The Impact of Chinese Public Environmental Awareness on Environmental Behavior: An Analysis Based on China National Surveys in 2003, 2010 and 2021

Abstract

Environmental awareness and behavior play a crucial role in the improvement of the environment. Five dimensions of environmental awareness are considered here including environmental values, environmental knowledge, environmental responsibility, environmental ability and environmental concern while we also distinguish between private and public environmental behaviors. Based on data from the China General Social Survey (CGSS) in 2003, 2010 and 2021, this research investigated changes in Chinese residents’ environmental awareness and environmental behavior and the relationship between them. The relevant empirical methods were the MV-Probit model, the Poisson model and a mediating effect model. The results are as follows: (1) From 2003 to 2021, environmental awareness of the public has improved significantly. Among them, environmental values, environmental knowledge and environmental concern have increased by 24.5%, 12.6% and 12.77%, respectively, while environmental capacity has decreased by 7.27% from 2010 to 2021. The eastern region of China has the highest score of environmental awareness, while the western region has the lowest one. (2) From 2003 to 2021, the public’s green consumption behavior changed from more than half of the public never consuming green products in 2003 to nearly 75% always or often consuming green products in 2021. Increases in recycling were also found over time. However, the participation rate in public environmental behavior declined. In 2021, environmental behaviors were lowest in the eastern region and highest in the western region. (3) Environmental responsibility, environmental concern and environmental capacity direct and indirectly affect private and public environmental behaviors and the number of environmental behaviors through the willingness to protect the environment, which plays a semi-mediating role. Environmental knowledge only has a direct impact on private green consumption behavior. (4) There is a correlation between public and private environmental behaviors. Given the results of this study, a series of policy implications and recommendations are provided.

1. Introduction

The ecological crisis is becoming increasingly serious due to the rapid development of the global economy. In the 1970s, the United Nations issued the “Declaration on the Human Environment” to tackle significant environmental problems. This marks a global consensus toward sustainable development and achieving harmonious development between humans and nature. The Global Environment Outlook released by the Fourth United Nations Environment Assembly highlighted that anthropogenic pollution and environmental damage are responsible for one-quarter of premature deaths and diseases worldwide [1]. How to manage environmental issues and achieve sustainable development has become a major issue that countries worldwide are facing. In the past two decades, China has promoted economic growth at the expense of the environment [2]. The Chinese government must actively seek countermeasures to deal with environmental problems. In 2012, the report of the 18th CPC National Congress incorporated the concept of ‘ecological civilization construction’ into the “five-in-one” general layout of socialism with Chinese characteristics. The essence of ecological civilization construction is the cultivation of public environmental awareness and the conscious implementation of ecological behavior [3].

The environmental awareness and behaviors of individuals play an important role in achieving sustainable development [3]. However, relevant studies have indicated that the environmental awareness of Chinese citizens has hampered the promotion of ecological civilization construction and become the biggest obstacle to improving environmental behavior [4]. In addition, rather than being static, the environmental awareness and behaviors of individuals are dynamic [5]. Therefore, it is of practical significance to understand their changes over time, as well as the relationship between them for building a sound environmental governance system.

There are two views of the definition of environmental awareness in the existing literature. One view is that environmental awareness is the unity of “knowing” and “doing” [6,7] but that the environmental awareness level ultimately depends on the “doing” [8]. For example, Hong [9] put forward that environmental awareness includes ecological knowledge, basic environmental values, attitudes to participate in environmental protection and environmental protection behavior [10]. The first three belong to knowing, and the latter one belongs to doing. The other view is that environmental awareness includes only the knowing part, that is, environmental knowledge, ecological awareness, environmental responsibility and value awareness, and the doing part is not included. Environmental awareness and environmental behavior are two independent variables [11,12]. The second view has gradually forged a consensus in academia. The reason is that the purpose of studying environmental awareness is to explore the extent to which it can be transformed into environmental behavior. The first view avoids this problem [13]. Moreover, the “knowing” and “doing” of environmental awareness in China are not always consistent [14], which is manifested as “knowing is strong and doing is weak” [15]. From the perspective of “cognition determines behavior”, some researchers [16] have analyzed the extent to which environmental awareness could affect individual environmental behavior.

The existing literature has examined the environmental awareness and behavior of the Chinese public at a single point in time, but changes in environmental awareness and behavior over time still remain unknown [17,18,19,20]. Therefore, understanding increases and reductions in public environmental awareness and behavior plays a significant role in the improvement of environmental quality in China. “Awareness determines behavior” [14,15], but environmental awareness is a multidimensional variable and the effects of different dimensions of environmental awareness on public environmental behavior may vary. Most of the current literature treats environmental awareness as a single variable [21,22,23] or has adopted multi-dimensional variables to characterize environmental awareness [24,25] to analyze the extent of its influence on the public’s environmental behavior. Clarifying the influence of various dimensions of environmental awareness on the Chinese public’s environmental behavior will identify key areas for enhancing their environmental practices. Furthermore, environmental behavior includes both public and private domains, which can often be interchanged [20]. However, the existing literature primarily focuses on either public or private environmental behaviors [26,27]. Investigating the relationship between the two can provide valuable insights for improving public environmental behavior.

This study aims to address the identified knowledge gap through three specific objectives: (1) Using data from the CGSS surveys in 2003, 2010 and 2021, we analyzed the changes in public environmental awareness and behavior in China over the past 20 years, and we identified their spatial differences. (2) Environmental awareness is divided into five dimensions: environmental values, environmental knowledge, environmental responsibility, environmental concern and environmental competence, and the influences of different dimensions of environmental awareness on environmental behavior were analyzed. (3) The correlation between public and private environmental behaviors, and between different environmental behaviors in the private domain are explored. Theoretically, this study can extend the literature on environmental behavior by determining the influence of environmental awareness sub-dimensions on public environmental behavior, and the correlation between different environmental behaviors. In practical terms, revealing changes and spatial differences in public environmental awareness and behavior in China will not only facilitate an understanding of the effectiveness of the Chinese government’s environmental protection efforts but also provide a reference for comprehensively improving public environmental awareness and behavior in China.

In summary, the overall content of this study is organized as follows: First, based on the available data from the CGSS2003, CGSS2010 and CGSS2021, the environmental awareness and environmental behavior of the Chinese public were measured after screening the samples. Then changes in the Chinese public’s environmental awareness and environmental behaviors over the past 20 years were analyzed. Considering the spatial differences in public environmental awareness and environmental behavior in China, spatial differences were mapped. Secondly, various dimensions of environmental awareness influence environmental behavior differently. To examine the correlation between the environmental behaviors of the Chinese public, we employed the MV-Probit model for further analysis. At the same time, to more comprehensively understand the influencing factors of environmental awareness on environmental behavior, we added a mediating effect model. Finally, this study employed the Poisson model to examine how each dimension of environmental awareness influences the number of environmental behaviors undertaken by the Chinese public.

This paper is organized as follows: the second section is the theoretical analysis; the third section is the materials and methods; the fourth section is the results; the fifth section is the discussion; the sixth section is the conclusions and policy implications; and the seventh section contains the limitations and future perspectives.

2. Theoretical Analysis

Environmental awareness usually refers to people’s views, cognition, attention to environmental issues and the resulting sense of responsibility and environmental literacy [28,29]. It includes five dimensions: environmental values, environmental knowledge, environmental responsibility, environmental concern and environmental capacity [30,31]. Environmental values reflect people’s rational thinking about the environment. Individuals who hold the value of harmony between humans and nature are becoming more conscious of the significance of environment protection and are more likely to have more environmentally friendly behaviors [32]. The implementation of more environmentally friendly behavior requires environmental knowledge and skills [33,34,35]. Individuals with greater environmental knowledge and skills can implement more environmentally friendly behaviors better. Environmental capability refers to the ability of actors to transform their own knowledge and skills into related behaviors, which is the determinant of behavior implementation. In general, residents’ strong environmental capability can provide methodological guidance for the implementation of environmental behavior [36,37]. Environmental behavior is external and altruistic, and altruism is an expression of individual responsibility. Individuals with high environmental responsibility are more likely to implement environmental behavior [38,39,40]. The principle of implementing environmental behavior involves paying attention to and recognizing environmental problems [41]. The higher the level of environmental concern, the easier it is for individuals to respond positively to environmental problems and take corresponding actions [42]. In summary, this paper proposes the following research hypothesis:

H1 :

Environmental awareness (environmental values, environmental knowledge, environmental responsibility, environmental concern, environmental capacity) has a positive impact on Chinese citizens’ environmental behavior.

Hines et al. [33] introduced the model of responsible environmental behavior, which shows that the dimensions of environmental awareness, including environmental values, environmental knowledge, environmental responsibility, environmental concern and environmental capacity, all impact environmental behavior via behavioral intention. The theory of planned behavior suggests that behavioral willingness is the most direct antecedent variable of behavior and other subjective psychological factors indirectly affect behavior through behavioral willingness [43], In other words, an individual’s environmental awareness through environmental protection willingness, in turn, influences the occurrence of an individual’s environmental behavior. When members of the public who value harmonious coexistence between humans and nature become concerned about environmental issues, a strong sense of responsibility, extensive environmental knowledge and high environmental competence will spur them to develop the will to engage in more environmentally friendly behaviors as will individuals with stronger environmental intentions [44]. To summarize, this paper proposes the following research hypothesis:

H2 :

Environmental willingness plays a mediating role between environmental awareness and environmental behavior.

Environmental behavior includes two dimensions: private environmental behavior and public environmental behavior [33]. Private environmental behavior refers to an individual’s daily life practices, whereas public environmental behavior pertains to actions that benefit the community. The values–beliefs–norms theory suggests that different environmental behaviors are collectively shaped by factors such as values, beliefs and sense of responsibility. Under the same factors, there exists a certain correlation between different environmental behaviors [45]. Private environmental behavior tends to be more likely to be adopted by individuals because of self-interest. Individuals will acquire more environmental knowledge and environmental protection skills in the process of implementing private environmental behavior and they will have a deeper understanding and experience of respecting the law of environmental development. Consequently, individuals who engage in private environmental behavior may be more inclined to undertake public environmental behavior, driven by a robust sense of environmental responsibility. Due to the stronger non-economic and altruistic nature of public environmental behavior, the individual who is willing to implement public environmental behavior is more conscious of environmental protection. Individuals who choose to implement public environmental behavior will also be more likely to implement private environmental behavior [20]. In summary, this paper proposes the following research hypothesis:

H3 :

There is a correlation between different public environmental behaviors.

In the summary, the influence of public environmental awareness on environmental behavior and the mediating role of environmental willingness in China is shown in Figure 1.

3. Materials and Methods

3.1. Research Area and Data Sources

The subject of environmental awareness is the individual, so nationwide survey data on environmental awareness and behavior are used here. The data in this paper are from China General Social Survey (CGSS). The CGSS is implemented by the China Survey and Data Center of Renmin University of China. There have been thirteen CGSS surveys to date. The CGSS questionnaire is divided into a core module, additional modules and theme modules. Surveys on individual environmental awareness and behavior are part of the theme modules. There are five more comprehensive surveys involving public environmental awareness and behavior but the issues involved in each survey are not the same. Data from the 2021, CGSS is the latest national survey data available for the study of public environmental awareness, and 2003 and 2010 are roughly consistent with the survey indicators from 2021. Therefore, this paper has used the CGSS2003, 2010 and 2021 data to analyze changes in public environmental awareness and behavior in China.

In accordance with international standards, the CGSS also conducted a continuous cross-sectional survey of more than 10,000 households in various provinces, municipalities and autonomous regions of mainland China in the form of household surveys. To date, the CGSS has used three different sampling schemes. These three sampling schemes all adopt multi-stage stratified PPS random sampling in principle, but they are different in their sampling frames, stratified variables and sampling stages, mainly for the purpose of effectively representing all aspects of Chinese Society. The data of 2003 used the sampling scheme of 2003–2006, which adopted stratified four-stage unequal probability sampling, and the comparison between urban samples and rural samples in the final sampling unit was 5900 and 4100, respectively. The data of 2010 and 2021 adopted the stratified three-stage probability sampling scheme of 2010. Based on the stratified layer used, the sampling units of each stage are slightly different. The subjects of the CGSS survey are the public aged over 17 years old. In 2003, the survey covered the whole country, including 22 provinces, 4 autonomous regions and 4 municipalities directly under the central government. The Tibet Autonomous Region, Hong Kong, Macao and Taiwan were excluded. The target of the survey in 2010 and 2021 was to deduct all urban and rural households in 31 provinces, autonomous regions and municipalities directly under the central government. Any samples with missing research variables of interested were eliminated for the purpose of this study to help ensure the accuracy of the research. After eliminating the samples with missing values, the number of samples that met the analysis requirements in 2003, 2019 and 2021 was 5048, 3669 and 2739, respectively.

Additionally, the data are collected through household surveys targeting college and graduate students, which bolsters the accuracy of the findings. Previous studies by Lu [46], Tang [47] and other scholars have examined environmental awareness and behavior using similar data, and their conclusions have been widely recognized. However, variations in sample characteristics across different years may introduce potential errors in the results. To address this issue, this paper incorporates a year variable in the regression analysis to control for the influence of unobservable factors, such as the survey year and the demographic characteristics of the respondents.

3.2. Variable Setting

The dependent variable in this paper is environmental behavior. According to the research of Hines et al. [33] and based on the variables in the CGSS questionnaire, environmental behavior was divided into public and private environmental behavior. Public environmental behaviors included: “In the past five years, have you signed a petition on an environmental issue; have you donated money to an environmental group; have you participated in a protest or demonstration for an environmental issue; have you joined an environmental organization?” If the respondent answered ‘Yes’ to any of the above four options, then public environmental behavior was assigned a value of 1, otherwise, it was 0. Private environmental behavior was measured by these two aspects: green consumption and recycling. The former was measured by “Do you often deliberately avoid buying certain products for the sake of environmental protection”, while the latter was measured by “Do you often deliberately separate glass, aluminum cans, plastics or newspapers for recycling?”. For the above questions, the answers “Always” and “Often” are assigned a value of 1; “Sometime” and “Never” are assigned a value of 0.

The core independent variable is environmental awareness, which is divided into five dimensions: environmental values, environmental knowledge, environmental responsibility, environmental capacity and environmental concern [30,31]. Environmental values were measured by the NEP scale. Environmental knowledge is measured through the following questions in the questionnaire: “What do you think is the harm of automobile exhaust to the environment? What is the degree of harm of industrial exhaust gas to the environment? How harmful are the pesticides and fertilizers used in agricultural production to the environment? How harmful is the air pollution caused by the pollution of rivers and lakes in China to the environment? What is the degree of harm of the global temperature rise caused by climate change to the environment? How harmful are genetically modified crops to the environment? What is the degree of harm of nuclear power plants to the environment?” The issues of environmental responsibility included: “It is difficult for me to do anything for environmental protection; unless all people do, otherwise if I protect the environment, it is meaningless; there are more important things in life than environmental protection; even if it takes a lot of money and time, I will do something good for the environment”. The issues of environmental capacity included: “I don‘t know whether my lifestyle is harmful or beneficial to the environment; how much do you know about the causes of environmental problems? How many ways do you have to solve environmental problems?” The issues of environmental concern included: “How concerned are you about environmental problems?” The options for answers to the core independent variables were a five-point Likert scale, and the scores for the different variables were obtained by the entropy method.

The mediating variable is willingness to protect the environment. The specific issues were “To what extent are you willing to pay a higher price to protect the environment? To what extent are you willing to pay higher taxes to protect the environment? To what extent are you willing to lower your living standards to protect the environment?” The answers were provided using a five-point Likert scale.

In this paper, the gender, age, political status, education level, cognition, income, class, household registration of the respondents and year variables were selected as control variables [48,49]. To ensure data integrity and availability, the year variable was designated as a dummy variable, with 0 assigned for 2010 and 1 for 2021. The definitions and assignments of each variable are presented in

Table 1. Variable description and descriptive statistics.

Variable Type	Variable	Definition and Assignment	Mean Value	Standard Deviation
Dependent variable	Green consumption	Whether to carry out green consumption? always and often = 1; sometimes and never = 0	0.321	0.467
	Recycling	Whether to carry out recycle? always and often = 1; sometimes and never = 0; No = 0	0.275	0.446
	Public environmental behavior	Whether to implement public environmental behavior? Yes = 1; No = 0	0.115	0.319
Independent variable	Environmental awareness
	Environmental value	Calculated according to the entropy method	0.623	0.213
	Environmental knowledge	Calculated according to the entropy method	0.428	0.200
	Environmental responsibility	Calculated according to the entropy method	0.434	0.179
	Environmental concern	How concerned are you about environmental problems? Completely not =1; Relatively not = 2; Uncertain = 3; Relatively = 4; Completely = 5	3.625	0.967
	Environmental capacity	Calculated according to the entropy method	0.371	0.202
Mediating variable	Willingness to protect the environment	Calculated according to the entropy method	0.474	0.235
Control variable	Gender	Female = 0; male = 1	0.467	0.498
	Age	Unit: years	49.1	16.68
	Political status	CPC members = 1, others = 0	0.128	0.335
	Education level	Uneducated = 1; junior high school and below = 2; high school and vocational school = 3; junior college = 4; bachelor degree and above = 5	2.54	1.11
	Cognition	What do you think of your ability to listen and speak Mandarin? What do you think of your ability to listen and speak English? Very poor = 1; Relatively poor = 2; General = 3; Better = 4; Very good = 5 (calculated by entropy method)	0.30	0.195
	Income	What was your personal total income last year?	9.18	4.37
	Class	Which class do you think you are in at present? low level = 1; Medium low = 2; Middle layer = 3; Middle upper layer = 4; Upper layer = 5	2.34	0.95
	Household registration	Rural = 0; city = 1	0.368	0.543

.

SPSS24.0 was utilized to assess the reliability and validity of environmental values, environmental knowledge, environmental responsibility, environmental capacity and environmental willingness. The results showed that Cronbach’s α was above 0.7 for all dimensions except for environmental responsibility, which had a slightly lower value. The KMO values of environmental values, environmental knowledge and environmental capacity were all above 0.8. The KMO value of the willingness to protect the environment was between 0.7–0.8. The value of environmental capacity was the lowest, between 0.6 and 0.7. In addition, the results of Bartley‘s sphericity test for all items were 0. This showed that the null hypothesis was rejected. The reliability and validity of each part of the scale were good.

3.3. Analysis Method

For public and private environmental behavior, the Probit model is used so that the analysis of the two will be independent of each other. When individuals implement an environmental behavior, they will also adopt other environmental behaviors and these behaviors are not exclusive [50]. The MV-Probit model does not only estimate the regression results of different environmental behaviors, but it can also be used to judge the relationship between different environmental behaviors through the likelihood ratio [51]. The formulas are as follows:

$$y^{\mathrm{*}}={\alpha }_0+{\sum }_i{\beta }_i{x}_i+{\sum }_j{\gamma }_j{Z}_j+\epsilon ,$$

(1)

$$y=\left\{\begin{array}{c}1,y^{\mathrm{*}}>0\\ 0, \mathrm{o}\mathrm{t}\mathrm{h}\mathrm{e}\mathrm{r}\end{array}\right. ,$$

(2)

where is y* the latent variable; y is the observed variable of environmental behavior; $x_i$ is the core independent variable; Z is the control variable; $i$ is the number of core independent variables; j is the number of control variables; α, β and γ are the parameters to be estimated; and ε is the error term, which satisfies ε ~ MVN (0, ψ).

The number of environmental behaviors implemented by individuals belongs to the counting model, so the Poisson model is an ideal estimation method. The dependent variable y represents the number of individual environmental behaviors, which obeys the Poisson distribution with expectation and variance both λ. The formulas are as follows:

$$P=\left\{Y=y\left|x\right.\right\}={\displaystyle \frac{e^{-\lambda }{\lambda }^y}{y!}},$$

(3)

$$E(Y\left|x)=\lambda =\mathit{exp}(x^{\prime }\beta )\right.,$$

(4)

$$\mathit{ln}\left(E\left(Y\left|x\right.\right)\right)=\mathit{ln}(\lambda )={\sum }_k{\beta }_{k}x ,$$

(5)

where $E\left(Y\left|x\right.\right)=Var\left(Y\left|x\right.\right)=\lambda$; and ${\beta }_k$ means that the number of environmental behaviors will change to the original $\mathit{exp}({\beta }_k)$ times when x changes by one unit.

4. Results

4.1. Changes in Environmental Awareness and Environmental Behavior

4.1.1. Analysis on the Changes of Public Environmental Awareness and Behavior at the National Level

Table 2. Environmental awareness in China in 2003, 2010 and 2021.

Environmental Awareness	2003	2010	2003–2010	2021	2010–2021
Environmental value	2.77	3.45	+24.5%	3.47	+0.59%
Environmental knowledge	3.15	3.55	+12.6%	3.56	+0.28%
Environmental responsibility	/	2.78	/	3.01	+8.27%
Environmental capacity	/	2.75	/	2.55	−7.27%
Environmental concern	3.17	3.67	+15.77%	3.57	−2.72%

shows the status of environmental awareness from 2003 to 2021, i.e., the average score of each of the five dimensions.

Analysis of the Change of Public Awareness at the National Level

In terms of changing trends, there were differences in the environmental awareness of Chinese citizens in the five different dimensions from 2003 to 2021. Environmental values and knowledge showed a continuous upward trend. There was a trend of environmental concern rising first and then falling. Compared to 2010, the environmental responsibility of Chinese citizens increased in 2021, while their environmental capabilities slightly declined. The reason may be attributed to the fact that with the development of refined environmental governance and the increasing diversity of environmental issues coming into the public view, most members of the public believe that environmental protection requires professional, personnel and technical expertise, leading to a slight decline in the environmental capabilities of the general public [52]. Chinese citizens’ environmental concerns decreased between 2010 and 2021. This may be related to the continuous improvement in China’s overall ecological environment and the gradual reduction of negative news reports related to the environment.

From the perspective of the magnitude of change, the environmental values, environmental knowledge and environmental concerns of Chinese citizens all illustrated a rapid growth rate from 2003 to 2010. However, between 2010 and 2021, the changes were slow. Environmental concerns have even shown a slight decline. The release of the “China Environmental Status Bulletin” in 2003 catapulted environmental issues to the forefront of societal discourse, maintaining this attention from 2006 to 2008. Chinese citizens gained a more objective understanding of the harmonious coexistence between humans and nature, as well as the pressing crisis in resources and ecology. Therefore, during this period, there has been a rapid improvement in environmental concern, environmental values, and environmental knowledge. Since the 18th National Congress of the Communist Party of China, the strategic position of ‘ecological civilization construction’ has been significantly improved, public environmental advocates have increased, and the collective sense of environmental stewardship has been steadily boosted. However, due to the decline in environmental protection technology acquisition, the public’s awareness of environmental protection ability has shown a weakening trend.

In terms of overall scores, almost all dimensions of environmental awareness scored 3 points or more in 2021 (with a maximum of 5). This indicates that after nearly 20 years of constant efforts, the Chinese citizens’ environmental awareness has reached the ideal level. Environmental concern scored the highest, with a value of 3.57. The score of environmental capacity was the lowest, at only 2.55. This indicates that Chinese citizens lack confidence in improving the environmental situation via their own efforts. Overall, the environmental awareness of Chinese citizens presents a state of high identity, high knowledge, high concern, strong responsibility and low ability How to improve the environmental capability of Chinese citizens should be the focus of guiding them to practice more ecological and environmental behaviors.

Analysis on the Change of Public Environmental Behavior at the National Level

Table 3. Environmental behavior in China in 2003, 2010 and 2021.

	Private Environmental Behavior
	2003	2010	2021	2003	2010	2021
	Green consumption			Recycling
Always	/	33.6%	36.0%	/	44.6%	32.5%
Often	22.6%	41.2%	39.7%	15.1%	24.0%	28.2%
Sometime	27.5%	16.7%	14.3%	21.7%	19.7%	23.1%
Never	49.6%	8.5%	8.0%	62.9%	9.7%	16.2%
	Public environmental behavior
	2003	2010	2021	2003	2010	2021
	Joining an environmental organization			Signing a petition on an environmental issue
Yes	41.3%	1.7%	5.3%	17.3%	1.3%	2.8%
No	58.7%	98.3%	94.7%	82.6%	98.7%	97.2%
	Donating money to an environmental group			Participating in a protest or demonstration for an environmental issue
Yes	31.9%	5.3%	8.6%	/	0.2%	0.5%
No	69.1%	94.7%	91.4%	/	99.8%	99.5%

shows the state of environmental behavior from 2003 to 2021, i.e., the participation rate of the Chinese public in environmental behavior in the public domain and in environmental behavior in the private domain.

From the perspective of green consumption, nearly half of the citizens never engaged in green consumption in 2003, 22.6% of the public “often” implement green consumption behavior, but in 2010 and 2021, nearly 75% of the public “always” and “often” made green consumption choices. About 15% of citizens “sometime” engage in green consumption, and the number of people who “always” consume green in 2021 increased by 1.4% compared with 2010. From the perspective of recycling, in 2003, over 60% of citizens never engaged in recycling. “Always” recycling accounted for 15.1% of the public, and this continued to rise in 2010 and 2021 to 28.2%. In 2010, the proportion of the public who “always” recycled was as high as 44.6%. Although it decreased in 2021, it was also close to 30%. As a result, the frequency of citizens’ green consumption and recycling behavior rose from 2003 to 2021. Moreover, in 2003–2010, this increase was more evident. It is worth noting that the proportion of citizens who never engaged in green consumption and recycling in 2021 has increased compared to 2010.

From the perspective of public environmental behavior, the proportion of citizens engaging in various environmental activities has risen since 2010, yet it remains notably low. This could be attributed to the public’s insufficient access to relevant information channels, potentially due to the underdevelopment of environmental protection organizations and the limited participation in environmental protection actions. The participation rates of “donating money” and “joining an organization” showed a slight upward trend, but the participation rate of “participating in a protest or demonstration” was nearly negligible. This discrepancy likely stems from the social and political risks associated with citizens’ participation in protests [53].

Most citizens do not choose activities with certain risk characteristics due to considerations of self-interest. For the years 2003 and 2010, there was a significant decrease in the participation rates of public environmental behaviors, which aligns with the previous research [54]. There are two reasons for this. First, in 1998, China designated environmental protection as a key infrastructure investment area. The State Environmental Protection Administration of China promulgated the “Outline of National Environmental Protection Work (1998–2002)”. This move elevated environmental issues to a focal point, fostering a positive atmosphere that spurred citizen involvement in environmental protection activities. However, with the decline in the popularity of environmental protection, citizens returned to their original state, leading to a reduction in the implementation of environmental behavior. Second, people tend to assess the impact of repeated behaviors. If the effect of the behavior does not meet their expectations, they may doubt the efficacy of the behavior [55]. The “Songhua River Water Pollution Incident” and “Tai hu Lake Water Pollution Incident” that occurred from 2005 to 2010 elicited negative sentiments among citizens, thereby affecting the implementation rate of environmental behavior.

By comparing private and public environmental behaviors, it is found that citizens had a higher participation rate in private environmental behavior. This indicated that the environmental behavior implemented by citizens is more at the individual level and less at the public level.

4.1.2. Analysis of Changes in Public Environmental Awareness and Behavior in Different Regions

Due to the lack of some data in 2003, we only used data from 2010 and 2021 to produce a spatial difference map for China by province. (According to China’s National Development and Reform Commission’s (NDRC) division of the country into east, center, and west, the east of the country includes 11 provinces (municipalities) including Beijing, Tianjin, Hebei, Liaoning, Shanghai, Jiangsu, Zhejiang, Fujian, Shandong, Guangdong and Hainan; the center of the country includes Shanxi, Jilin, Heilongjiang, Anhui, Jiangxi, Henan, Hubei and Hunan; and the west of the country includes Sichuan, Chongqing, Guizhou, Yunnan, Xizang, Shaanxi, Gansu, Qinghai and Ningxia, Xinjiang, Guangxi and Inner Mongolia.). To more clearly show the spatial and temporal patterns of Chinese public environmental awareness and environmental behavior in different regions, we used the average value of each dimension of public environmental awareness as a numerical value, and the assignment of private and public environmental behavior as outlined previously. Public environmental behavior is represented by a numerical value, while the recycling and green consumption behaviors as part of private environmental behavior still use the dichotomous variable. The details are shown in Figure 2 and Figure 3.

Analysis of the Changes in Environmental Awareness in Different Regions

The overall score indicates an improvement in the environmental awareness of the Chinese public. In the eastern region, scores are consistently above 3 points (out of a maximum of 5), which is significantly higher than those in the central and western regions. Notably, the western region recorded the lowest environmental awareness score in 2010, with scores nearly all below 3 points. Over the ten-year period from 2010 to 2021, the trends in public environmental awareness varied across the eastern, central and western regions. Specifically, the western region experienced a decline in environmental awareness, while both the central and eastern regions saw an increase during the same period.

Analysis of Environmental Behavior Change in Different Regions

From the perspective of green consumption behavior in the public domain, the eastern region exhibits the highest proportion of green consumption among the three regions studied. Between 2010 and 2021, this proportion has consistently increased in the eastern region. In contrast, the central region, particularly in provinces such as Shanxi, Henan, Hunan and Hubei, demonstrates the lowest levels of public green consumption. Regarding recycling behavior, the eastern region again leads with the highest participation rates, while the central region shows the least engagement in recycling activities. Notably, the overall trend in recycling behavior across the eastern, central and western regions has seen a decline.

Examining public environmental behavior from 2010 to 2021 reveals that participation rates have increased across all three regions. However, the overall participation remains low. Interestingly, the eastern region has the lowest participation rate in public environmental behavior, while the western region has the highest. Despite the low rates, the upward trend in participation indicates significant progress in the establishment of environmental protection organizations in China during this period.

4.2. The Relationship between Environmental Awareness and Environmental Behavior

4.2.1. Direct Effect Analysis

First, environmental value had a significant positive impact on private environmental behavior. However, it had no significant effect on public environmental behavior and the number of environmental behaviors. The possible reason is that the uncontrollability, invisibility and possible negative effects of public environmental behavior are contrary to the utilitarianism of the holders of environmental values [56]. As a result, the possibility of citizens holding environmental values to implement public environmental behavior is reduced. Secondly, environmental knowledge had a significant positive impact on public environmental behavior and green consumption behavior. However, the impact on recycling behavior was not significant. This could be attributed to the fact that individuals who engage in recycling are predominantly older, and they tend to have a limited capacity to assimilate environmental knowledge. This results in the disconnection between environmental knowledge and recycling behavior. Thirdly, environmental responsibility, environmental capacity and environmental concern all have a significant positive impact on public and private environmental behaviors as well as the number of environmental behaviors.

Among the control variables, age, individual cognition, household registration, gender and education level had a significant impact on private environmental behavior. The findings suggested that younger females with higher levels of cognition, education and urban residency were more inclined to engage in private environmental behavior. Class and political status have a significant impact on public environmental behavior. The analysis showed that Chinese Communist Party members and citizens in a high socioeconomic class were more inclined to implement public environmental behaviors. Individual cognition, class and household registration were significantly and positively correlated with the number of environmental behaviors implemented by the public. This indicates that individuals with higher cognition and social class, as well as urban residents, are more prone to undertake multiple environmental behaviors, and the urban public has more environmental behaviors than rural ones. The year variable had a significant positive impact on green consumption, public environmental behavior and the number of environmental behaviors. This showed that compared with 2010, there was a substantial increase in citizens’ green consumption and public environmental behavior in 2021.

4.2.2. Analysis of Mediating Effects

Based on the research of Wen et al. [57], this paper examined the mediating effect of the willingness to protect the environment. First, Probit regressions were run on all variables except environmental willingness. Secondly, according to the method of Baron and Kenny [58], the influence of environmental awareness variables and other control variables on willingness to protect the environment was assessed. The findings indicated that at the significance level of 5%, environmental responsibility, environmental concern and environmental ability had a significant impact on willingness to protect the environment, but environmental values and environmental knowledge had no significant impact. Therefore, the efficacy of this test was low and the Sobel test was required. Thirdly, the Sobel test is used to identify variables that were significant in the first step and not significant in the second step. The result showed that environmental value indirectly affects green consumption behavior and recycling behavior through willingness to protect the environment through Z scores of 2.74 and 2.70, respectively. This showed that the mediating effect of environmental willingness on the relationship between environmental value and private environmental behavior was significant. The Z value of the impact of environmental knowledge on public environmental behavior was 2.68, and the Z value of the impact on green consumption behavior was 0.64. This shows that the mediating effect of the willingness to protect the environment in the relationship between environmental knowledge and public environmental behavior was significant, but the mediating effect on green consumption behavior was not significant.

In summary, environmental value had both direct and indirect effects on citizens’ private environmental behaviors. Environmental knowledge had both direct and indirect effects on public environmental behaviors. However, it had only a direct impact on green consumption behavior. In addition, variables that had both direct and indirect effects in the test could not be estimated in the same equation [59]. To ensure the accuracy of the results, the equations with and without variables related to the willingness to protect the environment are regressed, respectively. See

Table 4. The Regression results of environmental behaviors with and without variables related to willingness to protect the environment in China.

Without Willingness to Protect the Environment	Model 1			Model 2
	Green Consumption	Recycling	Public Environmental Behavior	Number of Environmental Behaviors
Environmental value	0.196 ** (0.081)	0.268 ** (0.084)	0.019 (0.1)	0.016 (0.085)
Environmental knowledge	0.968 *** (0.087)	0.060 (0.089)	0.667 *** (0.108)	0.731 *** (0.078)
Environmental responsibility	0.832 *** (0.101)	0.519 *** (0.088)	0.527 *** (0.127)	0.689 *** (0.079)
Environmental concern	0.220 *** (0.019)	0.220 *** (0.019)	0.145 *** (0.124)	0.269 *** (0.017)
Environmental capacity	0.399 *** (0.093)	0.449 *** (0.094)	0.776 *** (0.116)	0.768 *** (0.074)
Control variable
Gender	−0.022 (0.035)	−0.068 * (0.035)	0.046 (0.044)	0.007 (0.028)
Age	0.010 *** (0.001)	0.005 *** (0.001)	−0.012 *** (0.001)	−0.001 (0.001)
Political status	0.013 (0.054)	−0.061 (0.55)	0.152 ** (0.065)	0.074 (0.040)
Education level	0.060 ** (0.022)	0.008 (0.022)	0.015 (0.027)	−0.004 (0.017)
Cognition	0.392 ** (0.126)	0.822 *** (0.127)	−0.094 (0.154)	0.557 *** (0.099)
Income	0.0002 (0.003)	0.002 (0.004)	0.001 (0.005)	0.002 (0.003)
Class identity	−0.019 (0.018)	0.007 (0.018)	0.083 *** (0.023)	0.030 * (0.016)
Household registration	0.066 * (0.033)	0.161 *** (0.033)	−0.043 (0.041)	0.065 ** (0.027)
year variable	0.694 *** (0.053)	−0.050 (0.054)	0.352 *** (0.068)	0.466 *** (0.043)
Constant	−3.133 *** (0.137)	−2.729 *** (0.138)	−2.239 *** (0.174)	−2.42 *** (0.113)
atrho21	0.484 *** (0.022)			-
atrho31	0.218 *** (0.026)			-
atrho32	0.157 *** (0.026)			-
Waldχ2	890.02			1217.03
Prob > χ2	0.00			0.00
Log Likelihood	−9135			-
Pseudo R2	-			0.073
With Willingness to Protect the Environment	Model 3			Model 4
	Green Consumption	Recycling	Public Environmental Behavior	Number of Environmental Behaviors
Environmental value			0.010 (0.101)	0.018 (0.084)
Environmental knowledge	0.911 *** (0.086)	−0.02 (0.087)
Environmental responsibility
Environmental concern
Environmental capacity
Willingness to protect the environment	1.086 *** (0.727)	0.805 *** (0.072)	1.005 *** (0.094)	1.253 *** (0.068)
Control variable	Controlled			Controlled

Ps: 1. The dependent variable regression likelihood ratio test showed that rho21 = rho31 = rho32 = 0.000, c2(3) = 19.65 and was significant at the 1% level. Therefore, the null hypothesis that “the dependent variable correlation coefficient is 0” was rejected, indicating the effectiveness of using the Mv probit model. 2. ***, ** and *, respectively indicated significance at the 1%, 5% and 10% levels. 3. The numbers in parentheses represent robust standard errors.

for the results.

4.2.3. Relationship between Different Environmental Behaviors

The results in Table 4 showed that the correlation coefficients atrho21, atrho32 and atrho31 for public and private environmental behaviors passed the significance test. This showed that there were correlations not only between recycling and green consumption but also between public and private environmental behaviors. This confirms the view that the implementation of environmental behavior by citizens is not an independent decision-making process [11].

4.3. Robustness Test

A robustness test of the model regression results was carried out by adding additional variables to control for factors that are not considered in the baseline regression. On one hand, mass media, particularly the internet, have emerged as significant educational tools that promote environmental behavior among individuals. The frequency of using the Internet is closely related to the implementation of individual environmental behavior [60,61]. The variable “Frequency of Internet use” was measured in the survey by “In the past year, how often did you use the Internet?”. On the other hand, government policies can limit individual actions at the macro level, and the implementation of these policies will influence individual environmental behavior. In other words, the more effective the implementation of environmental policies, the more likely environmental behavior is to occur [62]. This variable was measured by “How do you think the environmental protection work of the central and local governments has been done in the past five years?” From the perspective of robustness testing, the study’s results in

Table 5. Robustness test.

Variables	Mv-Probit			Poisson
	Green Consumption	Recycling	Public Environmental Behavior	Number of Environmental Behaviors
Environmental awareness
Environmental value	0.3274 *** (0.082)	0.312 *** (0.084)	0.041 (0.101)	0.018 (0.084)
Environmental knowledge	0.546 *** (0.097)	−0.035 (0.098)	0.531 ** (0.120)	0.748 *** (0.088)
Environmental responsibility	0.764 *** (0.012)	0.491 *** (0.102)	0.507 *** (0.125)	0.709 ** (0.092)
Environmental concern	0.210 *** (0.019)	0.212 *** (0.020)	0.142 *** (0.026)	0.253 *** (0.021)
Environmental capacity	0.428 *** (0.094)	0.467 *** (0.094)	0.78 *** (0.116)	0.719 *** (0.086)
Other variables	Controlled	Controlled	Controlled	Controlled
Control variable	−3.479 *** (0.148)	−2.829 *** (0.141)	−2.335 *** (0.179)	−2.71 *** (0.132)
atrho21	0.467 *** (0.022)			-
atrho31	0.227 *** (0.026)			-
atrho32	0.153 *** (0.026)			-
Waldχ2	1272.18			989.09
Prob > χ2	0.00			0.00
Log Likelihood	−8924.6			-
Pseudo R2	-			0.071

Ps: *** and **, respectively indicated significance at the 1% and 5% levels.

showed no significant changes, indicating that the estimation results are robust.

5. Discussion

The level of citizen’s environmental awareness is an important indicator for measuring the degree of civilization of a modern country. China has entered a stage of high-quality development, and the supporting role of the ecological environment is becoming increasingly evident. The improvement and maintenance of the environment depend on the implementation of citizens’ “environmental behavior”. Cognition determines behavior. If citizens lack environmental awareness, it is difficult for them to implement environmental behavior [63]. The initiation, formation and maintenance of environmental awareness and environmental behavior is a dynamic process. If it is ignored, the guiding significance of relevant research conclusions to reality will be significantly weakened [64]. However, most of the existing research is limited to a certain group of people at a single point in time, which lacks horizontal comparison and vertical change research [9,10,39]. As a result, we do not know the factors that affect citizens’ environmental awareness over time. Hence, this paper investigated the dynamic process of national public environmental awareness and behavior using data from three national surveys from 2003 to 2021 to analyze the degree of influence and path of different dimensions of environmental awareness on different environmental behaviors.

The results showed that the environmental protection work carried out in China in the past 20 years has achieved remarkable results in improving citizens’ environmental awareness and private environmental behavior. From a diachronic perspective, the public’s environmental values, environmental knowledge and environmental responsibility have increased from 2003 to 2010, which is the same as the results found in the “2010 National Public Environmental Awareness Survey Report “released by the China Environmental Awareness Project Office. The decline in environmental concerns in the last decade should be taken seriously by policymakers. Citizens’ environmental concerns represent their level of attentiveness to ecological matters and potential remedies, which is a prerequisite for ensuring the implementation of environmental behavior [65]. Consequently, the pivotal focus of policy development should be on how to raise citizens’ environmental awareness and stimulate them to actively pursue solutions to these problems. At present, beyond the low engagement of residents in public environmental behavior, the weak environmental capacity of the public is also recognized to be one of the obstacles to the construction of ecological civilization in the new era in China. Therefore, while concentrating on boosting the participation rate in public environmental behavior, the improvement of citizens’ environmental capabilities also needs to be taken into account. The improvement of citizens’ environmental capacity, on the one hand, comes from the environmental knowledge they have learned, and on the other hand, from their practice of transforming environmental knowledge into capacity [66]. According to the results of this survey, citizens’ environmental knowledge is generally robust.

From 2003 to 2021, the public’s private environmental behavior increased rapidly, but public environmental behavior participation rates showed a downward trend. This is due to the fact that the Chinese public regards themselves more as the victims of environmental pollution than the initiators of environmental protection in environmental issues. The public’s focus is on major pollution incidents, while they are less aware of the environmental pollution around them, which makes it extremely difficult to implement environmental behavior [67]. Another major finding of the research results is that the public environmental behavior of the Chinese public has increased in the past decade. China’s current bottom-up environmental protection movement and the constant development of non-governmental environmental protection organizations indicate that the environmental awareness of Chinese citizens is increasing. However, compared with the participation of environmental behavior in the private domain, the participation of environmental behavior in the public domain is extremely low. This is because environmentally friendly public actions are often carried out at the expense of their own interests. Under the implementation of government-led environmental governance and environmental policies, certain environmental behaviors in the commons, such as ‘demonstrations to protect the environment’, are considered to be rebellious activities. This requires political risks and economic costs, resulting in the majority of the public not being willing to implement corresponding environmental behaviors.

This paper confirms that environmental awareness has a significant impact on environmental behavior, which is consistent with the conclusion of Ghorbani et al. [68] and Luo et al. that [69] environmental awareness plays an important role in the implementation of individual environmental behavior. First, environmental values have a significant positive impact on private environmental behavior, but they have no significant impact on public environmental behavior. This is inconsistent with the research results of Janmaimool and Khajohnmanee [70] in the Thai student group. They believe that the moral concept of altruism will promote an individual’s public environmental behavior. Thinking about the reasons behind these results, we believe that private environmental behavior occurs on a daily basis and is easy to implement, and the implementation of private environmental behavior can have a positive impact on public environmental behaviors. However, the behavioral effect of implementing the environmental behavior of the commons is likely to be negative, which is contrary to the utilitarianism of the holders of environmental value [51]. This means that residents who hold environmental values do not necessarily implement public environmental behaviors. Secondly, environmental knowledge has a significant positive impact on public environmental behavior and green consumption behavior, but there is no significant impact on recycling behavior. This is inconsistent with the conclusion of Zhu et al. [71], i.e., that subjective knowledge has a positive correlation with individual recycling behavior intentions. The current situation in China is that recycling behavior is less prevalent among the elderly, but the main source of environmental knowledge in China comes from school education, which leads to fewer channels for the elderly to acquire environmental knowledge, resulting in the disconnect between environmental knowledge and recycling behavior. Although the public’s environmental behavior in 2010–2021 is ‘high private domain environmental behavior, low public domain environmental behavior’, the empirical results show that there is a correlation between the public’s public and private environmental behaviors, which is consistent with previous research results [42]. Finally, environmental intention plays a mediating role in the relationship between environmental awareness and environmental behavior. The results of Chang et al. also support this result [72].

It is noteworthy that the impact of environmental awareness on environmental behavior shows significant differences in gender, age, education and other demographic characteristics, which goes beyond findings from previous studies. The results also indicate that urban residents are more willing to implement environmental behavior than rural residents. This could be attributed to the rapid popularization of the Internet in urban areas [61], where non-agricultural residents have increased access to online resources [62], newspapers, magazines and environmental organizations, leading to a higher degree of awareness of environmental issues and environmental knowledge. Thus, they are more likely to implement environmental behaviors.

6. Conclusions and Policy Implications

6.1. Conclusions

Guided by the theory of planned behavior, based on the data from the CGSS in 2003, 2010 and 2021, this paper employed the Probit model and Poisson model for empirical analysis. The results are as follows. First, the environmental awareness of Chinese citizens has improved in the past 20 years, presenting a state of high identity, high knowledge, high concern, strong responsibility, low ability Environmental values, environmental knowledge and environmental concerns increased by 24.5%, 12.6% and 12.77%, respectively, while environmental capacity decreased by 7.27% between 2010 and 2021. The environmental capacity is low and demonstrates a downward trend. In 2021, the environmental capacity was only 2.55, decreasing by 7.27% from 2010 to 2021. Environmental attention has increased rapidly from 2003 to 2010 and then experienced a slight downturn from 2010 to 2021, but the overall trend is upward. From the perspective of the differences in public environmental awareness in different regions, the eastern region has the highest score for environmental awareness, and the scores of environmental values, environmental knowledge and environmental responsibility in 2021 are all greater than 3.5, while the western region has the lowest score.

Secondly, there exists not only a correlation between green consumption and recycling but also between public environmental behavior and private environmental behavior. Citizens’ participation in environmental behavior is manifested as “high private, low public”. In private environmental behavior, the participation rate of green consumption increased over time, in 2021, with nearly 75% of the public always consuming green but the participation rate of recycling decreased from 2010 to 2021. The participation rate of joining environmental protection organizations decreased from 41.3% in 2003 to 1.7% in 2021, a decrease of 41.3%, but there was a slight increase from 2010–2021. In 2021, nearly 50% of people in the eastern region engaged in green consumption, but the participation rate of environmental behavior in the eastern region is among the lowest of the three regions, while environmental behaviors in the western region are the highest.

Thirdly, different dimensions of environmental awareness have different effects on public and environmental behaviors, and the number of environmental behaviors. Environmental responsibility, environmental concern and environmental capacity not only directly affect but also indirectly affect citizens’ public and private environmental behaviors, and the number of behaviors through willingness to protect the environment. Environmental values only have direct and indirect effects on private environmental behavior. Environmental knowledge has both direct and indirect effects on public environmental behavior and the number of environmental behaviors but only a direct impact on green consumption.

6.2. Policy Implications

Based on the above analysis, there are three main policy implications. First, recognize the all-around improvement in public environmental awareness but seek ways to improve environmental capacity, which is a current weakness of citizens’ environmental awareness. To do this, it is necessary to improve citizens’ operational skills and lower the technical barriers associated with citizens’ environmental behavior. The government can influence the development of environmental protection practice courses in schools or help in the establishment of community and village-based environmental organizations. The government can also provide guidance and financial support in the process of public implementation of environmental behavior, such as providing garbage bin classification indicators, green product purchase subsidies, etc. Finally, the government can improve citizens’ access to and ease of use of environmental protection technologies.

Secondly, guide groups with high participation rates in private environmental behaviors to implement public environmental behaviors to improve citizens’ public and private environmental behaviors more generally. At the same time, the government should pay attention to the guidance of public environmental behavior and the response to citizens’ participation in public environmental behavior. When citizens exhibit a high willingness to protect the environment, the government needs to improve the conditions and expand opportunities for citizen participation in public environmental behavior. At the same time, it is necessary to improve the public’s environmental behavior according to local conditions. In the western region, the government should pay attention to the guidance of green consumption behavior in the private domain, while the eastern region needs to pay more attention to improving the participation rate of various environmental behaviors in the public domain. Different environmental policies should be designed for urban and rural residents. The policy focus of urban residents should be on garbage collection and green purchase behavior, with the elderly as the main target of publicity. People with higher education levels and individual cognition will obtain higher policy effects mainly from environmental–mental protection purchase behavior. People with a high sense of class identity mainly start with the environmental policy of the commons Thirdly, the government should prioritize environmental awareness as a foundation for encouraging citizens to adhere to the “Ten Norms of Citizens’ Ecological Environment Behavior”. The implementation of citizens’ environmental behavior norms depends on the improvement of citizens’ environmental awareness. On the one hand, adopting various methods that resonate with citizens’ daily lives can promote ecological civilization. This can improve citizens’ recognition and attention to the value of the environment, thereby encouraging them to engage in energy conservation, green consumption, low-carbon travel and recycling. On the other hand, fostering an environmentally friendly social atmosphere is imperative. Encouraging environmental volunteers and penalizing those who damage the environment can ensure that everyone has the opportunity to participate in environmental protection, working together towards the harmonious coexistence of humanity and nature.

7. Limitations and Future Research Directions

We acknowledge that the decision-making process behind the public’s implementation of environmental behavior is complex, and many of their decisions and actions are shaped by personal interests and situational factors [73]. The selection of influencing factors based on previous studies may, therefore, have led to omissions. The results of this study indicated that even when citizens possess strong values, they may refrain from engaging in public environmental behavior because the negative impact of public environmental behavior may be detrimental to citizens’ own interests [74]. Cost, income and suitability are crucial factors in the implementation of behavior [75]. Future research on environmental awareness and behavior should investigate the complex interactions among environmental factors, the costs of environmental behavior and other potential influences. Understanding how these factors influence behavior can aid in the development of effective policies and encourage citizens to engage in environmentally responsible actions.