
0.16
-13.616
(0.26)
-4340.81
6.45
(0.011)**
75.99
[71.14 83.44]
0.16
-3.330
(0.07)
-4337.29
13.49
(0.000)***
75.99
[71.17 83.4]
0.16
-8.561
(5.16)**
8.396
(0.17)
-4330.93
26.20
(0.000)***
75.99
[71.21 83.34]
0.16
Tables A6.13 Robustness check: substitute variable: population (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.025
(34.88)**
-0.226
(2.50)*
0.038
(0.48)
0.165
(3.95)**
0.233
(2.86)**
-0.001
(0.39)
0.011
(0.85)
0.051
(4.63)**
0.300
(3.06)**
-0.042
(0.54)
-0.080
(0.59)
-0.100
(0.78)
0.121
(1.27)
0.184
(5.46)**
-0.001
(0.23)
-0.027
(7.07)**
-0.013
(0.53)
Total_population
H2
-0.025
(34.87)**
-0.230
(2.55)*
0.097
(1.21)
0.123
(2.88)**
0.192
(2.35)*
-0.001
(0.34)
0.013
(0.98)
0.057
(5.20)**
0.313
(3.19)**
-0.055
(0.71)
-0.041
(0.31)
-0.048
(0.37)
0.155
(1.61)
0.164
(4.83)**
-0.005
(1.07)
-0.026
(6.55)**
-0.023
(0.91)
-0.015
(5.08)**
H3
-0.025
(34.87)**
-0.225
(2.49)*
0.099
(1.22)
0.130
(3.07)**
0.195
(2.39)*
-0.001
(0.34)
0.014
(1.08)
0.058
(5.25)**
0.291
(2.97)**
-0.054
(0.70)
-0.060
(0.44)
-0.071
(0.55)
0.154
(1.61)
0.195
(5.78)**
-0.011
(2.34)*
-0.021
(4.94)**
-0.029
(1.16)
-2.185
(4.70)**
pop_pop
H4
-0.025
(34.89)**
-0.229
(2.54)*
0.045
(0.57)
0.127
(3.01)**
0.216
(2.66)**
-0.002
(0.46)
0.005
(0.39)
0.055
(5.01)**
0.338
(3.43)**
-0.081
(1.04)
-0.055
(0.41)
-0.099
(0.77)
0.129
(1.35)
0.177
(5.26)**
-0.001
(0.18)
-0.030
(7.72)**
-0.059
(2.25)*
Def Yes
H4
-0.025
(34.86)**
-0.252
(2.79)**
0.075
(0.94)
0.136
(3.18)**
0.245
(3.02)**
-0.001
(0.21)
0.010
(0.81)
0.051
(4.66)**
0.307
(3.12)**
-0.047
(0.60)
-0.073
(0.54)
-0.102
(0.80)
0.130
(1.37)
0.177
(5.24)**
-0.007
(1.43)
-0.029
(7.49)**
-0.007
(0.29)
-0.014
(5.42)**
pop_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.025
(34.88)**
-0.227
(2.51)*
0.080
(1.00)
0.125
(2.95)**
0.219
(2.70)**
-0.001
(0.39)
0.009
(0.67)
0.057
(5.16)**
0.306
(3.12)**
-0.064
(0.83)
-0.064
(0.48)
-0.092
(0.72)
0.143
(1.50)
0.207
(6.07)**
-0.010
(2.25)*
-0.023
(5.80)**
-0.053
(2.06)*
-0.009
(3.35)**
-0.306
(0.58)
-2184.3
27.21
[24.04 30.52]
0.24
0.274
(0.51)
-2171.37
25.86
(0.000)***
27.3
[25.32 30.43]
0.19
0.422
(0.77)
-2173.25
22.1
(0.000)***
27.29
[25.31 30.45]
0.19
0.792
(1.41)
-2169.55
29.5
(0.014)**
27.26
[25.26 30.38]
0.19
H5
-0.025
(34.86)**
-0.237
(2.63)**
0.095
(1.17)
0.128
(3.02)**
0.213
(2.62)**
-0.001
(0.28)
0.013
(1.03)
0.057
(5.14)**
0.291
(2.96)**
-0.054
(0.70)
-0.065
(0.48)
-0.086
(0.67)
0.153
(1.60)
0.193
(5.72)**
-0.012
(2.39)*
-0.024
(6.01)**
-0.026
(1.03)
H5
-0.025
(34.88)**
-0.226
(2.50)*
0.038
(0.48)
0.165
(3.95)**
0.233
(2.86)**
-0.001
(0.39)
0.011
(0.85)
0.051
(4.63)**
0.300
(3.06)**
-0.042
(0.54)
-0.080
(0.59)
-0.100
(0.78)
0.121
(1.27)
0.184
(5.46)**
-0.001
(0.23)
-0.027
(7.07)**
-0.013
(0.53)
-3.111
(5.24)**
pop_gdpgwr
Constant
H4
-0.025
(34.86)**
-0.235
(2.61)**
0.098
(1.21)
0.123
(2.89)**
0.187
(2.29)*
-0.001
(0.38)
0.014
(1.08)
0.059
(5.33)**
0.310
(3.16)**
-0.060
(0.78)
-0.046
(0.34)
-0.054
(0.42)
0.162
(1.69)
0.168
(4.96)**
-0.006
(1.41)
-0.025
(6.43)**
-0.034
(1.36)
0.147
(0.27)
-2178.69
11.23
(0.217)
27.25
[25.21 30.41]
0.19
-2.461
(4.32)**
-0.012
(5.10)**
0.469
(0.86)
-2171.27
26.06
(0.000)***
27.3
[25.32 30.43]
0.19
0.873
(1.53)
-2170.57
27.48
(0.011)**
27.28
[25.29 30.41]
0.19
0.487
(0.88)
-2174.99
18.63
(0.000)***
27.28
[25.28 30.44]
0.19
-1.630
(4.47)**
0.464
(0.84)
-2174.29
20.03
(0.000)***
27.28
[25.29 30.44]
0.19
Tables A6.14 Robustness check: substitute variable: population (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.016
(36.10)**
-0.310
(3.45)**
0.027
(0.34)
0.169
(4.09)**
0.305
(3.77)**
-0.003
(0.95)
0.002
(0.12)
0.039
(3.62)**
0.371
(3.79)**
0.026
(0.34)
-0.163
(1.22)
-0.191
(1.51)
0.180
(1.89)
0.137
(4.11)**
-0.005
(1.21)
-0.022
(5.59)**
0.057
(2.31)*
Total_population
H2
-0.017
(36.09)**
-0.316
(3.51)**
0.088
(1.10)
0.127
(2.99)**
0.266
(3.27)**
-0.003
(0.89)
0.003
(0.26)
0.046
(4.20)**
0.387
(3.95)**
0.013
(0.17)
-0.123
(0.92)
-0.140
(1.10)
0.215
(2.25)*
0.117
(3.49)**
-0.009
(2.05)*
-0.019
(5.04)**
0.049
(1.98)*
-0.015
(5.20)**
H3
-0.017
(36.09)**
-0.311
(3.46)**
0.095
(1.18)
0.131
(3.11)**
0.266
(3.27)**
-0.003
(0.89)
0.005
(0.38)
0.047
(4.32)**
0.365
(3.72)**
0.012
(0.16)
-0.140
(1.05)
-0.161
(1.27)
0.218
(2.28)*
0.150
(4.49)**
-0.017
(3.46)**
-0.014
(3.33)**
0.041
(1.66)
-2.432
(5.25)**
pop_pop
H4
-0.017
(36.09)**
-0.313
(3.48)**
0.032
(0.40)
0.146
(3.47)**
0.295
(3.64)**
-0.004
(0.99)
-0.002
(0.15)
0.042
(3.82)**
0.394
(4.01)**
0.002
(0.03)
-0.146
(1.09)
-0.190
(1.50)
0.185
(1.94)
0.132
(3.97)**
-0.005
(1.18)
-0.023
(5.96)**
0.030
(1.15)
Prob Yes
H4
-0.016
(36.10)**
-0.313
(3.46)**
0.031
(0.38)
0.167
(3.92)**
0.306
(3.78)**
-0.003
(0.94)
0.002
(0.12)
0.039
(3.63)**
0.372
(3.79)**
0.026
(0.34)
-0.162
(1.21)
-0.191
(1.51)
0.180
(1.89)
0.136
(4.08)**
-0.006
(1.24)
-0.022
(5.58)**
0.057
(2.32)*
-0.008
(3.35)**
pop_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.017
(36.09)**
-0.312
(3.47)**
0.062
(0.78)
0.137
(3.24)**
0.295
(3.65)**
-0.003
(0.95)
-0.000
(0.02)
0.044
(4.03)**
0.378
(3.85)**
0.008
(0.10)
-0.149
(1.12)
-0.185
(1.46)
0.198
(2.08)*
0.155
(4.62)**
-0.013
(2.77)**
-0.018
(4.52)**
0.025
(0.99)
-0.001
(0.30)
-0.557
(1.07)
-2202.27
52.68
[47.89 57.61]
0.18
0.005
(0.01)
-2188.74
27.06
(0.000)***
52.71
[49.65 57.39]
0.15
0.226
(0.42)
-2188.47
27.6
(0.000)***
52.7
[49.65 57.37]
0.15
0.100
(0.18)
-2196.65
11.25
(0.014)**
52.68
[49.55 57.43]
0.15
H5
-0.017
(36.09)**
-0.320
(3.56)**
0.077
(0.96)
0.137
(3.24)**
0.289
(3.56)**
-0.003
(0.86)
0.003
(0.27)
0.044
(4.06)**
0.364
(3.72)**
0.015
(0.20)
-0.149
(1.12)
-0.180
(1.42)
0.207
(2.17)*
0.145
(4.34)**
-0.015
(2.95)**
-0.018
(4.64)**
0.047
(1.90)
H5
-0.017
(36.09)**
-0.314
(3.49)**
0.091
(1.14)
0.133
(3.16)**
0.263
(3.23)**
-0.003
(0.92)
0.006
(0.44)
0.048
(4.40)**
0.362
(3.69)**
0.009
(0.12)
-0.142
(1.07)
-0.166
(1.31)
0.224
(2.34)*
0.150
(4.49)**
-0.017
(3.51)**
-0.014
(3.53)**
0.036
(1.46)
-2.528
(4.30)**
pop_gdpgwr
Constant
H4
-0.017
(36.08)**
-0.322
(3.58)**
0.089
(1.11)
0.126
(2.99)**
0.261
(3.20)**
-0.003
(0.94)
0.005
(0.36)
0.048
(4.34)**
0.384
(3.91)**
0.007
(0.09)
-0.127
(0.95)
-0.146
(1.15)
0.223
(2.33)*
0.121
(3.61)**
-0.011
(2.40)*
-0.019
(4.91)**
0.037
(1.49)
-0.517
(0.96)
-2202.23
0.09
(0.766)
52.68
[49.49 57.47]
0.15
-2.139
(3.78)**
-0.013
(5.27)**
0.210
(0.39)
-2188.39
27.77
(0.000)***
52.7
[49.64 57.38]
0.15
0.380
(0.67)
-2193.02
18.51
(0.000)**
52.69
[49.61 57.39]
0.15
0.117
(0.21)
-2195.13
14.29
(0.000)***
52.69
[49.59 57.41]
0.15
-1.860
(5.12)**
0.294
(0.54)
-2189.15
26.25
(0.000)***
52.7
[49.64 57.37]
0.15
Tables A6.15 Robustness check: substitute variable: population (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.009
(36.11)**
-0.126
(1.41)
0.067
(0.86)
0.167
(4.04)**
0.222
(2.76)**
-0.003
(0.73)
-0.003
(0.27)
0.036
(3.28)**
0.291
(2.99)**
-0.054
(0.70)
-0.325
(2.43)*
-0.395
(3.11)**
0.057
(0.60)
0.049
(1.49)
-0.011
(2.71)**
-0.009
(2.30)*
0.117
(4.74)**
Total_population
H2
-0.009
(36.11)**
-0.127
(1.42)
0.116
(1.45)
0.133
(3.15)**
0.190
(2.34)*
-0.002
(0.67)
-0.002
(0.16)
0.040
(3.71)**
0.302
(3.09)**
-0.065
(0.84)
-0.296
(2.21)*
-0.357
(2.80)**
0.082
(0.86)
0.032
(0.96)
-0.014
(3.36)**
-0.007
(1.80)
0.111
(4.50)**
-0.012
(4.13)**
H3
-0.009
(36.11)**
-0.123
(1.38)
0.130
(1.63)
0.132
(3.13)**
0.185
(2.28)*
-0.002
(0.67)
-0.001
(0.04)
0.043
(3.91)**
0.284
(2.90)**
-0.068
(0.88)
-0.308
(2.31)*
-0.372
(2.92)**
0.089
(0.93)
0.059
(1.80)
-0.022
(4.61)**
-0.001
(0.32)
0.103
(4.16)**
-2.236
(4.85)**
pop_pop
H4
-0.009
(36.11)**
-0.126
(1.41)
0.070
(0.89)
0.150
(3.58)**
0.215
(2.67)**
-0.003
(0.75)
-0.006
(0.46)
0.037
(3.41)**
0.307
(3.13)**
-0.071
(0.92)
-0.314
(2.35)*
-0.395
(3.11)**
0.059
(0.62)
0.045
(1.37)
-0.011
(2.69)**
-0.010
(2.55)*
0.098
(3.78)**
Not Sure
H4
-0.009
(36.11)**
-0.128
(1.43)
0.071
(0.89)
0.165
(3.88)**
0.223
(2.77)**
-0.003
(0.72)
-0.003
(0.27)
0.036
(3.28)**
0.292
(2.99)**
-0.054
(0.71)
-0.324
(2.43)*
-0.395
(3.11)**
0.057
(0.60)
0.049
(1.47)
-0.012
(2.63)**
-0.009
(2.32)*
0.117
(4.75)**
-0.006
(2.37)*
pop_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.009
(36.11)**
-0.125
(1.40)
0.096
(1.21)
0.140
(3.32)**
0.213
(2.65)**
-0.003
(0.72)
-0.005
(0.39)
0.039
(3.60)**
0.296
(3.03)**
-0.070
(0.91)
-0.315
(2.36)*
-0.392
(3.09)**
0.070
(0.73)
0.063
(1.90)
-0.018
(3.88)**
-0.006
(1.41)
0.091
(3.55)**
-0.001
(0.28)
-0.749
(1.44)
-2312
92.56
[83.87 101.56]
0.19
-0.305
(0.58)
-2303.46
17.08
(0.000)***
92.58
[86.97 101.12]
0.15
-0.034
(0.06)
-2300.26
23.49
(0.000)***
92.58
[86.98 101.1]
0.15
-0.287
(0.52)
-2309.2
5.6
(0.018)**
92.56
[86.9 101.16]
0.15
H5
-0.009
(36.10)**
-0.134
(1.50)
0.120
(1.49)
0.134
(3.17)**
0.205
(2.54)*
-0.002
(0.63)
-0.002
(0.12)
0.041
(3.73)**
0.284
(2.91)**
-0.066
(0.86)
-0.314
(2.35)*
-0.387
(3.04)**
0.083
(0.88)
0.056
(1.71)
-0.021
(4.32)**
-0.005
(1.34)
0.107
(4.33)**
H5
-0.009
(36.11)**
-0.126
(1.41)
0.067
(0.86)
0.167
(4.04)**
0.222
(2.76)**
-0.003
(0.73)
-0.003
(0.27)
0.036
(3.28)**
0.291
(2.99)**
-0.054
(0.70)
-0.325
(2.43)*
-0.395
(3.11)**
0.057
(0.60)
0.049
(1.49)
-0.011
(2.71)**
-0.009
(2.30)*
0.117
(4.74)**
-2.096
(3.59)**
pop_gdpgwr
Constant
H4
-0.009
(36.10)**
-0.132
(1.48)
0.123
(1.54)
0.128
(3.04)**
0.181
(2.24)*
-0.003
(0.71)
-0.001
(0.06)
0.043
(3.91)**
0.301
(3.08)**
-0.072
(0.94)
-0.297
(2.22)*
-0.358
(2.81)**
0.092
(0.97)
0.033
(1.00)
-0.016
(3.74)**
-0.006
(1.63)
0.100
(4.02)**
-0.712
(1.33)
-2311.96
0.08
(0.779)
92.56
[86.8 101.2]
0.16
-2.215
(3.92)**
-0.011
(4.72)**
-0.066
(0.12)
-2300.84
22.32
(0.000)***
92.57
[86.99 101.09]
0.15
0.027
(0.05)
-2305.57
12.86
(0.011)**
92.57
[86.93 101.14]
0.15
-0.055
(0.10)
-2304.3
15.4
(0.000)***
92.56
[86.91 101.12]
0.15
-1.865
(5.15)**
0.099
(0.18)
-2298.74
26.51
(0.000)***
92.57
[86.99 101.07]
0.15
Tables A6.16 Robustness check: substitute variable: population (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.011
(38.13)**
-18.436
(2.19)*
7.080
(0.95)
15.093
(3.87)**
20.245
(2.67)**
-0.271
(0.79)
-0.554
(0.46)
4.019
(3.96)**
27.176
(2.96)**
-5.829
(0.81)
-25.250
(2.02)*
-28.295
(2.38)*
10.770
(1.20)
5.134
(1.65)
-0.917
(2.29)*
-1.011
(2.83)**
9.605
(4.16)**
Total_population
H2
-0.011
(38.13)**
-18.309
(2.20)*
11.967
(1.61)
11.441
(2.91)**
16.625
(2.21)*
-0.248
(0.73)
-0.421
(0.35)
4.475
(4.44)**
27.850
(3.08)**
-6.868
(0.96)
-22.002
(1.78)
-23.977
(2.04)*
13.295
(1.50)
3.276
(1.06)
-1.200
(3.00)**
-0.811
(2.28)*
8.853
(3.88)**
-1.177
(4.50)**
H3
-0.011
(38.13)**
-17.863
(2.16)*
13.057
(1.76)
11.481
(2.94)**
16.239
(2.17)*
-0.246
(0.73)
-0.279
(0.23)
4.654
(4.62)**
25.907
(2.87)**
-7.071
(0.99)
-23.194
(1.88)
-25.467
(2.17)*
13.718
(1.55)
5.971
(1.95)
-1.896
(4.31)**
-0.287
(0.76)
8.127
(3.55)**
-213.065
(5.00)**
pop_pop
H4
-0.011
(38.13)**
-18.371
(2.19)*
7.382
(1.00)
13.414
(3.40)**
19.430
(2.57)*
-0.278
(0.82)
-0.798
(0.66)
4.151
(4.10)**
28.503
(3.11)**
-7.457
(1.03)
-24.105
(1.93)
-28.118
(2.37)*
10.969
(1.23)
4.725
(1.52)
-0.903
(2.26)*
-1.104
(3.08)**
7.744
(3.19)**
-0.565
(2.43)*
pop_gdppc
Wang and HE
H4
-0.011
(38.13)**
-18.420
(2.18)*
7.057
(0.94)
15.112
(3.79)**
20.234
(2.67)**
-0.272
(0.79)
-0.554
(0.46)
4.019
(3.96)**
27.173
(2.96)**
-5.827
(0.81)
-25.254
(2.02)*
-28.290
(2.38)*
10.768
(1.20)
5.137
(1.65)
-0.913
(2.13)*
-1.010
(2.79)**
9.601
(4.14)**
0.006
(0.02)
pop_gdpgwr
Constant
Log likelihood
LR
Mean WTP
KR CI
Ratio
-62.418
(1.27)
-4344.03
75.99
[71.19 83.37]
0.2
-16.358
(0.33)
-4334.06
19.95
(0.000)***
75.99
[71.2 83.35]
0.16
6.984
(0.14)
-4331.72
24.62
(0.000)***
75.99
[71.14 83.44]
0.16
-17.693
(0.34)
-4341.09
5.88
(0.015)**
75.99
[71.12 83.47]
0.16
-62.681
(1.24)
-4344.03
0
(0.982)
75.99
[71.2 83.35]
0.16
H4
-0.011
(38.13)**
-18.785
(2.27)*
12.442
(1.68)
11.085
(2.83)**
15.856
(2.11)*
-0.261
(0.77)
-0.299
(0.25)
4.669
(4.63)**
27.591
(3.06)**
-7.486
(1.05)
-22.120
(1.79)
-24.120
(2.05)*
14.170
(1.60)
3.446
(1.12)
-1.368
(3.38)**
-0.754
(2.12)*
7.783
(3.38)**
H5
-0.011
(38.13)**
-18.143
(2.17)*
9.832
(1.33)
12.354
(3.14)**
19.137
(2.55)*
-0.266
(0.78)
-0.709
(0.59)
4.343
(4.30)**
27.275
(3.00)**
-7.309
(1.02)
-24.141
(1.94)
-27.681
(2.35)*
11.983
(1.35)
6.364
(2.05)*
-1.507
(3.52)**
-0.694
(1.90)
7.073
(2.96)**
H5
-0.008
(23.57)**
-0.239
(2.38)*
0.004
(0.04)
0.347
(7.35)**
0.223
(2.42)*
-0.003
(0.66)
-0.007
(0.50)
0.060
(5.04)**
0.218
(1.96)*
-0.057
(0.65)
-0.224
(1.43)
-0.119
(0.80)
0.177
(1.65)
0.143
(3.97)**
-0.009
(1.53)
-0.001
(0.15)
-0.022
(0.80)
H5
-0.008
(23.37)**
-0.236
(2.34)*
-0.014
(0.15)
0.355
(7.57)**
0.235
(2.55)*
-0.003
(0.65)
-0.009
(0.70)
0.059
(4.98)**
0.202
(1.82)
-0.060
(0.69)
-0.239
(1.53)
-0.169
(1.17)
0.187
(1.76)
0.137
(3.85)**
-0.004
(0.89)
-0.002
(0.48)
-0.019
(0.70)
-199.340
(3.66)**
-1.107
(4.98)**
5.330
(0.11)
-4331.85
24.36
(0.000)***
75.99
[71.17 83.4]
0.16
-238.695
(3.71)**
11.867
(0.23)
-4337.41
13.24
(0.000)***
75.99
[71.17 83.4]
0.16
0.125
(0.00)
-4337.2
13.67
(0.000)***
75.99
[71.21 83.33]
0.16
-174.520
(5.23)**
18.174
(0.36)
-4330.6
26.86
(0.000)***
75.99
[71.21 83.34]
0.16
Tables A6.17 Robustness check: substitute variable: population (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.005
(13.01)**
-0.265
(2.37)*
0.087
(0.84)
0.352
(6.69)**
0.370
(3.60)**
-0.002
(0.40)
0.000
(0.02)
0.055
(4.09)**
0.187
(1.49)
0.031
(0.32)
-0.079
(0.45)
-0.020
(0.12)
0.174
(1.47)
0.137
(3.45)**
-0.003
(0.58)
-0.002
(0.38)
-0.017
(0.54)
Total_population
H2
-0.005
(13.03)**
-0.274
(2.45)*
0.100
(0.97)
0.333
(6.25)**
0.359
(3.49)**
-0.002
(0.48)
0.007
(0.48)
0.056
(4.16)**
0.207
(1.64)
0.022
(0.23)
-0.046
(0.26)
0.051
(0.30)
0.158
(1.32)
0.128
(3.18)**
-0.004
(0.74)
-0.001
(0.25)
-0.023
(0.73)
-0.008
(2.11)*
H3
-0.005
(13.02)**
-0.272
(2.43)*
0.106
(1.02)
0.336
(6.33)**
0.347
(3.35)**
-0.002
(0.42)
0.005
(0.35)
0.057
(4.19)**
0.213
(1.69)
0.028
(0.29)
-0.046
(0.26)
0.059
(0.35)
0.153
(1.29)
0.147
(3.63)**
-0.008
(1.34)
0.002
(0.36)
-0.024
(0.78)
-1.315
(2.07)*
pop_pop
H4
-0.005
(13.04)**
-0.281
(2.49)*
0.084
(0.81)
0.341
(6.42)**
0.366
(3.56)**
-0.002
(0.47)
0.003
(0.21)
0.055
(4.12)**
0.191
(1.52)
0.024
(0.24)
-0.058
(0.33)
0.001
(0.01)
0.165
(1.39)
0.138
(3.44)**
-0.003
(0.56)
-0.003
(0.60)
-0.027
(0.84)
Single Bound
H4
-0.005
(13.01)**
-0.267
(2.38)*
0.093
(0.89)
0.350
(6.65)**
0.373
(3.62)**
-0.002
(0.41)
0.000
(0.03)
0.055
(4.08)**
0.191
(1.52)
0.033
(0.34)
-0.078
(0.44)
-0.015
(0.09)
0.174
(1.47)
0.136
(3.39)**
-0.004
(0.70)
-0.002
(0.45)
-0.015
(0.48)
-0.004
(1.22)
pop_gdppc
Log likelihood
LR
Mean WTP
KR CI
Ratio
H5
-0.005
(13.05)**
-0.286
(2.55)*
0.101
(0.97)
0.334
(6.29)**
0.351
(3.40)**
-0.002
(0.44)
0.004
(0.31)
0.056
(4.16)**
0.208
(1.65)
0.024
(0.25)
-0.038
(0.22)
0.047
(0.29)
0.154
(1.29)
0.153
(3.73)**
-0.007
(1.26)
-0.000
(0.02)
-0.034
(1.08)
-0.002
(0.46)
-1.130
(1.73)
-457.41
134.23
[113.95
0.29
152.41]
-0.931
(1.41)
-455.19
4.45
(0.035)**
134.48
[121.52
0.25
155.05]
-0.839
(1.26)
-455.26
4.3
(0.038)**
134.61
[121.63 155.12]
0.25
-0.920
(1.36)
-456.67
1.48
(0.224)
134.45
[121.44
0.25
155.1]
H5
-0.005
(13.01)**
-0.270
(2.41)*
0.110
(1.06)
0.341
(6.44)**
0.358
(3.48)**
-0.002
(0.42)
0.003
(0.22)
0.056
(4.14)**
0.212
(1.68)
0.034
(0.35)
-0.059
(0.34)
0.037
(0.22)
0.162
(1.36)
0.144
(3.57)**
-0.009
(1.38)
-0.000
(0.02)
-0.020
(0.65)
H5
-0.005
(13.00)**
-0.270
(2.41)*
0.104
(1.00)
0.337
(6.36)**
0.343
(3.31)**
-0.002
(0.43)
0.005
(0.34)
0.057
(4.20)**
0.213
(1.69)
0.029
(0.30)
-0.049
(0.28)
0.055
(0.33)
0.156
(1.31)
0.147
(3.62)**
-0.009
(1.38)
0.002
(0.31)
-0.026
(0.84)
-1.675
(2.19)*
pop_gdpgwr
Constant
H4
-0.005
(13.01)**
-0.274
(2.44)*
0.097
(0.94)
0.335
(6.29)**
0.353
(3.42)**
-0.002
(0.50)
0.007
(0.45)
0.057
(4.17)**
0.207
(1.64)
0.025
(0.26)
-0.050
(0.28)
0.045
(0.27)
0.160
(1.34)
0.130
(3.23)**
-0.005
(0.84)
-0.001
(0.20)
-0.027
(0.87)
-1.088
(1.65)
-457.31
0.21
(0.649)
134.2
[120.67 154.73]
0.25
-1.731
(1.86)
-0.006
(2.00)*
-0.851
(1.27)
-455.4
4.02
(0.045)**
134.51
[121.6 155.11]
0.25
-0.683
(1.00)
-454.99
4.84
(0.028)**
134.81
[121.87 155.31]
0.25
-0.827
(1.23)
-455.67
3.48
(0.062)*
134.47
[121.32 154.93]
0.25
-1.016
(2.04)*
-0.800
(1.19)
-455.34
4.15
(0.042)**
134.62
[121.65 155.15]
0.25
Tables A6.18 Robustness check: substitute variable: population (continue)
Bid price
rep_gov
water_problem
will_service
quality_deg
age
education
income_level
Income significant higher
than need
male
Can see the river
Far from the river
d_fish
Degree
share2nd
pop_density
gdp_growth
H1
-0.008
(23.48)**
-0.236
(2.34)*
-0.014
(0.15)
0.355
(7.57)**
0.235
(2.55)*
-0.003
(0.65)
-0.009
(0.70)
0.059
(4.98)**
0.202
(1.82)
-0.060
(0.69)
-0.239
(1.53)
-0.169
(1.17)
0.187
(1.76)
0.137
(3.85)**
-0.004
(0.89)
-0.002
(0.48)
-0.019
(0.70)
Total_population
H2
-0.008
(23.28)**
-0.241
(2.39)*
-0.007
(0.08)
0.343
(7.20)**
0.226
(2.45)*
-0.003
(0.71)
-0.005
(0.36)
0.060
(5.04)**
0.214
(1.93)
-0.065
(0.74)
-0.219
(1.39)
-0.119
(0.80)
0.179
(1.67)
0.131
(3.65)**
-0.005
(1.00)
-0.002
(0.37)
-0.023
(0.83)
-0.005
(1.51)
H3
-0.008
(23.40)**
-0.239
(2.38)*
-0.001
(0.01)
0.344
(7.25)**
0.216
(2.33)*
-0.003
(0.66)
-0.005
(0.41)
0.061
(5.06)**
0.219
(1.97)*
-0.062
(0.71)
-0.216
(1.38)
-0.107
(0.72)
0.173
(1.62)
0.144
(4.00)**
-0.008
(1.47)
0.001
(0.15)
-0.025
(0.89)
-0.963
(1.67)
pop_pop
H4
-0.008
(23.57)**
-0.241
(2.39)*
-0.015
(0.16)
0.350
(7.37)**
0.233
(2.52)*
-0.003
(0.68)
-0.008
(0.60)
0.060
(5.00)**
0.202
(1.83)
-0.063
(0.72)
-0.232
(1.48)
-0.160
(1.09)
0.184
(1.72)
0.138
(3.85)**
-0.004
(0.88)
-0.003
(0.57)
-0.023
(0.81)
-0.001
(0.55)
pop_gdppc
Double Bound
H4
-0.008
(23.65)**
-0.237
(2.36)*
-0.009
(0.09)
0.354
(7.53)**
0.237
(2.57)*
-0.003
(0.65)
-0.009
(0.68)
0.059
(4.98)**
0.204
(1.85)
-0.057
(0.65)
-0.238
(1.53)
-0.162
(1.11)
0.186
(1.75)
0.136
(3.78)**
-0.005
(0.99)
-0.002
(0.54)
-0.018
(0.63)
-0.002
(0.48)
pop_gdpgwr
Constant
Log likelihood
LR
Mean WTP
KR CI
Ratio
-0.516
(0.89)
-1045.41
103.43
[91.57 114.44]
0.22
-0.389
(0.67)
-1044.26
2.29
(0.131)
103.51
[95.68 115.01]
0.19
-0.312
(0.53)
-1044.01
2.79
(0.095)*
103.55
[95.47 114.94]
0.19
-0.431
(0.72)
-1045.25
0.3
(0.581)
103.47
[95.7 115]
0.19
-0.477
(0.82)
-1045.29
0.23
(0.63)
103.37
[95.15 114.63]
0.19
H4
-0.008
(23.21)**
-0.241
(2.40)*
-0.009
(0.09)
0.344
(7.23)**
0.222
(2.40)*
-0.003
(0.72)
-0.005
(0.38)
0.061
(5.05)**
0.214
(1.93)
-0.064
(0.73)
-0.221
(1.41)
-0.121
(0.82)
0.179
(1.68)
0.133
(3.69)**
-0.005
(1.07)
-0.002
(0.34)
-0.026
(0.93)
H5
-0.008
(23.31)**
-0.248
(2.46)*
-0.005
(0.05)
0.343
(7.23)**
0.220
(2.38)*
-0.003
(0.68)
-0.006
(0.47)
0.060
(5.04)**
0.214
(1.93)
-0.064
(0.73)
-0.214
(1.36)
-0.120
(0.81)
0.174
(1.62)
0.148
(4.05)**
-0.007
(1.38)
-0.001
(0.19)
-0.031
(1.07)
-1.132
(1.62)
-0.004
(1.45)
-0.336
(0.57)
-1044.36
2.1
(0.147)
103.51
[95.68 115.01]
0.19
-0.226
(0.37)
-1044.09
2.64
(0. 104)
103.55
[95.47 114.94]
0.19
H5
-0.008
(23.57)**
-0.239
(2.38)*
0.004
(0.04)
0.347
(7.35)**
0.223
(2.42)*
-0.003
(0.66)
-0.007
(0.50)
0.060
(5.04)**
0.218
(1.96)*
-0.057
(0.65)
-0.224
(1.43)
-0.119
(0.80)
0.177
(1.65)
0.143
(3.97)**
-0.009
(1.53)
-0.001
(0.15)
-0.022
(0.80)
-1.364
(1.61)
-0.289
(0.49)
-1044.11
2.58
(0.107)
103.47
[95.7 115]
0.19
H5
-0.008
(23.37)**
-0.236
(2.34)*
-0.014
(0.15)
0.355
(7.57)**
0.235
(2.55)*
-0.003
(0.65)
-0.009
(0.70)
0.059
(4.98)**
0.202
(1.82)
-0.060
(0.69)
-0.239
(1.53)
-0.169
(1.17)
0.187
(1.76)
0.137
(3.85)**
-0.004
(0.89)
-0.002
(0.48)
-0.019
(0.70)
-0.754
(1.66)
-0.279
(0.47)
-1044.02
2.77
(0.009)*
103.37
[95.15 114.63]
0.19