Article

Cite This: Environ. Sci. Technol. 2018, 52, 13824−13833 pubs.acs.org/est

Quantitative Analysis of Health Risk Perception, Exposure Levels,

and Willingness to Pay/Accept of PM2.5 during the 2014 Nanjing
Youth Olympic Games
Lei Huang,*,†,‡ Jie Li,† Ruoying He,† Chao Rao,† Tsering J. van der Kuijp,§ and Jun Bi*,†
†
State Key Laboratory of Pollution Control & Resource Reuse, School of the Environment, Nanjing University, Nanjing 210023,
China
‡
Lamont-Doherty Earth Observatory, Columbia University, P.O. Box 1000, 61 Route 9W, Palisades, New York 10964, United States
§
Department of Environmental Science and Public Policy, Harvard University, Cambridge, Massachusetts 02138, United States
See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

*
S Supporting Information

ABSTRACT: Local governments in China regularly imple-

Downloaded via NAGOYA UNIV on January 14, 2019 at 11:03:59 (UTC).

ment short-term emission control measures to improve air

quality during important sporting events. As a condition for
hosting the 2014 Youth Olympic Games (YOG), the Nanjing
government agreed to temporarily and substantially improve
air quality. Regression analysis, Spearman correlation analysis,
χ2 test, and the contingent valuation method were used to
explore the effects of robust, short-term air pollution control
measures on risk perception, daily exposure to PM2.5, risk
acceptance levels, and willingness to pay/accept (WTP/
WTA) for reductions in air pollution for the benefit of
reducing health risks. Postimplementation, the respondents’
risk perception levels presented the following changes: during the YOG, the respondents perceived the lowest effects of haze
pollution while after the YOG, they perceived the highest effects. The changes in risk acceptance levels showed the same
tendency. Furthermore, after the YOG, the respondents asked for the most economic compensation, and their willingness to
pay for risk reduction also reached the highest level. This study reveals the need to increase the public’s understanding of the
health risks of air pollution, protect those populations most exposed to high levels of PM2.5, and take more effective long-term
measures to meet local residents’ demands for improved air quality.

1. INTRODUCTION 2014 YOG, the local government of Nanjing agreed to

China has been suffering from hazardous air pollution due to
the rapid development of its economy and heavy industry over
the past several decades. Haze pollution accompanied by high
concentrations of particulate matter (particularly PM2.5, which
has an aerodynamic particle diameter of less than 2.5 μm) in
the atmosphere have come under increased regulatory scrutiny
due to their adverse impacts on the environment and public
health.1−4 Severe haze pollution occurs frequently in city-
clusters such as the Beijing-Tianjin-Hebei Region, the Yangtze
River Delta (YRD), and the Pearl River Delta. Nanjing is one
of the fastest growing megacities in the YRD, as a result,
continued industrialization has increased air pollution and
induced frequent haze in Nanjing.
Previous studies have linked reductions in national or
regional air pollution to governmental policies, national
political realignments, and even large-scale sporting events.5−7
Robust, short-term emission control measures have frequently
been implemented to improve air quality during important
sporting events in China, most notably the 2008 Beijing
Olympic Games, the 2010 Guangzhou Asian Games, and the
2010 Shanghai World Expo.8−11 As a condition for hosting the
temporarily and substantially improve air quality in Nanjing for
the YOG. During the YOG from August 16−28 of 2014,
Nanjing met its commitment to holding a “green youth
Olympics”, and the number of days achieving the rate of
“good” air quality accounted for 93.5% in YOG period. “Good”
means that the air pollution index (API) is 0−50 based on the
laws promulgated by the Ministry of environmental protection
of China in 2012. In this case, the air quality is suitable for all
segments of a population to perform outdoor activities. After
the YOG, the Nanjing municipal government also made efforts
to attain the air pollution control goals of the Action Plan
issued by the Chinese State Council. Assessments of the effects
of these short-term intervening measures can provide critical
insights into the main sources of air pollution as well as how to
design the most effective environmental policies.12−16
Received: March 27, 2018
Revised: October 14, 2018
Accepted: October 16, 2018
Published: October 16, 2018

© 2018 American Chemical Society 13824 DOI: 10.1021/acs.est.8b01634

Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology
Figure 1. Study routes and framework.
Figure 2. Location of study areas and sample size.
Numerous health studies have indicated that the impacts of
air quality improvements during large-scale sporting events
have been associated with beneficial physiological health
effects, including reductions in acute myocardial infarction
rates17 and systematic inflammation,18,19 as well as improve-
ments in cardio-respiratory health.20 Other studies have found
that changes in certain behaviors were mostly affected by
individual ’s varying perception of smog pollution.21,22
Individuals with higher perceived concerns, greater knowledge
of smog pollution, and higher perceptions of its health risks
were more likely to view air pollution as an unacceptable risk.23
However, to the best of our knowledge, no studies have
13825
Article
focused on the effects of powerful, short-term air pollution
control measures on health risk perception, risk acceptance
levels, willingness to pay (WTP) for the measures of helping
reducing health risks to the population and willingness to
accept (WTA) compensation for the health loss. Effective
intervention strategies that seek to promote certain behaviors
must first address individuals’ subjective judgments regarding
air pollution and its related health risks.24 Therefore, the
unusually low levels of pollutants that arose postimplementa-
tion and favorable weather conditions during the YOG offered
a unique opportunity to combine an assessment of the YOG
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology
emission reduction measures with a quantitative analysis of
affected residents’ perceptions and behaviors.
This paper was set to explore changes in public attitudes,
including their health risk perception, acceptable risk levels of
air pollution, and their willingness to pay/accept for reductions
in air pollution for the benefit of reducing health risks before,
during, and after the YOG. Previous studies have established
the correlation between people’s behaviors and their
opinions;25 we therefore set out to determine whether this
correlation exists between people’s perception of the health
risks of air pollution and their average daily PM2.5 exposure.
Moreover, socioeconomic backgrounds can also alter respond-
ents’ cognition of air pollution and their behaviors.26 Thus, the
influencing factors of health perception were also explored.
The conceptual framework of our study is shown in Figure 1.
2. METHODOLOGY
2.1. Study Site. As the capital of Jiangsu Province, Nanjing
encompasses approximately 6600 km2, with a population of
more than 8.2 million in 2013. The study area and sampling
site of Nanjing are depicted in Figure 2.
2.2. Sample Selection. The respondents were recruited in
public places of residential area, such as parks and markets by a
stratified random sampling of those living in seven districts in
Nanjing. Senior students in Nanjing University who had been
well trained in survey techniques interviewed all respondents
face-to-face. The research was approved for human subjects by
institutional review board of Nanjing University. Three surveys
each administered to 250 adults were conducted before, during
(August 16−28, 2014), and after the YOG in Nanjing in 2014.
The first round of surveys was administered in January 2014,
with a total of 218 questionnaires returned (87.2% response
rate). The second round of surveys was conducted in August
2014, with 228 questionnaires returned (86.8% response rate).
The third round of surveys was administered in October 2014,
with 217 questionnaires returned (87.6% response rate).
2.3. Questionnaire Design. The questionnaire was
designed based on psychometric paradigm methods,27−29
with minor modifications based on Chinese residents’
circumstances. The questionnaire comprised five parts (see
Supporting Information, SI). The first part included 11
questions to measure health risk perception regarding air
pollution. The response to each question was ranked on a 5-
point Likert-type scale ranging from “1 = minimum” to “5 =
maximum”. The second part included an introduction to the
Action Plan, pictures showing the city landscape under
different PM2.5 concentrations, and questions measuring the
respondents’ acceptance of the PM2.5 concentration reductions.
The third part investigated the respondents’ daily time-activity
patterns. They were asked to recall their activities over the past
24 h in detail. The fourth part was a payment card to
investigate the respondents’ willingness to pay/accept in
response to the haze pollution. The last part of the
questionnaire was designed to collect the respondents’
demographic characteristics, including body weight (BW),
age, gender, education, income, marital status, smoking status,
previous experience with severe haze pollution (referenced to
SI Picture S3), preventative actions taken, and sources of air
pollution information.
2.4. Data Analysis. Principal component analysis and
independent-samples t tests were used for comparative analysis
of risk perception factors. Regression analysis and Spearman
correlation analysis were used to explore the effects of robust,
13826
Article
short-term air pollution control measures on health risk
perception and daily exposure to PM2.5. In addition, the
contingent valuation method was used to assess the amount
that respondents were willing to pay for air quality improve-
ments and willing to accept in compensation for the excess
deaths caused by air pollution.
2.4.1. Risk Perception Analysis. Risk perception factors
comprised three categories defined as Ef fect, Familiarity, and
Trust. The domain in questions 3−7 corresponds to Ef fect, the
domain in questions 1 and 2 corresponds to Familiarity, and
the domain in questions 8−11 corresponds to Trust. For
example, the items of domain “Ef fect” are as follows: Are the
effects of the risk associated with haze immediate or will they
take place in the future? Is the risk associated with haze a
common risk or a terrible risk? Are you familiar with health
risks associated with haze? How severe is the haze in your
residence? How severely will haze impact your health? The
other items are detailed in the SI. Confirmatory factor analysis
was conducted using Lisrel 8.70 software, and the independ-
ent-samples t tests conducted using SPSS 22.0 software were
used for comparative analysis of risk perception factors before,
during, and after the YOG.
2.4.2. Regression Analysis. Regression models were
employed to explore what factors influence the public’s
perception of air pollution effects. The independent variables
included individual characteristics. Age was defined as a
continuous variable. Education was divided into six groups: no
formal education = 1, primary school = 2, middle school = 3,
high school = 4, college = 5, and postgraduate and above = 6.
Monthly income was divided into nine ranges (Chinese Yuan,
or CNY): < 4000 = 1, 4000−12 000 = 2, 12 000−20 000 = 3,
20 000−100 000 = 4, and >100 000 = 5. Other binary variables
included gender, smoking, marriage, experience with severe
haze, harm experienced due to haze, and chronic diseases.
2.4.3. Daily PM2.5 Exposure. The respondents’ daily PM2.5
exposure levels were calculated using eq 1, which is
recommended by the U.S. Environmental Protection Agency
(U.S. EPA).30 The equation links time-activity patterns to the
exposure medium concentration normalized by weight.
ADD = (C1 × IR1 × EF/24 + C2 × R × IR 2 × EF2 /24)/BW (1)
1
where ADD is the average daily PM2.5 exposure (μg/kg·d); C1
is the average ambient PM2.5 concentration (μg/m3) of each
outdoor activity time; C2 is the average ambient PM2.5
concentration (μg/m3) of each indoor activity time; IR1 is
the respiratory rate of the outdoor activities (m3/d); IR2 is the
respiratory rate of the indoor activities (m3/d); EF1 is the
outdoor exposure time (h); EF2 is the indoor exposure time
(h); R is the ratio of indoor to outdoor PM2.5 concentrations;
and BW is the body weight of the respondent (kg).
The respiratory rates of outdoor and indoor activities for
males are 15.8m3/d and 15.1m3/d, respectively, and 14.1m3/d
and 13.5m3/d, respectively, for females.31 The ratio of indoor
and outdoor PM2.5 concentrations ranges from 1.01 to 1.08
according to Morawska et al.’s study, which has been
frequently referenced in other studies.32 The real-time ambient
PM2.5 concentrations were collected from the China Environ-
mental Monitoring Centre (CEMC) (http://113.108.142.
147:20035/emcpublish/), while outdoor and indoor exposure
times were collected from the questionnaire surveys. On the
basis of these variables, we estimated individual ADD levels,
and the average estimated ADD for all respondents represents
the general public’s ADD. We then stratified individual
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology Article

characteristics by gender, age, education, and income to

examine the variations in PM2.5 exposure among different
groups of respondents. One-way ANOVA analysis was used to
do so.
2.4.4. Public Risk Acceptance Levels. A series of questions
were set to evaluate the respondents’ opinions on acceptable
levels of PM2.5. Three pictures showed the city landscape under
the following scenarios: clean day (PM2.5 < 30 μg/m3), hazy
day (PM2.5 = 30−90 μg/m3), and severe hazy day (PM2.5 =
90−210 μg/m3). The respondents were asked to rate their
acceptance at five levels: “fully accept,” “easy to accept,”
“basically accept,” “hard to accept,” and “do not accept” (see
SI). According to the Action Plan issued by the Chinese State
Council, which required PM2.5 concentrations to be decreased
by 20% in 2017 compared with the mean level in the YRD in
2012 (67 μg/m3), we created two scenarios (2017 and 2022)
and seven different degrees of PM2.5 concentration reductions
for both scenarios. A scatter diagram with function fitting was
employed to calculate the median (50%) of the respondents’
acceptance levels based on the seven degrees of PM2.5
concentration reductions. The results were analyzed using
Sigma Plot 10.0 (Sigma Plot Software Inc.).
2.4.5. Respondents’ Willingness to Pay and Willingness to
Accept Compensation in Response to Haze Pollution. A
payment card was designed to investigate the respondents’
willingness to pay and willingness to accept. The respondents
were asked to choose their maximum willingness to pay for
three kinds of interventions that could protect their health
from air pollution as well as their willingness to accept
compensation for losses caused by air pollution. (See SI Part S-
IV). The contingent valuation method was used to assess the
amount that people were willing to pay for air quality
improvements and willing to accept in compensation for the
excess deaths caused by air pollution. The contingent valuation
method is based on constructing a hypothetical scenario that
investigates how people respond to changes in environmental
quality and how much they may pay for those changes, or their
willingness to accept compensation for the loss of this
good.33−36
The number of deaths caused by air pollution and the
corresponding PM2.5 concentrations were calculated according
to the following eq 2:37
case total = (caseap/((RR − 1) × E)) + caseap (2)

where casetotal is the total number of deaths of the target

population; caseap is the number of deaths caused by air
pollution; RR is the relative risk caused by the unit
concentration changes of pollutants; E is the target
population’s exposure concentration to PM 2.5. The casetotal
is the total number of nonaccidental deaths in 2012 in Nanjing,
which is calculated based on the baseline mortality rate of
nonaccidental death according to Jiangsu Provincial Center for
Disease Control and Prevention and the total number of death
according to Nanjing Statistical Yearbook in 2012. The RR
value was set to 1.0036 according to Xiaochuan et al.37
3. RESULTS AND DISCUSSION
3.1. Demographics. As shown in Table S1, the
demographic characteristics of the respondents were similar
to that of the local population in Nanjing. These characteristics
include gender, age, and monthly income, but exclude
education. The respondents were slightly more educated
13827
Figure 3. Regression analysis of influencing factors on health
perception (β is the regression coefficient for the relevant variables;
*p < 0.05, **p < 0.01).
than the rest of the city, likely attributable to the fact that the
undereducated are more likely to have difficulty understanding
survey questions and completing a questionnaire easily.
3.2. Comparative Analysis and Determining Factors
of Health Risk Perception. Figures S1, S2, and S3 present
the path diagram of the revised final confirmatory factor
analysis model on health risk perception of air pollution before,
during, and after the YOG. The goodness-of-fit result of the
confirmatory factor analysis can be seen in Table S2. The
values of χ2/df, GFI, AGFI, IFI, and CFI largely aligned with
the reference standard, indicating that the model aligned well
with the date.
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology Article

Figure 4. ADD levels during three periods.

Table 1. Demographic Stratification Analysis of ADD

one-way ANOVA
demographic characteristics stratification mean ADD μg/kg·d F(gag)
before YOG during YOG after YOG before YOG during YOG after YOG
gender male 34.16 4.67 16.54 1.616 18.258 56.171
female 35.72 5.26 19.73 (0.205) (0.000) (0.000)
age <20 36.85 4.88 17.43 1.262 (0.282) 18.971 (0.000) 11.362 (0.000)
20−29 35.53 5.24 18.37
30−39 33.06 5.20 19.68
40−49 33.77 5.09 16.99
50−60 34.57 4.22 18.00
>60 31.81 3.44 13.86
education below high school 37.77 5.27 19.33 2.396 (0.069) 1.717 (0.164) 4.338 (0.005)
high school 33.37 5.07 18.56
college 35.23 4.83 17.32
postgraduate 31.35 5.16 16.26
monthly income (yuan) <4000 35.13 4.79 19.50 0.180 (0.948) 1.162 (0.329) 11.127 (0.000)
4000−12 000 34.84 5.04 16.96
12 000−20 000 34.06 5.33 16.67
20 000−100 000 33.47 5.11 15.59
>100 000 34.30 5.29 15.18

Public risk perception toward PM2.5 is shown in Table S3. In
addition, we compared the differences in health risk perception
during the three periods through an independent-sample t-test
analysis. The results indicated that Ef fect during the YOG was
the lowest during the three periods due to strong and effective
emission reduction measures (before the YOG: p = 0.000; after
the YOG: p = 0.001), while residents perceived the highest
level of air pollution effects after the YOG. Meanwhile, Trust
before and after the YOG were both significantly lower than
Trust during the YOG (before the YOG: p = 0.000; after the
YOG: p = 0.000). Trust after the YOG was higher than before
the YOG, but there is no significant difference between them
(p = 0.120). However, Familiarity during and after the YOG
were both significantly higher than before the YOG (during
13828
YOG: p = 0.003; after YOG: p = 0.000). Residents became
more and more familiar with haze pollution over the course of
the study. These results reflect the public’s heightened risk
perception levels to PM2.5 after the government implemented
intervention and emission reduction measures during the
YOG. On the basis of these heightened levels, it becomes clear
that the short-term emission reductions do call the attention of
the public, which may increase support for long-term action.
Regression models were employed in our study to analyze
the determining factors for health risk perception of air
pollution, as shown in Figure 3a−c. The variable definition of
the regression analysis can be seen in S-II.
First, before the YOG, the selected variables had no
significant effect on perceived Ef fect. Ef fect during the YOG
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology Article

Table 2. Spearman Correlation Analysis among Public Risk Perception Factors, PM2.5 Concentrations, and ADDa
before YOG 1 2 3 4 5
1. Effect 1.000 0.150* 0.082 −0.015 −0.037
2. Familiarity 1.000 −0.375** −0.116 −0.144*
3. Trust 1.000 0.119 0.097
4. PM2.5 concentration 1.000 0.747**
5. ADD 1.000
during YOG
1. Effect 1.000 0.079 0.056 0.012 0.016
2. Familiarity 1.000 −0.153* −0.081 −0.035
3. Trust 1.000 −0.079 −0.085
4. PM2.5 concentration 1.000 0.661**
5. ADD 1.000
after YOG
1. Effect 1.000 0.369** 0.024 −0.056 −0.137*
2. Familiarity 1.000 −0.243** 0.101 0.083
3. Trust 1.000 0.004 −0.042
4. PM2.5 concentration 1.000 0.466**
5. ADD 1.000
a
(**. p < 0.01 *. p < 0.05).

related to income (β = 0.177, p = 0.001), experience with

Figure 5. Public risk acceptable levels of air pollution reduction
Figure 6. Public willingness to accept compensation (WTA) for air
pollution related deaths.
was significantly affected only by gender (β = −0.433, p =
0.000), which indicates that men may be more sensitive to the
effects of haze pollution. Ef fect after the YOG was significantly
13829
severe haze (β = −0.530, p = 0001), and chronic diseases (β =
−0.242, p = 0.033), suggesting that residents who have higher
income levels, have experienced severe haze pollution, and/or
have suffered from chronic diseases may be more sensitive to
Ef fect.
Second, Familiarity before the YOG was significantly
influenced by gender (β = 0.201, p = 0.010) and experience
with severe haze pollution (β = −0.247, p = 0.004), indicating
that women and residents who had experienced severe haze
pollution were more familiar with air pollution in general.
Familiarity during the YOG was significantly influenced only
by education (β = 0.122, p = 0.023), which indicates that the
more educated one is, the more knowledge of haze pollution
one tends to hold. Familiarity after the YOG was significantly
influenced only by chronic diseases (β = −0.773, p = 0.000),
suggesting that people who had suffered from chronic diseases
were more familiar with air pollution.
Third, the selected variables had no significant effect on
perceived Trust before and during the YOG. Trust after the
YOG was significantly influenced by smoking (β = 0.308, p =
0.020), education (β = −0.169, p = 0.015), and income levels
(β = 0.098, p = 0.040), which indicates that smokers hold less
trust in government. Moreover, the lower-educated and higher-
income respondents displayed greater trust in the government.
Previous researchers have discovered that the perception of
health risks from atmospheric pollution could be influenced by
many factors, such as gender, age, education level, income
level, and environmental behavior.38−40 Our study also
confirmed that differences in socio-demographic status may
cause individuals to hold different perceptions of and responses
to air pollution before, during, and after the YOG. Although
the influencing variables are not consistent across time periods,
the changes of these variables still emerged some regularity:
Before and during the YOG, the significant influencing factors
were less than those after the YOG, which indicated that the
personal difference of the risk perception factors was not
obvious during these two periods. While after the YOG, the
crowd disparity was more obvious since there were more
significant factors of risk perception of air pollution emerging
under the strong policy interventions, which implies that
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology
Figure 7. Comparison of risk perception, ADD, risk acceptance levels, and willingness to pay (WTP) and willingness to accept compensation
(WTA) during the three periods.
sensitive groups appeared in increasing numbers after the
intervention policies. Generally speaking, in addition to
traditionally sensitive groups such as residents with chronic
diseases,41 (1) men, (2) residents with higher income levels,
and (3) residents who experienced a severe haze pollution
perceived a greater risk of haze pollution in our study. People
(1) with different income, (2) education levels, as well as (3)
smoking status have significant changes of trust after YOG. In
addition, the government should prioritize its communications
and messaging regarding air pollution mitigation toward (1)
men, and (2) residents with lower education.
3.3. Comparative Analysis and One-Way ANOVA of
ADD. The mean values of the individual daily PM2.5 exposures
before, during, and after the YOG were 36.46 μg/kg·d, 4.95
μg/kg·d, and 17.92 μg/kg·d, respectively. As seen in Figure
4a−c, before the YOG, the public had the highest levels of
PM2.5 exposure, while they had the lowest levels during the
YOG. After the YOG, residents’ ADD levels increased but
remained lower than those before the YOG.
As shown in Table 1, according to our calculations and one-
way ANOVA of ADD during the three periods, there were
significant disparities among certain demographic character-
istics, indicating that these characteristics may influence the
respondents’ actual ADD levels. No significant association was
discovered between the public’s ADD and demographic
characteristics before the YOG in this study. The public’s
ADD during the YOG was significantly associated with gender
(F = 18.258, p = 0.000) and age (F = 18.971, p = 0.000),
demonstrating that women had higher levels of PM 2.5
exposure. In addition, people at the 20−29 age group had
the highest levels of PM2.5 exposure, while those older than 60
were exposed to the lowest levels of PM2.5. The public’s ADD
after the YOG was significantly associated with gender, age,
education level, and income level (F = 56.171, p = 0.000; F =
11.362, p = 0.000; F = 4.338, p = 0.005; F = 11.127, p =
0.000). These phenomena reveal that after the YOG, women
and residents with lower levels of education and income had
higher levels of PM2.5 exposure. Moreover, the results also
indicate that the residents aged 30−39 had the highest levels of
PM2.5 exposure, while those over 60 were exposed to the
13830
Article
lowest levels of PM2.5. These results are consistent with several
previous studies, which confirmed that age, occupation,
income level, and education level are associated with levels
of exposure to toxic air pollution.42,43
Although residents over 60 are especially vulnerable to the
adverse effects of haze pollution,44 they are exposed to the
lowest ambient concentrations of PM2.5 because of reduced
outdoor activities. The government’s attention is urgently
needed to take active measures to help those aged 20−39, who
face the highest level of PM2.5 exposure due to extended
periods of outdoor activity. For example, notifications and
warnings concerning haze days must be provided and the
target population ought to be informed to reduce outgoing
activities.
3.4. Correlation Analysis between Health Risk
Perception Factors and ADD. Spearman correlation analysis
was conducted to examine the relationship between public
health perception and ADD among the three periods. As
shown in Table 2, Familiarity before the YOG was significantly
and negatively correlated with ADD, while Ef fect after the
YOG was significantly and negatively correlated with ADD.
The health perception of air pollution during the YOG had no
significant correlation with ADD and PM2.5 concentrations.
As shown in Table 2, respondents who perceived the higher
Ef fect had the lower PM2.5 exposure after the YOG, which may
indicated that respondents tended to be more sensitive and
they were more likely to take protective measures to reduce the
personal exposure. Cole-Hunter et al. have discovered that the
public health risk perception of air pollution was positively
associated with the estimated level.45 However, Dorizas et al.
concluded that there was no significant correlation between
perception of air pollution and measured pollutant concen-
trations.46 According to our study, respondents who were more
familiar with and more sensitive to air pollution were exposed
to less air pollution before and after the YOG. We can also see
from Table 2 that Familiarity negatively correlates with Trust
during the three periods, because residents usually learn about
air pollution from all kinds of social medias, which has been
demonstrated a negative influence on political trust if they
know more about air pollution in previous studies.47,48 It
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology
indicated that this intervention policy did not effect the
relationship between Familiarity and Trust. Therefore, the
government should actively communicate with the public
through social media and let them know more about emission
reduction measures implemented to improve public trust
gradually. What’s more, measures that promote risk
communication and awareness can in turn reduce people’s
exposure to air pollution, as knowledge of its effects can induce
people to take active steps to protect themselves.
3.5. Analysis of Public Risk Acceptable Levels (PRAL)
of Air Pollution. The specific proportion of public attitudes
toward different air pollution control policies before, during,
and after the YOG is exhibited in Tables S4−S6. The average
PM2.5 levels during these three periods are shown in Table S7.
The scatter diagram with function fitting in Figure 5 reveals the
relationship between the reduction rate of PM2.5 concen-
trations and public acceptance levels for the two scenarios
before and during the YOG. As expected, risk acceptance
increased with rising reductions in PM2.5 concentrations in
each scenario. Before the YOG, 50% of the investigated
respondents considered it acceptable to have a PM2.5
concentration reduction to 57.58 μg/m3 by 2017 (a 9.22 μg/
m3 reduction since 2014). Furthermore, this concentration
must decrease to 52.08 μg/m3 by 2022 for 50% of the
respondents (a 14.92 μg/m3 reduction since 2014), which
requires the PM2.5 concentration reductions to follow a
continuous trend and for more mitigation measures regarding
air pollution to be developed.
The average PM2.5 level in Nanjing in 2017 was about 67
μg/m3. During the YOG, 68.54% of the respondents
considered that level to be acceptable even if the PM2.5
concentrations did not decline. However, according to Figure
5, the acceptable PM2.5 concentration should decrease to 59.2
μg/m3 by 2022 (a 7.80 μg/m3 reduction since 2014) in order
to satisfy 50% of the respondents.
After the YOG, 50% of the investigated respondents
considered it acceptable to have a PM2.5 concentration
reduction to 47.69 μg/m3 by 2017(a 19.31 μg/m3 reduction
since 2014). The acceptable PM2.5 concentration should
decrease to 37.50 μg/m3 (a 29.50 μg/m3 reduction since
2014) for 50% of the respondents by 2022.
During the YOG, residents’ risk acceptance level of PM2.5
was at its highest. The PRAL of air pollution after the YOG
required a higher PM2.5 concentration reduction rate than the
reduction rate required before the YOG by 2017. This same
tendency was found according to the second landscape shown
to the study respondents: a higher PM2.5 concentration
reduction rate needs to be attained after the YOG than before
the YOG by 2022. The results indicate that residents usually
hope to maintain the current good air quality status and will
even raise their standards for stricter environmental controls
after experiencing better air quality.
3.6. Estimates of Willingness to Pay. The aggregate
statistics of the willingness to pay distribution (see Table S8)
reveal the following insights: during the YOG, the proportion
of people willing to pay for early warning systems and
protective measures decreased, but the per capita payment
increased compared to that before the YOG, while the
situation for measure 3 showed different results. After the
YOG, the proportion of residents willing to pay for the three
measures all increased, as did the per capita payment.
Moreover, the results showed that the proportion of
respondents willing to pay for measure 2: giving out masks
13831
Article
and providing an air purifier were the highest during the
periods, and the proportion of respondents willing to pay for
measure 1: setting up an early warning system for extreme
weather took the second place. Therefore, a reminder of severe
haze weather and health protection measures should be
provided especially for the residents aged 20−39, who were
exposed to the highest level of PM2.5 concentrations. Besides,
the information concerning health risks from haze pollution
must be provided in a timely manner through education and
risk communication.
We found that the proportion of respondents willing to pay
for these three protective methods did not exceed 50% during
the YOG, and more than one-half of the respondents refused
to pay at all. Of these unwilling respondents, the majority
thought it was the government’s responsibility to pay for
protective measures in response to haze pollution.
The values for the proportion of respondents willing to pay,
the per capita payment, and the total willingness to pay for the
three measures all increased after the YOG compared to those
during the YOG. We concluded that after the government
implemented effective emission reduction measures, residents
became more willing to take measures to protect themselves
from the health risks of haze pollution.
3.7. Estimates of Willingness to Accept Compensa-
tion. Table S9 contains a statistical summary of the
compensation that people are willing to accept for each
scenario of risk damage. The cost of air pollution-induced
health-related losses could be described as a quadratic equation
of the PM2.5 concentrations (see Figure 6). We calculated the
total compensation that Nanjing residents were willing to
accept in accordance with a PM2.5 concentration of 74 μg/m3.
The value of health costs for the total population in Nanjing
was 123.62, 80.71, and 138.97 billion CNY before, during, and
after the YOG, respectively.
In summary, before the YOG, residents’ willingness to
accept compensation was higher than that during the YOG.
However, after the YOG, residents’ willingness to accept
compensation was the highest of the three periods. After the
YOG, residents required higher compensation for health-
related losses than before and during the YOG. We concluded
that after the YOG, residents realized that the government had
the capability as well as responsibility to take effective
measures to reduce PM2.5.
According to our study, the mean values for willingness to
accept compensation before, during, and after the YOG
obtained in this study were 200 times greater than the mean
values for willingness to pay (360.18 million, 282.53 million,
and 542.82 million CNY, respectively). Willingness to pay and
willingness to accept can be explained as the income
expenditure in support of environmental improvement and
the willingness to accept compensation after the occurrence of
environmental damage.49 A number of studies have found that
the value of willingness to accept compensation is far greater
than that of willingness to pay, which is consistent with our
results.50−52 The disparity between willingness to pay and
willingness to accept compensation has great importance for
environment management; the welfare loss caused by destroy-
ing or polluting the environment is much greater than the
welfare income generated by protecting and improving the
environment. Therefore, the government should take proactive
measures to prevent or mitigate environmental and health
losses. For example, the government can established an early
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology
warning system for extreme weather, distribute masks during
haze-polluted days, etc.
3.8. Implications. As seen in Figure 7, great differences in
health perception levels, ADD, public acceptance levels, and
willingness to pay and willingness to accept compensation for
air pollution before, during, and after the YOG were revealed.
During the YOG, the public was exposed to the lowest
concentrations of PM2.5 and was most accepting of the risks
posed by haze. After the YOG, however, residents became
more sensitive to haze pollution and demanded higher air
quality as the government implemented strong policy
interventions to cut emissions. The resulting change was
characterized by a greater willingness to pay for risk reduction
measures and a greater willingness to accept compensation for
health-related losses.
In order to meet the stronger demand for clean air, long-
term and effective measures should be taken, as opposed to
short-term interventions such as those taken during the YOG.
In addition to traditionally sensitive groups such as residents
with chronic diseases, (1) women, (2) the middle-aged, (3)
people with lower levels of education, and (4) lower-income
individuals also require special protections. Furthermore, there
was a significant negative correlation between ADD and Trust
before the YOG. To remedy the situation, targeted
communications and messaging aimed at these sensitive
groups should be employed to validate their perception of
air pollution risks and encourage them to proactively protect
themselves.
Although our results are slightly more applicable to a
population with higher overall education levels, they can help
local governments gain a better understanding of risk
management boundaries and develop tailored health risk
management strategies suitable for people of varying socio-
economic characteristics. This study would help local govern-
ments identify the most at-risk populations, improve awareness
of health risks, and influence citizens to take initiative and
protective measures to mitigate the effects of air pollution.
■
ASSOCIATED CONTENT
*
S Supporting Information
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.est.8b01634.
Questionnaires and the variable definition of the
regression analysis; tables showing demographic charac-
teristics of respondents, goodness-of-fit statistics, public
risk perception of PM2.5 concentrations, the average
daily PM2.5 concentrations, public risk acceptance of
different PM2.5 concentrations, and public willingness to
pay/accept during the three periods; and figures
showing path diagram of the revised final confirmatory
factor analysis model (PDF)
■
AUTHOR INFORMATION
Corresponding Authors
*Tel: 025-89680535; e-mail: huanglei@nju.edu.cn (L.H.).
*Tel: 025-89680535; e-mail: jbi@nju.edu.cn (J.B.) .
ORCID
Jie Li: 0000-0002-2739-9280
Notes
The authors declare no competing financial interest.
13832
Article
■ ACKNOWLEDGMENTS
We thank Wenjie Kuang, Can Zhang, Linli Liu, Qianqi Yang,
and Penghui Liu for questionnaire distribution assistance and
advisory services. This work was supported by the Chinese
Natural Science Foundation (41822709, 41571475, and
21637002) and the Fundamental Research Funds for the
Central Universities.
■
REFERENCES
(1) Cheng, Z.; Wang, S.; Jiang, J.; Fu, Q.; Chen, C.; Xu, B.; Yu, J.;
Fu, X.; Hao, J. Long-term trend of haze pollution and impact of
particulate matter in the Yangtze River Delta, China. Environ. Pollut.
2013, 182 (6), 101.
(2) Han, L.; Zhou, W.; Li, W.; Li, L. Impact of urbanization level on
urban air quality: A case of fine particles (PM 2.5) in Chinese cities.
Environ. Pollut. 2014, 194 (1), 163.
(3) Pui, D. Y. H.; Chen, S. C.; Zuo, Z. PM_(2.5) in
China:Measurements,sources,visibility and health effects,and mitiga-
tion. Particuology 2014, 13 (2), 1−26.
(4) Tai, A. P. K.; Mickley, L. J.; Jacob, D. J. Correlations between
fine particulate matter (PM2.5) and meteorological variables in the
United States: implications for the sensitivity of PM2.5 to climate
change. Atmos. Environ. 2010, 44 (32), 3976−3984.
(5) Hedley, A. J.; Wong, C. M.; Thach, T. Q.; Ma, S.; Lam, T. H.;
Anderson, H. R. Cardiorespiratory and all-cause mortality after
restrictions on sulphur content of fuel in Hong Kong: an intervention
study. Lancet 2002, 360 (9346), 1646−52.
(6) Clancy, L.; Goodman, P.; Sinclair, H.; Dockery, D. W. Effect of
air-pollution control on death rates in Dublin, Ireland: an intervention
study. Lancet 2002, 360 (9341), 1210−4.
(7) Lee, J.; Son, J.; Cho, Y. Benefits of Mitigated Ambient Air
Quality Due to Transportation Control on Childhood Asthma
Hospitalization during the 2002 Summer Asian Games in Busan,
Korea. J. Air Waste Manage. Assoc. 2007, 57 (8), 968−73.
(8) Lin, Y.; Huang, K.; Zhuang, G.; Fu, J. S.; Xu, C.; Shen, J.; Chen,
S. Air Quality over the Yangtze River Delta during the 2010 Shanghai
Expo. Aerosol Air Qual. Res. 2013, 13 (6), 1655−1666.
(9) Streets, D. G.; Fu, J. S.; Jang, C. J.; Hao, J. M.; He, K. B.; Tang,
X. Y.; Zhang, Y. H.; Wang, Z.; Li, Z. P.; Zhang, Q. A.; et al. Air quality
during the 2008 Beijing Olympic Games. Atmos. Environ. 2007, 41
(3), 480−492.
(10) Xu, H. M.; Tao, J.; Ho, S. S. H.; Ho, K. F.; Cao, J. J.; Li, N.;
Chow, J. C.; Wang, G. H.; Han, Y. M.; Zhang, R. J.; et al.
Characteristics of fine particulate non-polar organic compounds in
Guangzhou during the 16th Asian Games: Effectiveness of air
pollution controls. Atmos. Environ. 2013, 76 (5), 94−101.
(11) Zhang, X. Y.; Wang, Y. Q.; Lin, W. L.; Zhang, Y. M.; Zhang, X.
C.; Gong, S.; Zhao, P.; Yang, Y. Q.; Wang, J. Z.; Hou, Q.; et al.
Changes of Atmospheric Composition and Optical Properties Over
BEIJING2008 Olympic Monitoring Campaign. Bull. Am. Meteorol.
Soc. 2009, 90 (11), 1633−1652.
(12) Hao, N.; Valks, P.; Loyola, D.; Cheng, Y. F.; Zimmer, W.
Space-based measurements of air quality during the World Expo 2010
in Shanghai. Environ. Res. Lett. 2011, 6 (4), 044004.
(13) Huang, K.; Zhuang, G.; Lin, Y.; Wang, Q.; Fu, J. S.; Fu, Q.; Liu,
T.; Deng, C. How to improve the air quality over mega-cities in
China? − Pollution characterization and source analysis in Shanghai
before, during, and after the 2010 World Expo. Atmos. Chem. Phys.
2013, 13 (12), 5927−5942.
(14) Liu, H.; Wang, X.; Zhang, J.; He, K.; Wu, Y.; Xu, J. Emission
controls and changes in air quality in Guangzhou during the Asian
Games. Atmos. Environ. 2013, 76 (5), 81−93.
(15) Mijling, B.; van der A, J.; Boersma, K. F.; Van Roozendael, M.
V.; De Smedt, I. D.; Kelder, H. M. Reductions of NO2 detected from
space during the 2008 Beijing Olympic Games. Geophys. Res. Lett.
2009, 36 (13), 1−6.
(16) Wang, T.; Nie, W.; Gao, J.; Xue, L. K.; Gao, X. M.; Wang, X. F.;
Qiu, J.; Poon, C. N.; Meinardi, S.; Blake, D.; et al. Air quality during
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833
Environmental Science & Technology
the 2008 Beijing Olympics: Secondary pollutants and regional impact.
Atmos. Chem. Phys. 2010, 10 (16), 7603−7615.
(17) Wang, M.-W.; Chen, J.; Cai, R. Air quality and acute myocardial
infarction in adults during the 2016 Hangzhou G20 summit. Environ.
Sci. Pollut. Res. 2018, 25 (10), 9949−9956.
(18) Gao, J.; Xu, X.; Ying, Z.; Jiang, L.; Zhong, M.; Wang, A.; Chen,
L. C.; Lu, B.; Sun, Q. Post-Effect of Air Quality Improvement on
Biomarkers for Systemic Inflammation and Microparticles in Asthma
Patients After the 2008 Beijing Olympic Games: a Pilot Study.
Inflammation 2017, 40 (4), 1214−1224.
(19) Li, H.; Zhou, L.; Wang, C.; Chen, R.; Ma, X.; Xu, B.; Xiong, L.;
Ding, Z.; Chen, X.; Zhou, Y.; Xu, Y.; Kan, H. Associations Between
Air Quality Changes and Biomarkers of Systemic Inflammation
During the 2014 Nanjing Youth Olympics: A Quasi-Experimental
Study. Am. J. Epidemiol. 2017, 185 (12), 1290−1296.
(20) Altemose, B.; Robson, M. G.; Kipen, H. M.; Ohman Strickland,
P.; Meng, Q.; Gong, J.; Huang, W.; Wang, G.; Rich, D. Q.; Zhu, T.;
Zhang, J. Association of air pollution sources and aldehydes with
biomarkers of blood coagulation, pulmonary inflammation, and
systemic oxidative stress. J. Exposure Sci. Environ. Epidemiol. 2017,
27, 244.
(21) Batterman, S.; Godwin, C.; Jia, C. Long duration tests of room
air filters in cigarette smokers’ homes. Environ. Sci. Technol. 2005, 39
(18), 7260−8.
(22) Elias, W.; Shiftan, Y. The influence of individual’s risk
perception and attitudes on travel behavior. Transport. Res. A-Polym.
2012, 46 (8), 1241−1251.
(23) Ban, J.; Zhou, L.; Zhang, Y.; Brooke Anderson, A. G.; Li, T.
The health policy implications of individual adaptive behavior
responses to smog pollution in urban China. Environ. Int. 2017,
106, 144−152.
(24) Semenza, J. C.; Wilson, D. J.; Parra, J.; Bontempo, B. D.; Hart,
M.; Sailor, D. J.; George, L. A. Public perception and behavior change
in relationship to hot weather and air pollution. Environ. Res. 2008,
107 (3), 401.
(25) Daramola, O. P. Determinants of students perceived manmade
environmental hazards and risks in tertiary educational institutions.
Global J. Environ. Sci. Manage. 2017, 3 (1), 1−8.
(26) Alvarezpedrerol, M.; Rivas, I.; Ló p ezvicente, M.;
Suadesgonzález, E.; Donairegonzalez, D.; Cirach, M.; de Castro, M.;
Esnaola, M.; Basagaña, X.; Dadvand, P.; et al. Impact of commuting
exposure to traffic-related air pollution on cognitive development in
children walking to school. Environ. Pollut. 2017, 231 (1), 837−844.
(27) Huang, L.; Zhou, Y.; Han, Y.; Hammitt, J. K.; Bi, J.; Liu, Y.
Effect of the Fukushima nuclear accident on the risk perception of
residents near a nuclear power plant in China. Proc. Natl. Acad. Sci. U.
S. A. 2013, 110 (49), 19742−7.
(28) Sjoberg, L. Factors in risk perception. Risk. Anal. 2000, 20 (1),
1−12.
(29) Slovic, P. Perception of Risk. Science. Science 1987, 236 (4799),
280−285.
(30) Cullen, L.; Madison, R.; Means, B.; Norton, S.; Valaoras, G.;
Zamuda, C.; Zaragoza, L. Risk Assessment Guidance for Superfund Vol. I
Human Health Evaluation Manual (Part A) Interim Final; 1989.
(31) Duan, X. L. Research Methods of Exposure Factors and Its
Application in Environmental Health Risk Assessment; Science Press:
Beijing, China, 2012.
(32) Morawska, L.; He, C.; Hitchins, J.; Gilbert, D.; Parappukkaran,
S. The relationship between indoor and outdoor airborne particles in
the residential environment. Atmos. Environ. 2001, 35 (20), 3463−
3473.
(33) Hammitt, J. K.; Zhou, Y. The economic value of air-pollution-
related health risks in China: a contingent valuation study. Environ.
Resour Econ 2006, 33 (3), 399−423.
(34) Katz, K.; Sterner, T. The Value of Clean Air: Consumers’
Willingness to Pay for a Reduction in Gasoline Vapours at Filling
Stations. Energy Stud. Rev. 2014, 2, (1). DOI: 10.15173/esr.v2i1.204
13833
Article
(35) Wang, H.; Mullahy, J. Willingness to pay for reducing fatal risk
by improving air quality: a contingent valuation study in Chongqing,
China. Sci. Total Environ. 2006, 367 (1), 50−7.
(36) Brouwer, R.; Brander, L.; Van Beukering, P. A convenient
truth”: air travel passengers’ willingness to pay to offset their CO2
emissions. Clim. Change 2008, 90 (3), 299−313.
(37) Xiaochuan, P.; Guoxing, L.; Ting, G. Dangerous Breathing:
Health Hazard and Economic Loss Assessment of PM2.5; China
Environmental Science Press: 2012.
(38) Tvinnereim, E.; Liu, X.; Jamelske, E. M. Public perceptions of
air pollution and climate change: different manifestations, similar
causes, and concerns. Clim. Change 2017, 140 (3−4), 399−412.
(39) Daramola, O.; Odunsi, O. Determinants of students perceived
manmade environmental hazards and risks in tertiary educational
institutions. Global J. Environ. Sci. Manage. 2017, 3 (1), 43−50.
(40) Lan, G.-l.; Yuan, Z.-k.; Maddock, J. E.; Cook, A.; Chu, Y.-y.;
Pan, B.-b.; Tu, H.; Fan, S.; Liao, X.; Lu, Y. Public perception of air
pollution and health effects in Nanchang, China. Air Qual., Atmos.
Health 2016, 9 (8), 951−959.
(41) Su, C.; Hampel, R.; Franck, U.; Wiedensohler, A.; Cyrys, J.;
Pan, X.; Wichmann, H. E.; Peters, A.; Schneider, A.; Breitner, S.
Assessing responses of cardiovascular mortality to particulate matter
air pollution for pre-, during- and post-2008 Olympics periods.
Environ. Res. 2015, 142, 112−122.
(42) Yu, K.; Chen, Z.; Gao, J.; Zhang, Y.; Wang, S.; Chai, F.
Relationship between Objective and Subjective Atmospheric Visibility
and Its Influence on Willingness to Accept or Pay in China. PLoS One
2015, 10 (10), e0139495.
(43) Wenxiao, Z.; Huaizhong, M.; Hongmin, F. Analysis on the
difference of social status of air pollution exposure risk:a case study of
Liaoning Province. Environ. Pollut. Control. 2017, 39, 444−450.
(44) Mordukhovich, I.; Coull, B.; Kloog, I.; Koutrakis, P.; Vokonas,
P.; Schwartz, J. Exposure to sub-chronic and long-term particulate air
pollution and heart rate variability in an elderly cohort: the Normative
Aging Study. Environ. Health 2015, 14, 87.
(45) Cole-Hunter, T.; Morawska, L.; Solomon, C. Bicycle
Commuting and Exposure to Air Pollution: A Questionnaire-Based
Investigation of Perceptions, Symptoms, and Risk Management
Strategies. J. Phys. Act. Health 2015, 12 (4), 490−499.
(46) Dorizas, P. V.; Assimakopoulos, M. N.; Santamouris, M. A
holistic approach for the assessment of the indoor environmental
quality, student productivity, and energy consumption in primary
schools. Environ. Monit. Assess. 2015, 187 (5), 259.
(47) Ceron, A. Internet, News, and Political Trust: The Difference
Between Social Media and Online Media Outlets. J. Comput-Mediat.
Comm 2015, 20 (5), 487−503.
(48) Im, T.; Cho, W.; Porumbescu, G.; Park, J. Internet, Trust in
Government, and Citizen Compliance. J. Public Adm. Res. Theory
2014, 24 (3), 741−763.
(49) Freeman, A. M.; Herriges, J. A.; Kling, C. L. The measurement
of environmental and resource values: theory and methods. Resour.
Policy 1994, 20 (4), 281−282.
(50) Venkatachalam, L. The contingent valuation method: a review.
Environ. Impact Assessment Rev. 2004, 24 (1), 89−124.
(51) Huang, L.; Ban, J.; Duan, B.; Bi, J.; Yuan, Z. Public demand for
remediating a local ecosystem: comparing WTP and WTA at Hongze
Lake, China. Lake Reservoir Manage. 2013, 29 (1), 23−32.
(52) Horowitz, J. K.; McConnell, K. E. A Review of WTA/WTP
Studies. J. Environ. Econ. Manage. 2002, 44 (3), 426−447.
DOI: 10.1021/acs.est.8b01634
Environ. Sci. Technol. 2018, 52, 13824−13833