Introduction

The urban domestic waste generated in China has increased gradually in recent years. According to the Ministry of Ecology and Environment of China’s Annual Report on the Prevention and Control of Solid Waste Pollution in Large and Medium-sized Cities in 2020, 196 cities nationwide generated 235.602 million tons of domestic waste and disposed of 234.872 million tons in 2019. As waste disposal is imminent, waste classification and recycling have attracted unprecedented attention. Since the Chinese State Council promulgated the Implementation Plan of the Domestic Waste Classification System in April 2019, source classification of domestic waste has become a major environmental policy that comprehensively promoted in major cities in China1 In April 2019, the Ministry of Housing and Urban-Rural Development and other departments jointly issued the Notice on the Comprehensive Implementation of Domestic Waste Classification in Cities at and above the Prefecture Level Nationwide, requiring 46 key cities to basically establish a household waste disposal system. In July 2020, the Implementation Plan to Make Up for Shortcomings in Urban Domestic Waste Classification and Disposal Facilities instructed cities at and above the prefecture level to basically complete a domestic waste disposal system for classified dumping, classified collection, classified transportation and classified disposal before 2024. In November 2020, the Ministry of Housing and Urban-Rural Development of China, together with other departments, issued Several Opinions on Further Promoting Domestic Waste Classification, setting goals for the public to generally form the habit of domestic waste classification and for the recycling rate of urban domestic waste in the country to exceed 35%.

In recent years, the research on personal waste classification behaviors has become a new focus, mainly in terms of individual psychology, social psychology and the external context of personal involvement2,3, willingness to pay4 as well as institutions and policies5. However, relatively few studies have focused on the accuracy of waste classification, and there is no econometric analysis of the accuracy of waste classification and its influencing factors has been found in the previous literature. The accurate classification of domestic waste, a vital precondition for resource recycling, plays an undeniable role in promoting the recycling rate. In the context of increasing amount of domestic waste generation, continuous introduction of waste classification regulations, and the increasing attention to resource recycling, it is of crucial practical significance and theoretical value in paying due attention to the mechanism of environmental relations, institution-based trust and community attachment on the accuracy of waste classification.

First, environmental concern measures the degree to which a person recognizes environmental issues and supports addressing them6. Environmental concern positively impacts individual involvement in waste classification, and individuals with higher degree of environmental concern are more likely to engage in environmentally friendly behaviors7,8. Therefore, low environmental awareness will inhibit the individual’s understanding of the harms of domestic waste and the importance of resource recycling to a certain extent, which curbs their enthusiasm for involvement in domestic waste classification, and thus hindering the improvement of the accuracy of waste classification. Hence, it is well worth examining in-depth to unveil the relationship between environmental concern and the accuracy of domestic waste classification. Secondly, institution-based trust is defined as trust with the institution as its object9. The effect of institution-based trust on improving individual environmental concern and encouraging environmentally friendly behaviors, has been confirmed in variety of literature10,11. Through promoting environmental concern, institution-based trust can indirectly influence the individual’s motivation and willingness to engage in environmentally friendly behaviors12,13. Therefore, we not only explore the impact of institution-based trust on the accuracy of waste classification but also incorporate institution-based trust into our research framework of the relationship between environmental concern and waste classification accuracy in an attempt to reveal its moderating role better. Finally, in urban China, the community has become a tool and carrier for social reorganization as well as the re-socialization of urban individuals14, acting as a grass-root organization that arranges domestic waste classification, personal sense of belonging to the community may affect the individual’s accuracy of waste classification. Community attachment may also indirectly influence the individual’s motivation and willingness to engage in environmentally friendly behaviors due to its influence on institution trust15,16. Therefore, while exploring the influence of community attachment on waste classification accuracy, this paper will further integrate community attachment into the research framework, concerning the relationship between institution-based trust and waste classification accuracy, in order to shed light on the moderating role of community attachment.

To summarize, the current research manifests the following deficiencies: first, there are no studies that investigate the factors influencing the accuracy of waste classification; second, few scholars discussed personal environmentally friendly behaviors and waste classification from the perspective of community attachment, while even fewer tried to introduce arguments about the moderating effect of community attachment on the relationship between institution-based trust and waste classification accuracy. In this paper, through a multivariate ordered logistic model, we will use microscopic research data collected from urban communities in Hangzhou to empirically analyze the influence of community attachment, environmental concern, and institution-based trust on the accuracy of personal waste classification. We will also further discuss the moderating effect of institution-based trust on the relationship between environmental concern and personal waste classification accuracy and the moderating effect of community attachment on the relationship between institution-based trust and personal waste classification accuracy. In this work, the first section is the introduction. After this section is the theory review and the research hypothesis in Sect. 2. The model, variable and data are discussed in Sect. 3 while the results and analysis are presented in Sect. 4. Section 5 offered a series conclusion and policy implications.

Theory review and research hypothesis

In order to study the factors affecting the accuracy of garbage classification, this paper builds a theoretical framework (Fig. 1). That is, environmental concern directly affects personal waste classification accuracy and that the relationship between environmental concern and personal waste classification accuracy is moderated by institution-based trust, while the relationship between institution-based trust and personal waste classification accuracy is moderated by community attachment.

Fig. 1
figure 1

Analytical framework of the influencing factors and mechanism of the classification accuracy of personal waste of urban residents.

Full size image

Environmental concern and personal waste classification accuracy

The consistency between environmental concern and environmentally friendly behavior is controversial among academics. While some studies have verified the significant positive effect of environmental concern on environmentally friendly behavior and the significant positive impact of environmental concern on the waste classification intention8, others have shown inconsistencies between environmental concern and environmentally friendly behaviors17. The Specification-Activation theory holds that primary value orientation, including self-interested concern, altruistic concern and ecosphere concern, determines individual attitudes toward environmental protection and thus affects the individual’s environmentally friendly behaviors. The Value-Belief-Norm theory18 argues that values direct environmentally friendly behaviors. In other words, values generate beliefs, and beliefs generate standards of behavior, which in turn generate environmentally friendly behaviors. When the public has a deeper understanding of the harm caused by domestic waste, it will not only pay more attention to the ecological issue of environmental pollution, but will also be accompanied by a higher level of environmental concern and environmental awareness. As individuals with better awareness are more inclined to adopt green-oriented behaviors, it infers that those who pay more attention to urban domestic waste management, could implement relevant policies more thoroughly, which leads to the higher waste classification accuracy. Accordingly, this paper forwards the research hypothesis that:

H1 Environmental concern has a significant positive impact on the public’s waste classification accuracy.

Institution-based trust and personal waste classification accuracy

In modern society, institution-based trust is a non-face-to-face and non-interpersonal trust relationship19, which is a constraint force that facilitates regulation of social order and restrains human behavior effectively. Studies have shown that the stock of institution-based trust affects the individual’s expectation of policy risks and effects. The higher the institution-based trust, the less the risk an individual expects from a certain behavior,, the higher the return, which largely determines the behavior; institution-based trust promotes individual willingness to pay for environmental protection and participate in ecological compensation20,21 institution-based trust exerts a significant influence on environmentally friendly behaviors such as waste classification12,22, which is helpful for the government’s environmental governance10. These conclusions suggest that the increasing amount of recognition and trust each individual has in the overall institution building and specific policy measures related to waste classification, the higher the expectation on the effectiveness of waste classification, the higher the degree of participation and coordinates of policies, the higher the possibility of cooperation, the better the implementation and the effect of waste classification policies, and finally, the higher the accuracy of waste classification. Accordingly, this paper forwards the research hypothesis that:

H2 Institution-based trust has a significant positive impact on public’s waste classification accuracy.

Community attachment and personal waste classification accuracy

The Encyclopedia of China defines community attachment as “the psychology of community residents to identify themselves with a group of people in a certain region. This psychology, with represents a confirmation of one’s community identity, is tinged with individual feelings, including devotion, fondness, and attachment to the community.” Early studies mainly focused on measuring community attachment23,24. In recent years, with the ongoing agenda of rural revitalization, the impact of community attachment on community governance and rural tourism is drawing increasing attention25,26, and the majority of relevant studies implicate that community attachment exerts a significant positive influence on promoting tourism benefits and community involvement27. Besides, the issue of community attachment for landless peasants is also hotly debated among researchers28, with certain studies confirming that community attachment plays a significant role in promoting landless peasants’ integration into urban societies Nevertheless, in the field of research on environmentally friendly behaviors, the topic of community attachment is still not frequently studied. Without a doubt, the community serves as the most basic unit in organizing urban domestic waste classification. The stronger the individual’s sense of community attachment, the higher their willingness to involve in community activities29,30. Greater community attachment rightly arouses higher enthusiasm for participating in community activities, having individual more satisfied to follow the community’s agendas, which brings about higher policy effectiveness, leading to higher accuracy of the public’s waste classification. Based on this observation, this paper forwards the research hypothesis that:

H3 Community attachment has a significant positive impact on the public’s waste classification accuracy.

Moderating effect: the indirect impact of institution-based trust and community attachment

Institutional policies, together with other external environmental factors, not only directly affect the willingness and behavior of individuals to participate in environmental protection actions, but also indirectly influence them to a certain extent12. Studies empirically analyzed the moderating or mediation effects of institution-based trust on environmental concern in environmental governance dilemmas10,31. The Attitude-Context-Behavior theory32 argued that environmentally friendly behavior is a product of the interaction between intrinsic environmental attitudes and extrinsic situational factors, with situational factors having a moderating effect on the relationship between environmental attitudes and environmentally friendly behaviors. The integrated Concern-Situation-Behavior model32 pointed out that low-carbon awareness affects low-carbon consumer behaviors, while institution-based trust exerts a moderating effect on the Concern—Behavior relationship. Therefore, it seems that institution-based trust can play a certain role in promoting the relationship between environmental concern and waste classification accuracy. Besides, since the community acts as the most fundamental organizational unit for social governance in China, the stronger the individual’s community attachment, the higher the stock of social capital will be, such as trust in the community33. Furthermore, seeing that trust in the community, which is established on “non-interpersonal” relationships, can be regarded as a form of institution-based trust34,35, thus it can be considered that community attachment promotes the relationship between institution-based trust and waste classification accuracy to a certain extent. The two points mentioned above are also the key purport of this paper. Therefore, this paper forwards the research hypothesis that:

H4 Institution-based trust moderates the relationship between environmental concern and waste classification accuracy and has a promoting effect on it.

H5 Community attachment moderates the relationship between institution-based trust and waste classification accuracy and has a promoting effect on it.

Model construction, variable selection, and data sources

Model construction

The dependent variables in this study are multivariate discrete variables, and the relatively widely used model for this type of variable is the logistic model, commonly applied to predict the probability of behavior occurs in social sciences36,37. To reflect the regression results more intuitively, this study follows the approach of Cosco et al.38 and uses the sequential logistic model (ordinal regression) for regression analysis. Meanwhile, the maximum likelihood method is applied to fit the regression parameter model.

Using the logistic function:

$${\text{P}}({{\text{y}}} = j|xi) = \frac{1}{{1 + e^{ - (\alpha + \beta Xi)} }}$$
(1)

In Eq. (1), \({\text{y}}\) represents five levels of waste classification accuracy, and the value \({\text{j}}(j = 1,2,3,4,5)\) is assigned according to the level of y; \(xi\)represents the factor i that affects waste classification accuracy. The accumulated logistic model is established as follows:

$${\text{Logit}}\;({\text{Pj}}) = \ln [{\text{P}}(y \leq j)/{\text{P}}({\text{y}}\geq j + 1)] = \alpha {\text{j}} + \beta {\text{X}}$$
(2)

In Eq. (2), \({\text{Pj}}\) represents the probability that the individual belongs to level j, and \({\text{Pj}} = {\text{P}}(y = j),j = 1,2,3,4,5,(x1,x2,...,xi)^{\text{T}}\) represents a set of independent variables; \(\alpha{\text{j}}\) is the intercept of the model; v is a set of corresponding regression coefficients. After obtaining the parameter estimates of \(\alpha{\text{j}}\) and \(\beta\), the probability of a particular situation (e.g. \(y = j\)) can be acquired through the following equation:

$$P(y \leq j/x) = \frac{{e^{ - (\alpha + \beta xi)} }}{{1 + e^{ - (\alpha + \beta xi)} }}$$
(3)

Selection and description of variables

Dependent variable

The dependent variable in this paper is the personal waste classification accuracy rate, a discrete variable characterized by the interval in which the respondents independently decide their accuracy rates. Combined with the preliminary investigation results and the actual situation, the question of personal waste classification accuracy is designed as “Your waste classification accuracy rate is about ( )”. The answer options are “Below 20%, 20%—Below 40%, 40%—Below 60%, 60%—Below 80%, 80%–100%.” The higher the accuracy of waste classification, the higher the corresponding value.

Independent variables

All variables are presented in Table 1.

(1) Environmental concern. Since its introduction in 1978, Dunlap’s NEP (New Environmental Paradigm) single-dimension scale has been widely used worldwide. This paper sets up five topics to measure environmental concerns according to the particularity of waste classification. Each question corresponds to an orientation: ①Does human destruction of nature usually lead to catastrophic consequences? ② Are humans currently abusing and destroying the environment? ③ Do animals and plants have the same right to live as humans? ④ Must humans obey the laws of nature? ⑤ Are resources (such as water) and space limited on earth? The options are “strongly disagreed “, " disagreed “, “unclear”, “agreed”, “strongly agreed”. Each option scores 1, 2, 3, 4, or 5 out of 5. Concern for the environmental is described as the average of the scores. The higher this figure, the higher the environmental concern index.

(2) Institution-based trust. Regarding the dimension and measurement of institution-based trust, some studies use direct questions to ask respondents whether they trust or not, and some studies choose the trust in the people or things in question as a proxy for institutional trust, depending on the content of the study35. By synthesizing the views of previous scholars, this paper summarizes the acceptance of the nature and purpose of institutions as the legitimacy dimension of institution-based trust and summarizes the institution-based trust generated by the effective implementation as the effectiveness dimension. For the legitimacy dimension of institution-based trust, two questions have been designed: “Do you think there is anything unreasonable about government policies?” “Do you think government policies are in our interest?“. For the effectiveness dimension, two questions have been designed: “What do you think of the overall implementation of government policies?” “Do you agree that most of government workers can effectively implement government policies?” Each option scores 1, 2, 3, 4, or 5 out of 5. The degree of institution-based trust is described with the average of the scores. The higher this figure, the higher the institution-based trust index.

(3) Community attachment. Previous scholars’ measurement methods of community attachment can be divided into three categories: the first is to cite scales designed by Kasarda&Janowitz36 and Gerson37; the second is to measure the three dimensions of dedication, fondness, and attachment to the community; the third is to measure the factors influencing community attachment. This paper mainly sets the dimensions of the factors affecting community attachment. We measured five dimensions: length of residence, interpersonal relationship, community involvement, community satisfaction, and emotional attachment. We designed five questions: “How many years have you lived in your community?” “How do you relate to others in your community?” “Do you regularly participate in activities organized in your community?” “How is your overall satisfaction with the community you live in?” “If you have to move away from this community for some reason, will you feel sorry when you leave?“. Each option scores 1, 2, 3, 4, or 5 out of 5. The degree of community attachment is described with the average of the scores. The higher this figure, the higher the community attachment index.

Table 1 Dimensions of variables.
Full size table

Control variables

Individual characteristics such as gender, age, education, annual household income and occupation will have an impact on individual environmentally friendly behaviors38,39,40. Therefore, gender, age, education, annual household income and occupation are taken as the control variables of this study. The meanings, assignments, and descriptive statistics of the variables above are shown in Table 2.

Table 2 Descriptive statistics of variables.
Full size table

Data sources

The data used in this research are derived from a questionnaire survey conducted in Hangzhou in January 2020. The main contents of the questionnaire include the basic information of the surveyed public, the accuracy of waste classification, community attachment, environmental concern, and institution-based trust. The design of the Environmental Care Scale draws on the NEP. Institution-based trust was measured from the two dimensions of legitimacy and validity; community attachment was measured from the five dimensions of the length of residency, interpersonal relationships, community involvement, community satisfaction, and emotional attachment. In this survey, we selected public individuals in Xihu District, Shangcheng District, Xiacheng District, Yuhang District, Gongshu District, Xiaoshan District, Fuyang District, and Jianggan District of Hangzhou City as respondents. A total of 423 questionnaires were posted, of which 418 were valid, with an effective rate of 98.82%. The average age of respondents was 33.79. Among the respondents, 236 were male, and 182 were female, with a gender ratio of 1.3. In the sample region, 97.61% of the respondents had a junior high school degree or above, and 61.24% had a college/bachelor’s degree, reflecting a relatively high overall educational level. The most significant number of respondents were employees, reaching 27.27%. The gender ratio, the average income and other relevant data are similar to those recorded in the Hangzhou Statistical Yearbook 2019, indicating that the sampling is representative and reasonable for the most part.

Reliability and validity test

The reliability and validity test results of this questionnaire shows that Cronbach’s alpha coefficient value is 0.793, reaching above 0.7, indicating that the questionnaire had high reliability and validity. The KMO (Kaiser-Meyer-Olkin) value of the questionnaire is 0.828, which is over 0.6. The chi-square value of Bartlett’s Test of Sphericity is significant, with a p-value well below 0.05, indicating that the questionnaire has good validity.

Statement regarding studies involving human participants

All the methods adopted in this study were carried out in strict accordance with the relevant guidelines and regulations. According to the Declaration of Helsinki, the respondents was informed orally and in writing that anonymous questionnaire survey was used to collect the data solely for academic researchFootnote 1. At the same time, according to the relevant provisions of laws and regulationsFootnote 2, this study does not involve ethical risks such as human health risks and privacy disclosure risks, so it does not involve ethical approval. Ethical approval was not sought for this paper due to the above reasons.

Results and analysis

Descriptive statistical analysis of variables

Descriptive statistical analysis of the personal waste classification accuracy rate

The accuracy rate of personal waste classification is distributed at all levels (Table 2), and the number of people at each level generally shows an upward trend as the accuracy rate increased. However, nearly 2/5 of the personal classification accuracy rate is below 60%, which shows a large room for improvement.

Table 3 Cumulative frequency distribution of the personal waste classification accuracy.
Full size table

Descriptive statistical analysis of community attachment

The average value of the public community attachment is 3.08, which is relatively weak currently. With the improvement of living standards, the public pays more attention to constructing their communities and pursues a living environment that better at meets their needs. At the same time, more attention is paid to community construction and services, leading to more problems being exposed. Increasing barriers to communication among people in urban communities, and increasing mobility of residents have hindered the enhancement of individual community attachment. .

Descriptive statistical analysis of environmental concerns

The average value of public environmental concern is 4.26, indicating a high level of public environmental concern. With the accelerated development of China and the significant improvement of the public’s living standards, the public’s initiative to understand environmental issues is increasing. Coupled with the increasing efforts of environmental protection propaganda at different levels, angles and in different ways by the government and society, and the increasingly diversified channels for acquiring environmental protection knowledge and receiving environmental protection education.

Descriptive statistical analysis of institution-based trust

The average value of the public institution-based trust is 3.54, which is above the average. It cannot be separated from the government’s long-term efforts to reach out to the public, actively explore and implement policies that benefit the people to ensure their effective implementation. The public gains from implementing these policies and is thus more willing to believe that the government’s policies are aimed at enhancing people’s well-being, which results in higher institutional trust. However, there is clearly still room for improvement.

Analysis of factors influencing the personal waste classification accuracy

The explained variables of this study are multivariate ordered discrete variables, so the multivariate ordered Logistic model is applied to analyze the influencing factors of the personal waste classification accuracy, and hierarchical regression is carried out through model 1 to 5: Model 1 is the main effect analysis of the waste classification accuracy on community attachment; Model 2 integrates environmental concern into Model 1 for main effects analysis; Model 3 integrates institution-based trust into Model 2 for main effect analysis; Model 4 introduces the interaction between environmental concern and institution-based trust into Model 3 to examine the moderating effect of institution-based trust on the relationship between environmental concern and the personal waste classification accuracy; Model 5 introduces the interaction between institution-based trust and community attachment into Model 3 to examine the moderating effect of community attachment on the relationship between institution-based trust and the personal waste classification accuracy. It should be noted that the independent and moderating variables were zero-centered in examining the moderating effects of institutional trust and community attachment. The regression analysis of the survey data was performed using Stata MP Version 16 statistical software (Table 3), and the likelihood ratio chi-square values of the five models all reached a significance level of 1%.

Table 4 Model estimation results of the influencing factors of personal waste classification accuracy.
Full size table

Analysis of estimated results

According to the estimates results of the four models in Table 4, this paper analyzes the factors influencing the accuracy of personal waste classification in six aspects: community attachment, environmental concern, institution-based trust, the moderating effects of institution-based trust and community attachment, and control variables.

The effect of environmental concern on personal waste classification accuracy is not significant. According to the “Attitude-Situation-Behavior” theory, situational factors such as sorting facilities, sanitary conditions at the drop-off point, and technical support for responsibility tracing in real-life scenarios can result in waste sorting and its accuracy rate. In addition, some studies have found inconsistencies between environmental concern and environmentally friendly behavior41, so it is not a coincidence that environmental concern does not significantly affect environmentally friendly behavior. The insignificant effect of environmental concern on waste classification accuracy has not been verified by H1.

Institution-based trust has a significant positive impact on waste classification accuracy. In Model 3, institution-based trust is incorporated into Model 2 to analyze its effect on waste classification accuracy. The result shows that institution-based trust significantly impacts the accuracy of personal waste classification, that is, the higher the public’s institution-based trust in the government, the higher their waste classification accuracy. The uncertain risk faced by individuals in the social environment can be reduced under the control of the institution, and the expected income of individuals is more optimistic under the design of the new system of waste classification, which makes the public inclined to actively cooperate with the policy implementation. In addition, institution-based trust can form a “soft constraint” to discipline and shape social order42, guide the public to understand the relationship between humans and nature correctly, and promote the public to implement environmentally friendly behavior43,44. Therefore, the higher the institution-based trust, the more pronounced the disciplining effect of this “soft constraint” is, and the higher the waste classification accuracy. That is, institution-based trust can significantly and positively affect the accuracy of personal waste classification, which verifies the H2.

Community attachment has a significant positive impact on the accuracy of personal waste classification. The model estimation results show that community attachment significantly and positively affects the waste classification accuracy within a 5% statistical level in all five models, indicating that as the community attachment increases, the higher the likelihood of higher accuracy of personal waste classification. With the improvement of people’s living standards, people pay more attention to the governance and services in the community. The public community attachment leads to their willingness to participate in community activities or abide by community regulations. The stronger the public community attachment is, the more actively they participate in waste classification, the more serious they are about waste classification, so as to accumulate more knowledge and experience of waste classification, and the higher the accuracy rate of waste classification. That is, community attachment can positively affect the accuracy of personal waste classification, which verifies the H3.

The moderating effect of institution-based trust on the relationship between environmental concern-personal waste classification accuracy is not apparent. In Model 4, the interaction between environmental concern and institution-based trust are included to examine the moderating effect of institution-based trust on the relationship between “environmental concern—waste classification accuracy”. The interaction term is insignificant in the model, indicating that institution-based trust has no significant moderating effect on “environmental concern-personal waste classification accuracy”. The H4 has not been verified.

The moderating effect of community attachment on “institution-based trust—waste classification accuracy” is remarkable. In Model 5, the interaction term between community attachment and institution-based trust was included in Model 3 to investigate and examine the moderating effect of community attachment on the relationship between “institution-based trust—waste classification accuracy”. The interaction terms are prominent in the model, which shows that community attachment significantly moderates the relationship between institution-based trust and personal waste classification accuracy. The relationship between institution-based trust and waste classification accuracy regulated by community attachment is shown in the Fig. 2, as the impact on the slope of the line. Low community attachment makes institution-based trust compromise waste classification accuracy, while strong community attachment makes institution-based trust positively impact waste classification accuracy. It can be seen that the influence of community attachment on the relationship between “institution-based trust—waste classification accuracy” is positive. With the improvement of community attachment, the influence of institution-based trust on waste classification accuracy changes from negative to positive, and the influence degree shows a trend from large to small. After the direction of influence changes to positive, the influence degree changes to a trend from small to large. The H5 is verified.

Fig. 2
figure 2

Interactive relationship between institution-based trust and the accuracy rate of personal waste classification.

Full size image

Age and annual household income have significant positive effects on waste classification accuracy. Age has a significant positive impact on the accuracy of personal waste classification at the 1% level in all models, indicating that the older the age, the greater the probability that the accuracy of garbage classification falls in the higher range. Retirees have superior conditions in time and space and the need for spiritual self-realization, so they involve more in waste classification and thus have a higher waste classification accuracy rate. Annual household income positively impacts the accuracy of personal waste classification at a significant level of 5% in all models, indicating that the higher the annual household income, the greater the probability of higher waste classification accuracy. Annual household income represents the individual’s social class, and society tends to have greater expectations and requirements for people with higher social classes. According to Maslow’s demand theory, people with higher income pursue a better ecological environment but also have higher social recognition and self-realization needs. In addition, the field survey finds that the high-grade residential waste classification management is more stringent, providing a more active external scenario of classification compared to the low-grade residential areas. Gender, education and occupation do not significantly influence the accuracy of personal waste classification in all models.

Conclusion and policy enlightenment

Conclusion

This study uses the multivariate ordered Logistic model to analyze the data of 418 respondents in Hangzhou. The main effect of community attachment, environmental concern and institution-based trust on waste classification accuracy, the moderating effect of institution-based trust on the relationship between “environmental concern—waste classification accuracy”, and the moderating effect of community attachment on the relationship between “institution-based trust—waste classification accuracy” are investigated. The main conclusions are as follows:

  1. 1.

    The accuracy of waste classification is generally acceptable, but there is still room for improvement. In the survey samples, the average accuracy of waste classification is 62.49%, indicating that implementing of the current waste classification policy is satisfactory. However, the accuracy of waste classification still has a large room for improvement and requires further development.

  2. 2.

    Community attachment has a significant positive impact on the accuracy of personal waste classification. The average value of the public community attachment is 3.08, which is relatively general. The regression analysis results show that community attachment significantly contributed to the accuracy of waste classification, the stronger the community attachment is, the higher the accuracy of waste classification will be. Community attachment has a significant positive moderating effect on “institution-based trust—waste classification accuracy”. Community attachment indirectly affects the accuracy of waste classification by regulating the influence of institution-based trust on the accuracy of waste classification. Therefore, enhancing community attachment can be an essential means to improve the accuracy of waste classification.

  3. 3.

    Institution-based trust plays a significant role in improving the accuracy of waste classification. The mean value of institution-based trust is 3.54, and the public has a high degree of trust in the current government policy of waste classification and disposal. Regression analysis shows that institution-based trust has a significant positive effect on waste classification accuracy, which means the higher the degree of institution-based trust, the higher the classification accuracy. Therefore, creating an excellent institutional environment to obtain public trust can effectively improve the accuracy rate of waste classification.

  4. 4.

    Among the control variables, age and annual household income have a significant positive impact on the accuracy of public waste classification. People aged 46 to 69 have a higher rate of waste classification accuracy because they have sufficient time and the subjective will of self-realization. In terms of annual household income, the value pursuit of high-income people and the scene provided by the community are more likely to stimulate their active waste classification behavior.

Policy implications

  1. 1.

    Based on the community, the government should strengthen the essential organizational function of waste classification, strengthen community governance and service functions, provide a good community environment and public services, and increase public satisfaction with the community. Emphasis on the construction of community culture and the organization of activities, promote good communication among the community public, cultivate community consciousness and enhance the sense of community attachment. Build community media infrastructure, make full use of communication carriers of community organizations, disseminate practical life information and community planning governance information, cultivate and guide public awareness and behavior to participate in community affairs actively. Establish an organizational communication mechanism for public opinion expression and enhance public autonomy and independent management of community affairs. Furthermore, external social supervision and expectations will regulate the behavior of the public virtually and improve the accuracy of waste classification.

  2. 2.

    Reasonable policy formulation and effective implementation should be ensured, and the foundation of trust in the waste classification system should be solidified. During the policy formulation and implementation, people’s livelihood and public sentiment should be emphasized, public’s acceptance of garbage sorting needs to be evaluated, and public’s interest demands and concerns should be timely responded. The policy implementation should be accurate and in place. At the same time, given the public institution-based trust as a whole, grassroots work other than waste classification and disposal should also receive equal attention. The study found that people such as retirees have a higher waste classification accuracy, so the communities or superior government should give subsidies to stimulate their initiative and creativity, and actively mobilize them to participate in community waste classification promotion, persuasion and management.

  3. 3.

    The system of charging for waste classification should be constructed, and the external situational stimuli for waste classification should be created. The public tends to rely on the government due to the institution-based trust, and individual environmental awareness behavior is weakly driven. Therefore, the government should gradually impose waste classification management fees and transfer the classification cost from the government or the cleaner to the public. In addition, the government should provide an effective external scenarios of classification. For example, upgrading waste cans and other classification facilities to reduce the cost of public sorting and focusing on the waste drop point and the surrounding environment to keep the sorting point clean and tidy. Although the effect of environmental relations on waste classification accuracy is not significant, initiatives related to enhancing environmental concern are generally also one of the important elements of extrinsic scenario creation and waste sorting policy implementation, which is useful for enhancing both institutional trust and high community belonging and deserves attention.