Exploring the Impact of Socioeconomic Status on Farmers’ Participation in Rural Living Environmental Governance Behavior—Evidence from Jiangsu Province, China

Abstract

The participation of farmers in environmental governance is not simply the behavior choice of individuals but is also influenced by their socioeconomic status (SES). Using survey data from 2088 peasant households in Jiangsu Province, China, this study used the ordered probit model to identify the differential impacts of objective socioeconomic status (OSES) and subjective socioeconomic status (SSES) on farmers’ participation in rural living environmental governance behavior (FPLEB). The mediation effect model was also used to explore the mediating role of environmental cognition in the relationship between SES and FPLEB. The results reveal that (1) SES (i.e., OSES and SSES) is a positive factor affecting farmers’ involvement in improving their living environment, with SSES demonstrating a stronger correlation with efforts to enhance rural living conditions than OSES, (2) environmental knowledge and pollution cognition serve as mediating factors, bridging the effect of OSES on the level of participation in environmental governance, and (3) SES (i.e., OSES and SSES) is more effective in fostering future-oriented environmental governance behaviors among farmers than those with a present-oriented mindset.

1. Introduction

In the past, people have paid more attention to urban environments and neglected the rural areas that account for 42.7% of the global population [1]. Data show that approximately 50% of rural populations in developing countries have no access to improved sanitation facilities, leading to health problems and increased pollution [2]. As a result, environmental governance has become a pressing problem in most developing economies around the world. As the largest developing country, China has a large population in rural areas, which is most closely related to the natural ecology [3]. In 2023, the availability of sanitary toilets in rural areas of China was less than 80%, and the centralized treatment rate of rural domestic sewage was about 40% [4]. Rural living environmental governance is a long-term and complex social governance work that needs to establish the multi-party co-construction and co-management pattern involving the government, external enterprises, villagers, and village collectives [5]. Among them, as the major beneficiaries of environmental governance, rural residents have the endogenous power to participate in environmental remediation. Therefore, the governance of the rural living environment should reflect the concept of “neutral education” and exert the principal role of rural residents in environmental governance [6].

Several studies have shown that perceived value [7], environmental cognition [8], and social capital [9] are important factors that influence the environmental behavior of rural residents. As differences in SES and wealth inequality grow, it becomes necessary to understand environmental behavior from the perspective of social class [10,11]. However, existing studies have not agreed on the pathways and effects of SES affecting individual environmental behavior.

Althoff and Greig (1977) believed that people in higher social classes would be more proactive in practicing environmental protection [12]; for example, they would be more likely to engage in green consumption [13], energy-saving efforts [14], and waste recycling [15]. However, Buttel and Flinn (1978) argued the opposite. They claimed that the lower and working classes usually live in heavily polluted areas with dirty working conditions and poor recreational facilities. Thus, they may be more concerned about poor environmental conditions than the middle and upper classes [16]. Eom et al. (2018) support this view by demonstrating that lower-class individuals tended to be more supportive of pro-environmental policies than upper-class individuals [17]. However, Chen et al. (2023) argued that the correlation between socioeconomic status (SES) and individual environmental behavior is not strictly linear but rather “U-shaped” [18].

Furthermore, objective socioeconomic status (OSES) and subjective socioeconomic status (SSES) may have differing effects on individual environmental behavior. The distinction of SES is formed by the differences between objective material resources and subjective perceptions of social status [19]. OSES is correlated with SSES but has independent effects on psychological and behavioral outcomes [20]. Generally speaking, SSES is a stronger predictor than OSES [21]. This may be because it is difficult for OSES indicators to fully depict social stratification, leading to biased research conclusions on SES. In contrast, Grandin et al. (2022) and Yan et al. (2021) recently found that OSES explains more than SSES [11,22]. In conclusion, the disparities between OSES and SSES should be strictly distinguished when studying the relationship between SES and environmental behavior.

Another important point is that few studies have investigated the link between OSES and FPLEB mediated by environmental cognition. Environmental cognition is the basis for implementing farmers’ environmental governance behaviors [23,24]. However, environmental cognition is influenced by a combination of sociocultural and economic contexts, and SES may be the psychological antecedent of individual differences in support of environmental action [17,25]. Therefore, we considered environmental cognition to be a mediating variable from the perspectives of environmental knowledge and pollution perception. The mediating role of environmental cognition in OSES and FPLEB was examined.

The existing research primarily focuses on whether SES is associated with environmental behavior, lacking comparative studies on the impact of OSES and SSES on FPLEB and rarely paying attention to the possible intermediary role of environmental cognition between OSES and FPLEB. Based on this finding, this study assembled 2101 farmers in Jiangsu Province of China as the study focus. Multivariate probit regression and bootstrap-mediated effects models explored the logical relationship between SES (objective and subjective) and residential environmental governance behaviors.

This paper has the following contributions to existing research areas: (1) The mutual demonstration of OSES and SSES is conducive to improving the reliability and consistency of the conclusion on the relationship between SES and FPLEB; (2) the importance of SSES to FPLEB is emphasized to provide a reference for policymakers to introduce more targeted environmental governance policies; and (3) this paper elucidates the mediating role of environmental cognition, provides a new way to understand the relationship between OSES and FPLEB, and expands the cognitive mechanism in related research fields.

2. Theoretical Analysis and Research Hypotheses

2.1. Objective Socioeconomic Status

OSES reflects individuals’ differences in the acquisition and control of resources. High OSES means being at the center of the social network [26], having sufficient resources [27] and engaging in superior power relations [28]. As a result, they face weaker information barriers [29], allowing farmers to gain more pro-environmental knowledge to implement pro-environmental behaviors. At the same time, having more wealth makes it easier for individuals to consume fewer resources in implementing sustainable behaviors [30]. As a result, they are more likely to invest time, energy, and resources to participate in specific actions, such as garbage classification, energy conservation, emission reduction, and sewage treatment. Moreover, they have a greater need for a high quality of life and a livable environment, leading to stronger environmental protection behaviors [31].

In contrast, individuals with lower OSES represent chronic resource scarcity, leading to pro-environmental behavior requiring higher risks and associated costs in the face of uncertainty [32]. They are more likely to have narrow-minded values and short-sighted behavior for goal achievement [33], resulting in a higher tolerance for environmental pollution. Therefore, their potential to implement environmental behaviors is further reduced. Accordingly, we propose the following hypothesis:

H1. 

OSES is positively associated with FPLEB.

2.2. Subjective Socioeconomic Status of the Farmers

SSES reflects people’s subjective assessment of their relative societal position [34]. The existing research has found that the higher the SSES of residents, the more environmental behaviors they implement. For one thing, status motives lead people to participate in environmental governance [35]. To maintain their current status and image, farmers with high SSES will actively participate in environmental governance. In addition, farmers with a higher SSES have better literacy and positive and stable socio-political attitudes; so, their positive environmental protection concepts in the political and social fields will naturally correspond to daily life, thus implementing more environmental behaviors [12]. Accordingly, we propose the following hypothesis:

H2. 

SSES is positively associated with FPLEB.

2.3. Mediating Effect of Environmental Cognition

Environmental cognition means an evaluation of the environment and an understanding of the relationship between the individual and the environment [36], which is mainly characterized by environmental knowledge and pollution perception [37]. Among them, environmental knowledge refers to the public’s cognition and understanding of environmental governance and government environmental protection policies [38], which is a key factor in environmental protection behavior [39]. Enhanced knowledge of human settlement environmental governance can better transform farmers’ “will” into “action.” Pollution perception refers to farmers’ perception of the threat of unfriendly environmental behavior and the degree of pollution of the surrounding environment, which is an important manifestation of farmers’ environmental awareness. The more farmers perceive that the random disposal and lack of sorting of household waste pose a serious threat to their healthy lives, the higher the potential for protecting the environment.

The environmental cognition level of individuals is associated with their SES [16]. The differentiation of SES results in the differentiation of farmers’ environmental cognition, which ultimately affects the degree of response of farmers’ environmentally friendly behaviors. At the level of environmental knowledge, the higher the farmers’ SES is, the easier it is for them to control and access key information and scarce resources by using their network [40]. This, in turn, helps to enhance farmers’ factual knowledge of environmental problems and strategic knowledge of environmental problem-solving and ultimately promotes changes in farmers’ pro-environmental behaviors and outcomes. At the level of pollution perception, the educational gap caused by the difference between the rich and the poor makes individuals with high SES more aware of the ecological hazards of non-friendly environmental behaviors [41], and they quickly analyze the pollution problems in the living area according to the existing knowledge value system. Therefore, when farmers realize the severity of environmental pollution in their villages, it is likely to stimulate their sense of ownership of environmental protection [36], which enhances their motivation to implement environmental behavior. Based on the above discussion, we propose the following hypotheses:

H3. 

Environmental knowledge plays a mediating role between OSES and FPLEB.

H4. 

Pollution perception plays a mediating role between OSES and FPLEB.

Figure 1 shows the research framework of the influence of SES on FPLEB.

3. Data and Methodology

3.1. Data

The data used in this study are from the China Land Economic Survey (CLES), which covers multiple individual, household, and village levels. In 2020, Nanjing Agricultural University adopted the Probability Proportionate to Size Sampling method to investigate the sample cities, counties, districts, and administrative villages in Jiangsu Province and conducted a tracking survey in 2021. Finally, 2600 farmers from 52 villages in 26 counties of 13 prefecture-level cities were selected. Given that the dimension data on rural residential environmental governance were launched in 2021, we selected the 2021 household data for the research. Simultaneously, to acquire the necessary village data for this study, a village questionnaire was matched with a peasant household questionnaire. After removing the samples lacking key information, 2088 valid samples were derived. Among them, men accounted for 92.48% and women accounted for 7.52%. The age range was 18–92 years (M = 63). The average household agricultural labor force was 2 persons, accounting for 44.41%.

3.2. Methodology

3.2.1. Probit Model

This study focused on the dependent variable of farmers’ participation in rural living environmental governance behavior (FPLEB), specifically measured through farmers’ participation in garbage classification, sewage discharge, and the number of sanitary latrines, reflected in the values of 0, 1, 2, and 3, respectively. This study applied the research practices of Shen et al. (2024) [42] and chose the ordered probit model for verification. The estimation formula is as follows:

$$F{PLEB}_i^{*}={\alpha }_0+{\alpha }_1{SES}_i+\beta control+{\epsilon }_i$$

(1)

where ${SES}_i$ is the farmer’s SES and $\mathrm{c}\mathrm{o}\mathrm{n}\mathrm{t}\mathrm{r}\mathrm{o}\mathrm{l}$ represents a series of control variables. The ordered probit model is expressed as follows:

$${FPLEB}_i=\left\{\begin{array}{c}0\left(non-participation\right), if {FPLEB}_i^{\mathrm{*}}\le {\gamma }_1\\ 1\left(participationin1type\right), if {\gamma }_1<{FPLEB}_i^{\mathrm{*}}\le {\gamma }_2\\ 2\left(participationin2types\right), if {\gamma }_2<{FPLEB}_i^{\mathrm{*}}\le {\gamma }_3\\ 3\left(participationin3types\right), if {\gamma }_3\le {FPLEB}_i^{\mathrm{*}}\end{array}\right.$$

(2)

where ${\gamma }_1{, \gamma }_2, \mathrm{a}\mathrm{n}\mathrm{d}{\gamma }_3$ are the unknown dividing points. The value of ${FPLEB}_i$ depends on the relationship between the latent ${FPLEB}_i$ variable and the cut-off point, and the probabilities of farmers’ non-participation, participation in one type, participation in two types, and participation in three types of rural living environmental governance behavior were obtained as follows:

$$P({FPLEB}_i=0|{SES}_i)=P({FPLEB}_i\le {\gamma }_1\left|{SES}_i\right)=\phi ({\gamma }_1-{\alpha }_1{SES}_i-\beta control)$$

(3)

$$P({FPLEB}_i=1|{SES}_i)=P({\gamma }_1<{FPLEB}_i\le {\gamma }_2\left|{SES}_i\right)=\phi ({\gamma }_2-{\alpha }_1{SES}_i-\beta control)-\phi ({\gamma }_1-{\alpha }_1{SES}_i-\beta control)$$

(4)

$$P({FPLEB}_i=3|{SES}_i)=P({\gamma }_3\le {FPLEB}_i\left|{SES}_i\right)=1-\phi ({\gamma }_3-{\alpha }_1{SES}_i-\beta control)$$

(5)

3.2.2. Mediation Effect Model

This study used the bootstrap method of Preacher et al. (2007) [43] to detect the mediating effects of environmental cognition (EN) between OSES and FPLEB. The bootstrap method uses the sample data to back the sampling process through the original repeated sampling to calculate the statistics and estimate the sample distribution, the overall repeat sampling 1000 times with a confidence interval of 95%.

$${FPLEB}_i={\alpha }_0+{\alpha }_{1}O{SES}_i+{\epsilon }_1$$

(6)

$$EN={\beta }_0+{\beta }_1{OSES}_i+{\epsilon }_2$$

(7)

$$FPLEB={\delta }_0+{\delta }_1{OSES}_i+{\delta }_2{EN+\epsilon }_3$$

(8)

3.3. Measures

3.3.1. Farmers’ Participation in Rural Living Environmental Governance Behavior (FPLEB)

According to the key tasks proposed in the Three-Year Action Plan for Rural Living Environment Improvement issued by the State Council and the existing research content, behavior to participate in the “toilet revolution”, “garbage management”, and “sewage treatment” were used to define FPLEB. The participation value was 1; otherwise, the value was 0. To reflect the level of participation, the quantitative method proposed by Shen et al. (2024) [42] was used as a reference, and the total number of FPLEB was used to represent their participation level.

3.3.2. Socioeconomic Status (SES)

Objective socioeconomic status (OSES): Combining the measurement concept of Qu (2020) and Liu (2023) [44,45], OSES was assessed using cadres’ status, level of education, relationship network, annual household income, financial assets, and housing assets. The entropy weight method was used to measure the weights of 6 variables, and then, the entropy value of each index was calculated and the comprehensive score of farmers’ socioeconomic status was obtained. Finally, the comprehensive score of socio-economic status is sorted by village, and the quartile method proposed by Shao et al. (2022) [46] is adopted to divide farmers into low class, middle class, middle class, upper middle class and upper class, with the values increasing from 1 to 5 to distinguish the coreness of each sample in different living spaces. The results are shown in

Table 1. OSES Evaluation Index System.

Primary Indicators	Secondary Indicators	Information Entropy e	Utility Value d	Weight Coefficient w
OSES	Cadres’ status	0.756	0.243	0.526
	Education level	0.977	0.023	0.049
	Relationship network	0.986	0.014	0.030
	Household income	0.996	0.012	0.011
	Financial assets	0.986	0.005	0.031
	Housing assets	0.837	0.163	0.353

.

Subjective socioeconomic status (SSES): SSES was measured using Question 9 in Part K of the CLES questionnaire, that is, the self-rating status of individuals in the village, as a metric. The option was assigned as follows: 1 = the lowest level and 5 = the highest level; the higher the score, the higher the rank and the higher the SES of the individual.

3.3.3. Environmental Cognition

We relied primarily on the research of Tang et al. (2018) to measure environmental knowledge and pollution perception [37]. Environmental knowledge was measured by “the degree to understand the rural living environmental governance.” Pollution perception was measured using “the evaluation of the impact of living waste being randomly dumped and not sorted on the community environment (village appearance and living order)” and “the evaluation of the impact of living waste being randomly dumped and not sorted on the rural ecological environment (deterioration of water quality and pollution of soil, etc.).” The indicators were summed and divided by two to obtain the pollution perception value. All answers to the questions used the Likert five-point scale, with the degree of intensity ranked from small (1) to large (5).

3.3.4. Controlled Variables

Additionally, many other factors affect FPLEB. With reference to relevant research, we selected the characteristics of the head of the household, family characteristics, and external characteristics for testing. Among them, the characteristics of the head of the household mainly included gender, age, and health status. Household characteristics included the number of agricultural laborers in the household and whether it was an ethnic minority household. External characteristics included the village’s topography, whether it is located on the city’s outskirts, the distance between the village committee and the county seat, social norms, and social pressures (

Table 2. Descriptive Analysis of Variables.

Variables	Definition	Evaluation	Mean	S.D.
FPLEB	Number of participants in three types of environmental governance behaviors, including sewage treatment, garbage sorting, and the use of flushable toilets	No participation = 0; 1 type of participation = 1; 2 types of participation = 2; and 3 types of participation = 3	2.223	0.818
SES	OSES	Lower level = 1; middle and lower = 2; middle = 3; middle and upper = 4; and upper = 5	2.966	1.414
	SSES	1 = very low; 2 = relatively low; 3 = general; 4 = relatively high; and 5 = very high	2.941	0.700
Environmental knowledge	Knowledge of rural habitat improvement	1 = have not heard of it; 2 = have heard of it but do not know much about it; 3 = know a little about it; 4 = Know a lot; and 5 = Very well understood	2.739	1.330
Pollution perception	Impacts on rural ecosystems (deterioration of water quality, contamination of soil, etc.) caused by indiscriminate dumping and non-separation of household waste	1 = very small; 2 = relatively small; 3 = average; 4 = relatively large; and 5 = very large	4.168	0.796
	Impacts on the community environment (village appearance and order of life in the village) caused by the haphazard accumulation and non-separation of household garbage	1 = very small; 2 = relatively small; 3 = average; 4 = relatively large; and 5 = very large	4.171	0.807
Gender	Gender	Male = 1 and female = 0	0.925	0.263
Age	Age	Year	63.207	10.311
Health	Self-identified health status	1 = incapacitated; 2 = poor; 3 = medium; 4 = good; and 5 = excellent	4.001	1.097
Minority households	Whether it is a minority household	1 = yes and 0 = no	0.030	0.171
Agricultural laborers	Several members of the family worked in agriculture	Number of people working in agriculture	1.418	0.997
Terrain	Topography of the village	1 = plains and 2 = hills	1.165	0.371
Location	The village is located in the city suburb	1 = yes and 0 = no	0.372	0.483
Distance	The distance between the village committee and the county seat	Kilometer	18.784	12.602
Social norm	Household waste sorting can be appreciated and praised	1 = Completely disagree; 2 = do not quite agree; 3 = general; 4 = agree more; and 5 = agree very much	4.152	0.857
Social pressure	Whether the government should publicize the classification of rural domestic waste	1 = yes and 0 = no	0.848	0.359

).

4. Results

4.1. Benchmark Regression Results

4.1.1. Oprobit Regression Result Analysis

Before proceeding with the model dissection, it was necessary to discuss whether there was a high degree of correlation between the explanatory variables. We needed to conduct a multicollinearity parametric test among the variables. The mean variance inflation factor (VIF) of each variable was 1.06, and the tolerance was greater than 0.1. Therefore, there was no obvious multicollinearity among the variables, and these data met the requirements of the regression analysis. Next, we used the Oprobit model to test the relationship between SES and FPLEB, and the results are given in

Table 3. Basic Regression Results of the Oprobit Model.

Variables	FPLEB
	(1)	(2)
OSES	0.091 ***
	(0.019)
SSES		0.188 ***
		(0.037)
Gender	0.041	0.065
	(0.095)	(0.094)
Age	−0.010 ***	−0.012 ***
	(0.003)	(0.003)
Health	0.131 ***	0.120 ***
	(0.024)	(0.024)
Minority households	−0.056	−0.048
	(0.146)	(0.146)
Agricultural laborers	−0.034	−0.043
	(0.026)	(0.026)
Terrain	0.574 ***	0.556 ***
	(0.077)	(0.077)
Location	0.374 ***	0.362 ***
	(0.054)	(0.054)
Distance	−0.002	−0.003
	(0.002)	(0.002)
Social norm	0.089 **	0.095 **
	(0.029)	(0.029)
Social pressure	0.754 ***	0.756 ***
	(0.068)	(0.068)
No. of respondents	2086	2086
Pseudo R2	0.097	0.097

Notes: ** p < 0.01, and *** p < 0.001.

.

According to Table 3, FPLEB was positively associated with OSES (β = 0.091, p < 0.001) and SSES (β = 0.188, p < 0.001), and the correlation with OSES was stronger than SSES. It demonstrated that improving OSES and SSES increases willingness to participate in environmental governance; however, SSES plays a greater role. Therefore, we concluded the following:

Hypotheses H1 and H2 are verified.

We found that a higher age led to less environmental governance behavior (p < 0.001). This may be because environmental behavior requires technology and knowledge, and older farmers have lower educational levels, limiting their ability to learn and apply environmental knowledge. In addition, health, terrain, location on a city’s outskirts, and social pressures positively influenced FPLEB at the 0.1% level.

4.1.2. Marginal Effect Analysis

Figure 2 shows the marginal impact of SES on FPLEB. The results show that the higher the OSES and SSES, the higher the probability of farmers participating in the three habitat environmental governance projects, and the test was passed at the significance level of 1%. Meanwhile, the marginal effect was SSES > OSES, which indirectly verifies the result of baseline regression.

4.1.3. Heterogeneity Analysis

The aforementioned empirical analysis confirmed that SES significantly promotes FPLEB and does not consider the differences in the impact of SES on the different time preferences on the farmers’ governance behaviors. Future-oriented individuals tend to exhibit more environmental attitudes and behaviors than present-oriented individuals (Grandin et al., 2022) [22]. In this study, time preference was divided into three categories: present-oriented, future-oriented, and both present- and future-oriented groups; we used group regression to further investigate the heterogeneity influence of SES on FPLEB.

As shown in

Table 4. Regression Results with Different Time Preferences.

Variables	Present-Oriented	Both Present- and Future-Oriented	Future-Oriented
OSES	0.049	0.090 **	0.166 **
	(0.028)	(0.029)	(0.054)
SSES	0.253 ***	0.046	0.263 *
	(0.058)	(0.060)	(0.104)
Control variables	Controlled	Controlled	Controlled
No. of respondents	879	898	264

Notes: * p < 0.05, ** p < 0.01, and *** p < 0.001.

, the coefficients of the impact of OSES on both present- and future-oriented and future-oriented preferences of FPLEB are 0.090 (p < 0.01) and 0.166 (p < 0.01), respectively. SSES has a positive impact on the present-oriented (β = 0.253, p < 0.001) and future-oriented preferences (β = 0.263, p < 0.05) of FPLEB. Overall, favoring future preferences will strengthen the facilitating effect of SES on FPLEB.

4.1.4. Robustness Test

The benchmark regression adopted the number of behaviors to define environmental governance behavior. However, there are uncontrollable factors that may lead to bias. To ensure reliable results, this study conducted robustness tests by changing the assignment of the explainaed variable and replacing the regression model. First, the explained variable “farmers’ participation in rural environmental governance” was treated as a binary variable (1 point was allocated if the farmer participated in at least one rural environmental governance project; otherwise, 0 points were allocated). Second, the causal relationship between SES and environmental behavior was retested using the OLS model. The results in

Table 5. Robustness Test Results.

Variables	OSES		SSES
	Replacement Variable	Replacement Model	Replacement Variable	Replacement Model
SES	0.145 ***	0.058 ***	0.167 *	0.126 ***
	(0.041)	(0.012)	(0.073)	(0.024)
Gender	−0.202	0.027	−0.131	0.039
	(0.215)	(0.062)	(0.211)	(0.061)
Age	−0.003	−0.005 **	−0.006	−0.007 ***
	(0.006)	(0.002)	(0.006)	(0.002)
Health	0.139 **	0.094 ***	0.136 **	0.086 ***
	(0.046)	(0.016)	(0.046)	(0.016)
Minority households	−0.135	−0.044	−0.129	−0.039
	(0.286)	(0.094)	(0.284)	(0.094)
Agricultural laborers	0.014	−0.017	0.005	−0.023
	(0.055)	(0.016)	(0.056)	(0.016)
Terrain	1.098 **	0.316 ***	1.057 **	0.305 ***
	(0.375)	(0.046)	(0.371)	(0.046)
Location	0.338 **	0.234 ***	0.316 *	0.227 ***
	(0.130)	(0.034)	(0.129)	(0.034)
Distance	0.005	−0.001	0.004	−0.002
	(0.004)	(0.001)	(0.004)	(0.001)
Social norm	0.113	0.063 **	0.120 *	0.067 ***
	(0.059)	(0.019)	(0.058)	(0.019)
Social pressure	0.361 **	0.537 ***	0.370 **	0.538 ***
	(0.122)	(0.046)	(0.122)	(0.046)
No. of respondents	2088	2086	2088	2086
R2		0.193		0.195
Pseudo R2	0.125		0.114

Notes: * p < 0.05, ** p < 0.01, and *** p < 0.001.

show that SES still positively affects FPLEB. This suggests that the results of the original regression model are robust and reliable.

4.2. Mediation Analysis

Table 6 shows the mediating role of environmental cognition in OSES and FPLEB. As the results show, the 95% confidence intervals for the indirect effects of environmental knowledge and pollution perception are [0.007, 0.018] and [0.000, 0.005], respectively, respectively, with significance at the at 0.1% and 5% level, indicating that the effects of OSES on FPLEB are partially mediated by environmental cognition (environmental knowledge and pollution perception). Therefore, we conclude the following:

Hypotheses H3 and H4 are partially confirmed.

Table 6. Intermediary Effect Regression Results.

Variables	Route	Effect Values	Boot Standard Error	95% Confidence Interval (BootCI) Lower Bound
				BootCI Lower	BootCI Upper
OSES	Direct effect	0.046 ***	0.013	0.021	0.070
Environmental knowledge	Indirect effect	0.013 ***	0.003	0.007	0.018
OSES	Direct effect	0.056 ***	0.013	0.030	0.081
Pollution cognition	Indirect effect	0.003 *	0.001	0.000	0.005

Notes: * p < 0.05 and *** p < 0.001.

Table 6. Intermediary Effect Regression Results.

Variables	Route	Effect Values	Boot Standard Error	95% Confidence Interval (BootCI) Lower Bound
				BootCI Lower	BootCI Upper
OSES	Direct effect	0.046 ***	0.013	0.021	0.070
Environmental knowledge	Indirect effect	0.013 ***	0.003	0.007	0.018
OSES	Direct effect	0.056 ***	0.013	0.030	0.081
Pollution cognition	Indirect effect	0.003 *	0.001	0.000	0.005

Notes: * p < 0.05 and *** p < 0.001.

5. Discussion

SES reflects the differences in social and economic resources owned by farmers and will affect the environmental cognition level and environmental goals of farmers with different social and economic statuses, leading to different behavior structures in environmental governance. This study investigated the relationship between SES (OSES and SSES) and FPLEB. It examined the mediating role of environmental cognition between OSES and FPLEB. The following conclusions were obtained:

First, SSES has a stronger impact on FPLEB than OSES. This view is consistent with Cohen et al. (2008) [47] that SSES was more predictive of FPLEB than OSES. A plausible explanation is that SSES is an individual’s subjective perception and evaluation of their own socioeconomic status, and this subjective perception may more directly affect the individual’s attitude and behavior. This implies that it is not enough to simply rely on social means such as increasing educational inputs [39] and enhancing social capital [15] and economic incentives [48] to enhance the motivation of FPLEB; enhancing the class identity of peasants is a key approach to promoting FPLEB.

Second, future-oriented farmers are more willing to participate in the environmental governance of the rural living environment when their SES improves. This is similar to previous research demonstrating that future preference is the main driving force to promote an individual’s behavior [49]. This may be because the management of rural living environmental governance is a typical intertemporal investment with high costs, and the governance benefits are mainly reflected in the future. Therefore, future-oriented farmers exhibit more environmental behavior [50]. While this provides valuable insights into improving FPLEB, few studies have provided empirical evidence on the correlation between farmers’ time preferences and environmental protection behavior. Therefore, future research should continue to explore differences in time preferences in favor of environmentalism.

Third, environmental cognition is an active mediator in the pathways through which SES affects FPLEB. This suggests that effective environmental policy information and knowledge as well as awareness of environmental problems are more likely to be transmitted among households with a higher OSES. A higher OSES means having better resources, social networks, and communication opportunities [51]. This is crucial in helping farmers acquire environmental knowledge and accurately assess environmental problems.

Our findings are different from those of Song et al. (2020) [52]. They argued that people with marginalized identities and lower incomes tend to perceive a higher risk of climate change. In other words, the results show that higher SES is associated with higher environmental cognition among farm households, indicating that farmers are more motivated to participate in environmental governance. Future attention should be paid to SES mobility and increased environmental cognition. It may be that there are great differences in economic development levels, environmental policies, and cultural traditions in different regions, which lead to differences in farmers’ environmental cognition level and environmental behavior.

6. Conclusions and Policy Implications

The quality of the rural living environment affects the health and life quality of rural residents. It is necessary to find incentives to promote FPLEB. In this study, we tried to compare the effects of OSES and SSES on FPLEB, and we considered environmental cognition to be a mediating variable of OSES affecting FPLEB from the perspectives of environmental knowledge and pollution perception. As expected, both OSES and SSES positively affected FPLEB, but the effect of SSES was greater. In addition, OSES also affects FPLEB by influencing their environmental cognition. More specifically, higher SES can promote higher levels of environmental knowledge and pollution perception among farmers, which makes them more actively involved in environmental governance.

Our research has important policy implications for environmental protection in rural China. Policy design to promote farmers’ pro-environmental behavior should focus on improving farmers’ SES. For farmers with low SES, education and ideological guidance can help change the negative SSES cognition. At the same time, the OSES level should be improved by promoting employment, organizing regular exchange activities, popularizing education, increasing government subsidies, etc., so as to improve farmers’ participation in living environment governance. For farmers with higher SES, strengthening supervision and management, setting up environmental demonstration families, and establishing incentive mechanisms are the means to encourage them to participate in housing environmental governance. In addition, we recommend promoting knowledge of environmental protection and sustainable development and helping farmers build awareness of the long-term benefits of environmental governance so as to increase the enthusiasm of farmers to participate in environmental governance.

While our article is good, there are several limitations. First, the scope of the sample data may be too small, and it is uncertain whether the results can be extended to other regions. Because there are obvious differences in individual perceptions and pursuits of status in different regions [18], their behavior patterns related to environmental protection will vary [48]. Because this study only used Chinese farmers as the research subject, the survey results may not accurately reflect the causal relationship between SES and FPLEB. Future studies should use more national datasets to enhance the reliability of the conclusions.

Second, although SSES is a better predictor of FPLEB than OSES, this study did not observe how SSES promotes FPLEB. Subsequent research could consider whether psychological factors, such as personal expectations, environmental satisfaction, and well-being, lead to differences in impact effects.

Third, the data on individual characteristics were collected primarily from the heads of households, and in the context of China, it is common for the heads of households to be predominantly male, which may limit the generality of the findings.