Abstract
Air pollution has been a major concern for people around the world, especially in urban areas of developing countries, such as Ha Noi city. Based on the choice experiment approach, this paper presents estimates of residents’ willingness-to-pay (WTP) for improving air quality of Ha Noi. Hanoi residents expressed their strong preferences for increase of green spaces and reduction of air pollution-related deaths. The mean marginal WTP for the increase of 1 m2 in per-capita tree cover is estimated at 2,256 VND per month; and for the reduction of 1 in 100,000 death related air pollution is about 1,865 VND per month. Hanoi residents appear to be willing to pay monthly 70,591 VND for the maximal improvements in air quality. This maximum amount of WTP accounts for about 0.5% of household income. The information on residents’ WTP for improving air quality would be useful for policy makers in investing effectively in controlling air pollution given the budget limitation.
Keywords: Air pollution, Choice experiment, Ha Noi
1. Introduction
Air pollution is one of the most serious problems in the world. According to World Health Organization (WHO), more than 80% of people in urban areas are living in an atmosphere with quality levels not satisfying the WHO recommended limits. Recent estimates by WHO show that ambient air pollution accounts for an estimated 4.2 million deaths per year. While ambient air pollution affects developed and developing countries alike, low- and middle-income countries experience the highest burden, with the greatest toll in the WHO Western Pacific and South-East Asia regions.1
- Air pollution in Hanoi
Rapid economic development poses a growing threat to environmental quality in Vietnam. Even at this early stage of development environmental pollution, especially air pollution is getting more severe in the big cities including Ha Noi, the capital city of Vietnam. Industrial production, increasing urbanization, and the rapid growth of individual vehicles are among the main factors contributed to intensity of urban air pollution in Vietnam (MONRE 2017).
The monitoring data of the Center for Environmental monitoring of Ministry of Environment and Natural Resources (MONRE) shows that, in the period from 2012 to 2016, air pollutants such as dust and particles at a number of locations in Ha Noi exceeded nationally stipulated standards for ambient air quality (QCVN05:2013) as seen in Fig.1-1. Like many other cities of Vienam, particulate matter (PM) is a major environmental problem of Hanoi (MONRE, 2017). The noise level is also persistently high. The concentration of other pollutants, such as NO2, SO2, CO in the air of Ha Noi have remained relatively stable in recent years and below the national standard mentioned above.
Figure 1. Air Pollution in Hanoi, 2012-2016
Dust pollution levels (µg/m3)
Source: Center for Environmental Monitoring of MONRE, various years.
PM2.5 pollution level in 1st quarter of 2018 (µg/m3)
Source: Green ID (2018)
Air pollution in Ha Noi is considered to be more serious when compared with other big cities in Vietnam (Luong et al., 2017). The WHO Global Ambient Air Quality Database (update 2018)1 shows that PM concentration in Ha Noi are usually higher than in other cities of Vietnam, such as Ho Chi Minh City, Da Nang and Ha Long. The annual concentration of PM2.5 (monitored by the Vietnam’s U.S. Embassy at 7 Lang Ha Street, Ha Noi) in 2016 reached 50.5 μg/m3, and in 2017 was 42.6 μg/m3 nearly twice as compared to the Vietnamese standard (25 μg/m3) and five times as recommended by WHO (10μg/m3) (GreenID, 2017, 2018). According to the Department of Natural Resources and Environment of Ha Noi city, 70% of air emissions are caused by traffic activities. Emissions from more than 4 million vehicles account for 85% of CO2 emissions and 95% of volatile organic compounds (Box 2.1, MONRE (2017))
- Health impact of air pollution
It is well known that people’s health is adversely affected by air pollutants from vehicles such as PM10 and PM2.5 (particulate matter less than 10 or 2.5 microns in diameter, respectively), nitrogen oxides, sulfur dioxide and carbon monoxide, as well as secondary pollutants such as ozone (WHO 2018). These cause respiratory problems, sinusitis, bronchitis, asthma, lung cancer, cardiovascular diseases and premature death. Particles have also been shown to increase the mortality rate. People with asthma and respiratory diseases in turn are highly susceptible to particles, nitrogen oxides, sulfur dioxide and ozone. In addition, lead particles have serious effects on children’s growth and development. Children with high lead levels in their blood are often deficient in weight and tend to have a low count of red blood cells. Their IQ levels on average are also lower than those with lower lead levels. In Vietnam and other countries in the region, air pollution is now acknowledged as a serious public health threat. WHO (2018) estimates that globally about 7 million people die every year from exposure to fine particles in polluted air that lead to diseases such as stroke, heart disease, lung cancer, chronic obstructive pulmonary diseases and respiratory infections, including pneumonia. Among them, about 90% are believed to be in Asia and Africa. According to the Vietnam Health Statistical Year Book 2015 (Ministry of Health, 2017), diseases of the respiratory system accounted for the highest numbers of both proportion morbidity and mortality by disease chapters (Ministry of Health, 2017).
The health effect of air pollution was first studied in Vietnam as early as in 1995 with a focus on traffic police officers (Dang, 1995). Due to extended exposure to high levels of air and noise pollution, 2.9% of traffic policemen were infected with tuberculosis, compared with an average infection rate of 0.075%. Moreover, 76% of traffic policemen suffered from ear, nose, and throat infection, and 32% of them had reduced hearing ability. Separately, National Institute of Occupational and Environmental Health (NIOEH) conducted a study on the health impacts of air pollution in 2005 (NIOEH, 2005). It showed that 83.1% of the respondents suspected that dust pollution came from transportation. Examination of persons who worked more than 8 hours per day on roadside found a significant difference in the health conditions between targeted and reference groups.
The health effects of air pollution to Hanoi citizens are considered to be serious. Hieu et al. (2013) estimated the number of deaths due to PM10 pollution from traffic in 2009 was 3200 people, greater than the number of deaths from traffic accidents. Luong et al. (2017) showed that in the period of 2010-2011, if the PM10, PM2.5 concentration increased to 10μg/m3, the number of children hospitalizations related to the respiratory diseases in Hanoi increased by 1.4% and 2.2%, respectively.
To cope with this situation, the Government of Viet Nam in June 2016 has issued the National Action Plan on Air Quality Management until 2020 with the main goal of strengthening air quality management based on controlling emission sources and monitoring ambient air quality. In recent years, Hanoi’s Government also has made efforts to implement measures for improving air quality such as cleaning dust on trucks before entering the city, installing additional air monitoring stations, planting one million trees in the period of 2016-2020. However, air pollution is still a major concern of the Hanoi citizens, demanding for more effective solutions to improve air quality.
The aim of this paper is to estimate households’ willingness-to-pay (WTP) for improvements in air quality in Ha Noi by using the choice experiment approach. Such information is important for policy makers when determining public investments and policy instruments in order to effectively improve air quality in Ha Noi city.
2. Choice experiment design and implementation
2.1. Theoretical framework for the choice experiment valuation
Air quality is a non-market commodity, so that market prices are not available to measure users’ WTP. Instead, non-market valuation method – measuring the monetary value of changes in individual welfare associated with the change in environmental quality – should be applied. Questionnaire surveys were conducted using choice experiment approach, a stated preference method, which involves the construction of a hypothetical market to obtain and analyse respondents' choices of an improved cyclone warning service.
Choice experiment (CE) has its roots in conjoint analysis where individuals make choices between multi-attribute goods and services (Adamowicz et al., 1994; Boxall et al., 1996; Adamowicz et al., 1998; Alriksson and Öberg, 2008). In a CE survey, individuals are requested to decide over a series of choice sets. Each choice set includes a number of alternatives, which are described by different levels of the attributes or characteristics of the good or service that is being valued. In choosing between the alternatives, the individuals also make a trade-off between the levels of the attributes. If a monetary (cost) attribute is included in the choice sets, the researchers can estimate the individual’s marginal willingness to pay for a change in each of the other non-market attributes.
CE is an application of Lancaster’s theory of value, combined with random utility theory (Hanley et al., 1998; Wang et al., 2007). According to Lancaster’s theory, individuals’ choices are determined by the utility or value that is derived from the attributes of the goods and services rather than directly from the goods and services themselves (Lancaster, 1966). CE is also based on the behavioral framework of random utility theory (RUT), which describes discrete choices in a utility maximizing framework. The researchers are able to observe only part of individuals’ utility, and the unobserved component is randomly distributed. Under the RUT, Uin, utility that individual n enjoys from choice alternative i can be decomposed into two parts:
Uin = Vin + εin (1)
where Vin is the systematic and observed component of the choice utility; and εin is the stochastic unobserved component.
The observed component of the choice utility can be disaggregated further, as utility can depend on the choice attributes (Zin) that may be viewed differently by different individuals and the characteristics of the individual (Sn). Then equation (1) can be rewritten as follows:
Uin = V(Zin, Sn) + εin (2)
Alternative i is chosen over some other option j if and only if Ui > Uj. Due to the unobserved component, the researchers are unable to predict choices perfectly. This uncertainty is expressed in terms of choice probability, and the probability that individual n will choose option i over other options j in choice set t is given by:
Prob(i | t) = Prob(Vin + εin > Vjn + εjn; all j Î t and j ¹ i) (3)
The individual's indirect utility function (Vi) in Equation (2) for a choice option can be modelled with various specifications. If assuming that the relationship between the utility and attributes of the choice is linear such that V = βZin, and that only the main effects are considered, the functional form of the indirect utility function is as follows:
Vin = βi + ΣkβkZkn + ΣpθpSpn (4)
where:
βi is vector of constant terms (alternative specific constants) for i = 1,……, I choice options;
βk is a vector of coefficients attached to the vector of attributes (Zkn) for k = 1, …., K;
θp is a vector of coefficients attached to the vector of respondent’s characteristics (Spn) for p = 1,….,P.
The utility function estimated for each alternative, therefore, contains a unique alternative specific constant (ASC), the effects of a choice’s attributes, and the individual’s characteristics. The ASCs represent the average effect on choices of any variation that cannot be explained by the observed attributes or the socio-economic characteristics.
With the assumption of linear indirect utility function, compensating surplus (CS) welfare estimates may be obtained in the following formula (Hoyos, 2010):
(5)
where α is the marginal utility income (represented by the β coefficient of the cost attribute), and V0n and V1n are indirect utility functions before and after a specified change in the non-market good or service, respectively.
The marginal benefit of an improvement on a single attribute can be estimated by the ratio of coefficient given in equation 6 (Hanley et al., 2001):
(6)
The above ratio is usually known as the implicit price of the non-market attribute. It shows the trade-off made between the non-market attribute and the cost attribute, and an estimate of the individual’s willingness to pay for a unit change in the non-market attribute (Bergmann et al., 2006).
2.2. Survey design and implementation
The choice experiment design
The design of choice experiment includes decisions about attributes and their levels, the design of choice tasks, and questionnaire design. The attributes are used to describe to the respondents a storm early warning service. For the estimated utility function of users, the attributes will be the observed independent variables. The appropriate selection of attributes is a critical component in a CE exercise, since the selected attributes affect respondents’ choices, as well as the policy under concern. Having defined the attributes, the levels of these attributes must be determined. Levels can be expressed qualitatively or quantitatively, and the quantitative attributes can be presented in absolute or relative terms (Bennett and Blamey, 2001). In this part of a CE exercise, a series of focus group studies should be conducted with the aim of selecting the relevant attributes and levels (Alpízar et al., 2001). The focus studies could be in the form of verbal group discussion or actual surveys. In order to obtain contrasting opinions and to obtain a representative sample of the population, the focus group composition should be heterogeneous in terms of occupation, background, age and gender (Suh, 2002).
The most notable disadvantage of CE approach is the cognitive burden associated making choices between bundles of attributes and levels. The larger the number of attributes and the levels, the bigger the cognitive burden that the respondents face. The solutions for this stage of survey design are to carefully select attributes and choose the optimal number of attributes (DeShazo and Fermo, 2002). One important lesson, learned from reviewing the previous studies, is that most CE studies in environmental and meteorological valuation have used 4-5 attributes including the cost attribute in each choice set.
Health Risk related to air pollution. Following the above instructions and lessons from the literature, this research started by studying the attributes and attribute levels used in previous studies. A key lesson is that attributes related to health effects of air pollution have been commonly selected in the design of CEs for air quality improvements (Yoo et al., 2008; Rizzi et al., 2014). This selection is reasonable, since many epidemiological studies have indicated that air pollutants such as particulate matter (PM), nitrogen dioxide (NO2), sulphur dioxide (SO2), and ozone (O3) are responsible for increasing mortality and morbidity in different populations around the world, especially from respiratory and cardiovascular diseases (CVD) (Phung et al., 2016).
Attributes selected should be both relevant and understandable to respondents. To collect residents’ desire for air quality improvements, two focus studies were conducted in the form of an internet survey with 191 respondents and 212 face-to-face interviews in Hanoi city. In the surveys, respondents were presented a list of measures, which were designed based on a rigorous review of international experiences and the Government’s plans on controlling air emission sources in order to improve urban air quality. Then, we asked respondents to choose their preferred measures that should be implemented at high priority to improve air quality of Ha Noi city. The most preferred measure chosen by more than 70% of respondents is the increase of green spaces.
The effects of tree on urban air pollution. In recent years, researchers have been looking into potential benefits of green space and vegetation, including temperature reduction and other microclimatic effects, removal of air pollutants, emission of volatile organic compounds and tree maintenance emissions and energy effects on buildings etc. Reduced air pollution was acknowledged and developed by several authors such as Antoine et al. (2017), Wissal et al. (2016), The Nature Conservancy (2016), Beckett et al. (2000) and Lovett (1994) among others. Authors generally agree that the use of urban vegetation is often promoted as an effective measure to reduce air pollutants concentrations. This measure is based on the underlying argument that trees (and vegetation in general) have the capability of cleaning the air by filtering out the pollutants. Different studies of Antoine et al. (2017), Wissal et al. (2016) have experimentally assessed the deposition rate at which pollutants are taken up by the urban vegetation and showed that trees trap air pollution by up to about 7%.
Based on the focus studies, in concert with an in-depth literature review, the proposed attributes and their levels are presented in Table 1. Having attributes and levels determined, an orthogonal choice task design was used, resulting in eighteen choice tasks. In order to reduce the cognitive burden on respondents, each respondent was randomly chosen to face a block of nine choice tasks. For each choice task, respondents indicated their preference between two alternatives: one potential improvement program and the status quo (i.e., keeping all levels at their current levels). The status quo option remains identical across tasks. An example of a choice task is presented in Figure 2.
Table 1: Attributes and levels
Attributes
|
Current levels
|
Improvement levels
|
Health Risk related to air pollution:
|
Out of 100,000 people:
|
People who get hospitalised due to air pollution-related diseases
People who die from air pollution-related diseases
|
200 people
36 people
|
200; 150; 100 people
36; 27; 18 people
|
Tree cover area
|
8 m2 per capita
|
10; 14; 18 m2 per capita
|
Change to household electricity bill, starting in 2020
|
No change
|
Increase of
5; 15; 25 thousand VND/month =
60; 180; 300 thousand VND/year
|
The structure of the questionnaire included three main components as follows:
+ The survey started with questions about respondents’ perception of air pollution in Ha Noi. The overview of the Government’s plans on improving air quality was presented to respondents. A list of relevant measures to improve air quality was also presented to respondents, who then were requested to make choices of three most preferred measures. Benefits of implementing those measures were discussed with respondents, and they also were requested to choose three types of benefits which are most relevant to their situations.
+ After being aware of benefits of improvements in air quality, respondents were requested to make choices in the series of nine choice tasks in Part 2. Respondents also were reminded about their budget constraints when making their choices. Follow up-questions were included to identify anomalies in the responses, such as reasons for zero bids, hypothetical bias (respondents may agree to pay because the payment is hypothetical).
+ The final part of the questionnaire collects respondents’ socio-economic and attitudinal information to analyse the factors affecting the WTP for improvements in air quality in Ha Noi.
Figure 2: An example of a choice task
Survey implementation
In January and February 2019, an online survey was designed using the Google Forms[1]. Residents of Hanoi city were invited to participate in the online survey. The survey was completed by 161 respondents. Table 2 presents a summary of the socio-economic characteristics of the sample in our CE exercise.
Table 2: Socio-economic characteristics of the surveyed sample
Socio-economic characteristics
|
Mean
|
Perception of air pollution levelsa
Perception of air pollution impactsa
Age (years)
Percentage of college degree
Monthly electricity bill (thousand VND per household)
|
4.07
4.06
33
92.5
780
|
Sample size
|
161
|
a Perception was measured using 5-point Likert scales ranging from very high (5) to very low (1)
3. Willingness-To-Pay for improving air quality in Hanoi city
To estimate the WTP for improving air quality in Hanoi, conditional logit (CL) model was developed using NLOGIT 5.0 software. Results of the CL model are presented in Table 3. The sign of all attribute variables in Table 3 confirming the prior expectation that the likelihood of choosing an improvement program decreases as the increased payment in the electricity bill rises; the likelihood of choosing an improvement program decreases as numbers of air pollution-related illness and death increase; and as tree cover area rises, the likelihood of supporting an improvement program increases. All variables are statistically significant at the 1% or 5% level, confirming effects of the independent variables on residents’ choices of supporting the improvements in air quality.
Table 3: Results of conditional logit model
Attribute variables
|
Coefficient (β)
|
Standard deviation
|
P-value
|
Air pollution-related Illness
|
-0.00313**
|
0,013
|
0.013
|
Air pollution-related Death
|
-0.04042***
|
0,007
|
0,000
|
Tree cover area
|
0.04889***
|
0,014
|
0,001
|
Change to household electricity bill
|
-0.02166***
|
0,006
|
0,000
|
Notes: *** = Significance at 1% level ; ** = Significance at 5% level
Mean marginal and total WTP per household for improvements in air quality in Hanoi city are reported in Tables 4 and 5, respectively. The total WTP was estimated for two improvement programs described as follows:
+ Medium improvement: number of people hospitalised 150/100,000, number of people died 27/100,000 and tree cover area of 14 m2 per capita.
+ Maximal improvement: number of people hospitalised 100/100,000, number of people died 18/100,000 and tree cover area of 18 m2 per capita.
Table 4: Mean marginal WTP estimates in thousand VND per month
Attribute variables
|
Coefficient (β)
|
Standard deviation
|
Reduction in air pollution-related Illness
|
0.144**
|
0.069
|
Reduction in air pollution-related Death
|
1.865***
|
0.660
|
Tree cover area
|
2.256***
|
0.722
|
Notes: *** = Significance at 1% level ; ** = Significance at 5% level
Table 5: Total WTP estimates in thousand VND per month
Program
|
Coefficient (β)
|
Standard deviation
|
Medium improvement
|
37.552***
|
10.930
|
Maximal improvement
|
70.591***
|
20.666
|
Notes: *** = Significance at 1% level
4. Conclusions
In this paper, residents’ WTP for improving air quality of Ha Noi was estimated. Hanoi residents expressed their strong preferences for increase of green spaces and reduction of air pollution-related death. The mean marginal WTP for the increase of 1 m2 in per-capita tree cover is estimated at 2,256 VND per month; and for the reduction of 1 in 100,000 death related air pollution is about 1,865 VND per month. Hanoi residents appear to be willing to pay monthly 70,591 VND for the maximal improvements in air quality. This maximum amount of WTP accounts for about 0.5% of household income. The estimate in our study is similar to the WTP values of 0.4-0.7% of household income, which were estimated for improvements in air quality in some Chinese cities (Wang and Mullahy, 2006; Wang et al., 2006; Wang and Zhang, 2009).
Acknowledgement
This research is funded by Vietnam National Foundation for Science and Technology Development (NAFOSTED) under grant number 502.99-2017.14.
Nguyen Cong Thanh1, Le Ha Thanh2
Faculty of Environmental, Climate Change and Urban Studies,
National Economics University, Vietnam
1 E-mail: thanhnc@neu.edu.vn, 2 E-mail: lehathanhneu@gmail.com
(International Conference ICSEED2019)
References
- Adamowicz, W., P. Boxall, M. Williams and J. Louviere (1998). "Stated Preference Approaches for Measuring Passive Use Values: Choice Experiments and Contingent Valuation." American Journal of Agricultural Economics 80(1): 64-75.
- Adamowicz, W., J. Louviere and M. Williams (1994). "Combining revealed and stated preference methods for valuing environment amenities." Journal of Environmental Economics and Management 26(3): 271-292.
- Alpízar, Francisco, Fredrik Carlsson and Peter Martinsson (2001). "Using Choice Experiments for Non-Market Valuation." Economic Issues 8(1): 83-110.
- Alriksson, Stina and Tomas Öberg (2008). "Conjoint analysis for environmental evaluation." Environmental Science and Pollution Research 15(3): 244-257.
- Antoine P.R. Jeanjean, Riccardo Buccolieri, James Eddy, Paul S. Monks, Roland J. Leigh (2017).“Air quality affected by trees in real street canyons: The case of Marylebone neighbourhood in central London”. Journal of Urban Forestry & Urban Greening 27(2017): 41-53
- Beckett, K.P., Freer-Smith, P., Taylor, G. (2000). Effective tree species for local air quality management. Journal of Arboric. 26, 12-19
- Bennett, Jeff and Russell K. Blamey (2001). The choice modelling approach to environmental valuation. Cheltenham, UK ; Northampton, MA, USA: E. Elgar Pub.
- Bergmann, Ariel, Nick Hanley and Robert Wright (2006). "Valuing the attributes of renewable energy investments." Energy Policy 34(9): 1004-1014.
- Boxall, Peter C., Wiktor L. Adamowicz, Joffre Swait, Michael Williams and Jordan Louviere (1996). "A comparison of stated preference methods for environmental valuation." Ecological Economics 18(3): 243-253.
- DeShazo, J. R. and German Fermo (2002). "Designing Choice Sets for Stated Preference Methods: The Effects of Complexity on Choice Consistency." Journal of Environmental Economics and Management 44(1): 123-143.
- GreenID (2017). GreenID: Chất lượng không khí Việt Nam 2016. Trung tâm Phát triển Sáng tạo Xanh (GreenID), Hà Nộ.
- GreenID (2018). GreenID: Chất lượng không khí Việt Nam 2017. Trung tâm Phát triển Sáng tạo Xanh (GreenID), Hà Nộ.
- Hanley, Nick, Susana Mourato and Robert E. Wright (2001). "Choice Modelling Approaches: A Superior Alternative for Environmental Valuation?" Journal of Economic Surveys 15(3): 435-462.
- Hanley, Nick, Robert Wright and Vic Adamowicz (1998). "Using Choice Experiments to Value the Environment: Design Issues, Current Experience and Future Prospects." Environmental and Resource Economics 11(3): 413-428.
- Hieu, Vu Van, Le Xuan Quynh, Pham Ngoc Ho and Luc Hens (2013). "Health Risk Assessment of Mobility-Related Air Pollution in Ha Noi, Vietnam." Journal of Environmental Protection 4(10): 1165-1172.
- Hoyos, David (2010). "The state of the art of environmental valuation with discrete choice experiments." Ecological Economics 69(8): 1595-1603.
- Lancaster, Kelvin J. (1966). "A New Approach to Consumer Theory." The Journal of Political Economy 74(2): 132-157.
- Lovett, G.M. (1994). Atmospheric deposition of nutrients and pollutants in North America: an ecological perspective. Journal of Ecol. Appl. 4, 629-650.
- Luong, Ly M. T., Dung Phung, Peter D. Sly, Lidia Morawska and Phong K. Thai (2017). "The association between particulate air pollution and respiratory admissions among young children in Hanoi, Vietnam." Science of The Total Environment 578: 249-255.
- MONRE (2017). National State of Environment 2016: Urban Environment. Ministry of Natural Resources and Environment, Hanoi (in Vietnamese).
- Ministry of Health (2017). Vietnam Health Statistical Year Book (2015), Health publishing house, Hanoi, Vietnam
- NIOEH (2005). Air pollution and health impact. NIOEH research report. Hanoi. Vietnam.
- Phung, Dung, To Thi Hien, Ho Nhut Linh, Ly M. T. Luong, Lidia Morawska, Cordia Chu, Nguyen Duy Binh and Phong K. Thai (2016). "Air pollution and risk of respiratory and cardiovascular hospitalizations in the most populous city in Vietnam." Science of The Total Environment 557-558: 322-330.
- Rizzi, Luis Ignacio, Cristóbal De La Maza, Luis Abdón Cifuentes and Jorge Gómez (2014). "Valuing air quality impacts using stated choice analysis: Trading off visibility against morbidity effects." Journal of Environmental Management 146: 470-480.
- Suh, Jungho (2002). Estimation of Non-market Forest Benefits Using Choice Modelling.Harrison, S. R., Herbohn, J. L. and Mangaoang, E. O., Socio-Economic Research Methods in Forestry: A Training Manual Leyte State University, ViSCA, Baybay, The Philippines, 149-164.
- The Nature Conservancy (2016) Planting Healthy Air: A global analysis of the role of urban trees in addressing particulate matter pollution and extreme heat . www.nature.org/healthyair
- Wang, Hong and John Mullahy (2006). "Willingness to pay for reducing fatal risk by improving air quality: A contingent valuation study in Chongqing, China." Science of The Total Environment 367(1): 50-57.
- Wang, X. J., W. Zhang, Y. Li, K. Z. Yang and M. Bai (2006). "Air quality improvement estimation and assessment using contingent valuation method, a case study in Beijing." Environ Monit Assess 120(1-3): 153-168.
- Wang, Xuehong, Jeff Bennett, Chen Xie, Zhitao Zhang and Dan Liang (2007). "Estimating non-market environmental benefits of the Conversion of Cropland to Forest and Grassland Program: A choice modeling approach." Ecological Economics 63(1): 114-125.
- Wang, Yan and Yi-Sheng Zhang (2009). "Air quality assessment by contingent valuation in Ji'nan, China." Journal of Environmental Management 90(2): 1022-1029.
- Wissal Selmi, Christiane Weber, Emmanuel Rivière, Nadège Blond, Lotfi Mehdi, David Nowak (2016). “Air pollution removal by trees in public green spaces in Strasbourg city, France”. Journal of Urban Forestry & Urban Greening 17(2016): 192-201
- Yoo, Seung-Hoon, Seung-Jun Kwak and Joo-Suk Lee (2008). "Using a choice experiment to measure the environmental costs of air pollution impacts in Seoul." Journal of Environmental Management 86(1): 308-318.