Research on the impact of heterogeneous environmental regulations on smallholders’ livelihood strategies from the perspective of environmental identity theory

Abstract

Based on environmental identity theory, the multiple mediation model was used to reveal the effects of heterogeneous environmental regulation on smallholder farmers’ livelihood strategies and behaviors, specifically examining the multiple mediating effects of two different dimensions of environmental identity and percived self-efficacy. This paper utilizes smallholder data from four provinces, including Heilongjiang, Jiangxi, Sichuan and Hunan, and uses the macro PROCESS 3.5 developed by Hayes for data analysis and processing. The results show that (1) informal environmental regulation has a significant positive effect on smallholder livelihood strategies, while the effect of formal environmental regulation is not significant. (2) Environmental identity strength fully mediated both between heterogeneous environmental regulation and livelihood strategies; environmental identity salience fully mediated the effect of formal environmental regulation and partially mediated the effect of informal environmental regulation. (3) Self-efficacy fully mediated the effect of formal environmental regulation on smallholder livelihood strategies, while partially mediating the role of informal environmental regulation. (4) Environmental identity strength, salience, and self-efficacy had parallel and chained multiple mediating effects and the parallel mediating path were more important in the relationship between heterogeneous environmental regulations and smallholder livelihood strategies. By clarifying the relationship between heterogeneous environmental regulation and smallholder livelihood strategies and the mediating mechanism of this relationship, the study provides theoretical references and empirical evidence for research on the construction of environmental governance systems and the sustainable development of smallholder livelihoods, enriches the Chinese micro-case studies on livelihood strategies, and expands the understanding of the relationship between environmental regulation and livelihood strategies.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-52328-7.

Introduction

Livelihood strategy refers to the rational combination of livelihood activities chosen by farmers based on their capital to optimize well-being¹. Under the pressure of heterogeneous environmental regulations, smallholders adopt different adaptation pathways due to differences in resource endowments and capabilities². However, differences in smallholders’ value identification³ of different environmental regulatory instruments (mandatory and non-mandatory) and trade-offs⁴ between the costs of compliance and the goals of regulation contribute to the acceptability⁵ and bindingness⁶ are different, which will ultimately lead smallholder farmers to follow different paths of livelihood adaptation and development with unsatisfactory results⁷. Therefore, how to stimulate the value identity of smallholders by optimizing the combination of regulatory tools, so that smallholders’ livelihood strategies resonate with environmental regulation policies and thus lead to consistent actions, becomes a realistic and significant problem for policy makers and smallholders. Scholars’ studies on the relationship between environmental regulation and farmers’ livelihood strategies are mostly based on the sustainable development framework^8,9. However, analytical frameworks risk inherent bias when studies disregard non-economic drivers that fundamentally shape farmers’ livelihood decision-making¹⁰.

Building upon extant scholarship, this study employs a multi-mediation model grounded in environmental identity theory to systematically unravel how heterogeneous environmental governance instruments shape smallholders’ livelihood strategies. Through empirical validation of multiple mediating effects of environmental identity and self-efficacy, the study also provides theoretical as well as practical references in optimizing the combination of environmental regulation tools to satisfy the needs of small farmers’ sustainable development, and more importantly to unveil the Black Box of environmental regulation in social psychology.

Literature review

Empirical studies reveal distinct operational logics between formal environmental regulations and informal institutions, with divergent policy instruments and enforcement mechanisms generating heterogeneous compliance patterns among agricultural stakeholders¹¹. This study adopts the classification framework proposed by Xia et al.⁶, and Wang et al.⁵, to categorize environmental regulation into two distinct types: formal and informal mechanisms. Specifically, formal environmental regulation operates through government-mandated institutional frameworks that employ legal statutes, policy interventions, administrative sanctions, and economic incentives to enforce compliance with environmental standards. In contrast, informal environmental regulation emerges from self-governing social systems, manifesting through farmers’ environmental consciousness, community-established norms, and traditional ecological practices that spontaneously guide agricultural activities through non-coercive social governance. Overall, factors driving livelihood strategy choice fall into endogenous and exogenous categories¹². Endogenous factors mainly derive from livelihood capital, while exogenous factors reflect the push–pull forces of organizational structures and government policies¹³. When ecological conditions, policies, or institutions change, farmers typically adjust their livelihood strategies to sustain or enhance their livelihoods and re-adapt to the ecological environment¹⁴.Accordingly, scholars have increasingly classified livelihood strategies by attitudes toward environmental regulation and household development orientation^15,16. This study categorizes farmers’ livelihood strategies into three types: Passive adaptation, Proactive adjustment and Positive utilization. Passive adaptation livelihood strategies are characterized by smallholder farmers’ passive acceptance and negative adaptation to environmental regulations. Such households rely heavily on traditional agriculture for their livelihoods, absorb the costs and risks induced by regulatory constraints internally, and maintain a low-level and highly vulnerable livelihood status. Proactive adjustment livelihood strategies denote that smallholder farmers maintain basic agricultural engagement while taking a pragmatic attitude toward environmental regulations through active adjustment and partial transformation. They increase non-agricultural income by outsourcing agricultural production links and engaging in part-time labor activities. Positive utilization livelihood strategies indicate that smallholder farmers hold an embracing and opportunity-exploring attitude toward environmental regulations, and turn ecological constraints into green development opportunities to achieve a win-win situation for both environmental protection and livelihood improvement.

In recent years, domestic and foreign scholars have studied the effects of different types of environmental regulation on pro-environmental behavior^5–7,11 as well as the effects of socio-economic factors on the livelihood strategies of farmers^1,2,17. However, there is less literature that directly integrates environmental regulation with farmers’ livelihood strategies. For example, Qiu et al.⁸, empirically examined the impact of grassland ecological compensation policies on income-generating activities and livelihood diversification of rural beneficiaries in China with a Tobit model and data from 722 herding households in rural China. Wu⁹, conducted a systematic analysis of coupling mechanisms between village-scale rural livelihood systems and specific environmental conditions within a geographical analytical framework.

Effective policy design for mitigating external pressures on smallholder farmers necessitates multilevel integration of complex behavioral change frameworks with environmental governance practices. This requires comprehensive understanding of agricultural producers’ motivational drivers, value orientations, and livelihood development imperatives across both societal and individual dimensions to formulate informed interventions targeting livelihood strategy optimization¹⁸.

Therefore, this study aims to elucidate the mechanism of heterogeneous environmental regulation on smallholder farmers’ livelihood strategies under the framework of environmental identity theory, and adopts a multiple mediation model to explore the multiple mediating effects of environmental identity and self-efficacy in the mechanism of action from the dimensions of environmental identity strength and environmental identity salience, so as to systematically reveal the influencing mechanism of psychosocial factors on smallholder farmers’ livelihood strategy choices.

Theoretical mechanisms and hypothesis formulation

Environmental identity theory

In 1997, the American scholar Weigeit formally proposed Environmental Identity (EI) theory from a psychological perspective in response to the increasingly severe global environmental crisis. The core objective of this theory is to reshape individual environmental identity and alter the maladaptive environmental behaviors of actors, thereby providing feasible psychological pathways for addressing environmental problems¹⁹. Rooted in social identity theory and self-categorization theory, EI theory is closely aligned with core concepts in environmental psychology such as self-nature connection and place identity. Its central premise is to incorporate the natural environment into the construction of individual self-concept, forming a psychological bridge for human-nature interaction. And scholars Stets and Biga²⁰ broke through the limitations of traditional psychological research by focusing on the social attributes of environmental identity. They clearly defined environmental identity as a set of meanings and social roles that individuals actively assign to themselves when establishing deep connections with the natural environment. This advancement shifted environmental identity research from a purely individual psychological perspective to a dual framework encompassing both psychological and social dimensions, further enriching the theoretical connotation of EI.

EI theory holds that environmental identity serves as a critical pathway for individuals to construct self-concept. It represents a deep psychological bond between individuals and the natural environment, formed through historical experience, emotional attachment to nature, and perceived similarity with the natural world. This form of identity directly shapes individual behavioral patterns^21,22 and is also referred to as pro-environmental self-identity²³, highlighting its predictive and guiding role in pro-environmental behavior. Furthermore, the theory distinguishes two independent dimensions of EI: environmental identity strength (EIT) and environmental identity salience (EIA), each of which exhibits unique associations with individual behavioral tendencies²⁴. Specifically, EIT refers to the subjective intensity of the psychological connection between the self and nature perceived by individuals, while EIA denotes the frequency with which environmental identity is activated in environmentally relevant behavioral domains. EI theory has evolved through three key stages: an initial stage centered on individual psychology represented by Weigeit; a social construction stage led by Stets & Biga that integrated sociological perspectives; and a current integrative stage in which EI is deeply merged with behavioral theories such as the theory of planned behavior and the norm activation model, becoming a cornerstone of pro-environmental behavior research^22,25. This evolutionary trajectory provides a scientific and feasible framework for analyzing the intrinsic link between environmental policy and livelihood behavior.

Effective activation of environmental identity can significantly promote pro-environmental conduct among individuals, driven primarily by two core mechanisms: green identity preference and moral credibility^2, 26;. Green identity preference provides clear normative and value guidance for individual behavior, effectively strengthening the strength and salience of environmental identity and constraining unsustainable actions^27,28. The formation of such preference depends largely on an individual’s inherent environmental sensitivity; individuals with green identity preferences are more inclined to reduce resource consumption and adhere to sustainable development principles (Almudí & Sánchez, 2011). Under the moral credibility mechanism, past pro-environmental behaviors drive the formation of self-identity and accumulate moral credit^29,30. This moral credibility encourages behavioral consistency and generates positive behavioral spillover effects in new contexts, where spillover effects generally refer to the likelihood that an individual engages in (positive spillover) or avoids (negative spillover) identical or related behaviors in another situation. In turn, this promotes sustained pro-environmental behavior³¹; Frezza^26,31–33

For the specific group of smallholder farmers, effective activation of their environmental identity, strengthened by recalling past environmental practices, can further consolidate their environmental self-identity. This encourages them to proactively adopt more active and sustainable pro-environmental livelihood strategies when faced with environmental regulatory perturbations. In summary, EI theory offers a comprehensive and in-depth theoretical explanation for the intrinsic relationship between environmental regulation and smallholders’ livelihood strategies, while also providing solid theoretical support for the optimization and implementation of environmental policies.

Environmental regulation and smallholder livelihood strategies

There are more complex psychosocial mechanisms behind environmental regulation that promote a more pro-environmental and sustainable path for smallholder livelihoods. First, the value recognition mechanism of environmental regulation - environmental identity. Increasing the perceived and incentivized identity of environmental regulatory policies of the regulatory client is an effective path to strengthen the acceptability of environmental regulation³. EI theory suggests that in the context of pro-environmental behaviors, environmental identity operates at two levels: behavior-specific and general²³. The former helps to explain the persistence of performing specific pro-environmental behaviors (the environmental identity preference mechanism), while the latter may explain spillovers between pro-environmental behaviors (the moral credit mechanism). There is evidence that place attachment and environmental perceptions have significant positive effects on pro-environmental behaviors such as organic fertilizer use among farmers³⁴. Second, self-efficacy, the cognitive mechanism behind behavior change. In describing the assumptions of cognitive theory, Bandura³⁵, introduced the term self-efficacy, which suggests that every individual possesses a self-system that provides some control over the self’s thoughts, feelings, motivations, and behaviors, and that this self-system provides a reference mechanism and a set of subfunctions for perceiving, regulating, and evaluating behaviors. Self-efficacy is beliefs about a person’s ability to produce certain effects of their behavior and their ability to cope with situations³², and these beliefs lead to perceived behavioral control or self-efficacy over behavior³⁶. Self-efficacy can provide individuals with the ability to influence their own cognitive processes and behaviors, thus functioning as a self-regulator for changing the environment³⁷. EI theory states that environmental behaviors are complex and dependent in nature, and that the degree of difficulty in controlling behaviors, as well as the degree of similarity across contexts, affects the spillover effects of environmental identity, and that a combination of multidimensionality of environmental identity with self-efficacy is measurement of environmental behavior is necessary³².

Therefore, activating environmental identity can motivate smallholders to recognize the value of environmental regulatory tools and enhance their self-efficacy in coping with the constraints of environmental regulation, thereby promoting consistency of action and continuity of livelihood strategies and identity across contexts. In other words, the higher the level of smallholders’ environmental identity and self-efficacy, the higher the acceptability of environmental regulation, the more they can adopt more pro-environmental behaviors in their actual livelihood strategy choices. EI theory emphasizes the norms and values provided by environmental identity for individuals, dynamically characterizes the identity construction process of assimilation-adaptation-assessment, and provides a theoretical basis for the study of livelihood strategies of smallholder farmers under the perturbation of environmental regulation. Based on this, this paper puts forward the following hypotheses:

H1:

Environmental regulation has a significant positive impact on smallholder farmers’ livelihood strategy choices. With higher regulatory intensity, smallholders are more inclined to transform from passive adaptation to proactive adjustment and positive utilization livelihood strategies.

Mediating effects of environmental identity

Under the external perturbations caused by environmental regulation, smallholder farmers do not merely passively endure constraints. Instead, they engage in adaptive interactions of regulation perception–identity activation–strategy response through the mediating role of environmental identity, thereby forming differentiated livelihood choices. As a critical bridge connecting external institutions and individual behaviors of smallholders, environmental identity transforms external regulatory pressure into internalized practical actions via two pathways: improving the acceptability of environmental policies and motivating farmers to implement pro-environmental behaviors. This process further drives the evolutionary progression of smallholders’ livelihood strategies from passive adaptation to proactive adjustment, and finally to positive utilization.

First, environmental identity exerts direct or indirect impacts on the acceptability of environmental policies³⁸. As an external social norm, environmental regulation is perceived by smallholders through policy dissemination, demonstration, and guidance, and is then internalized into their own behavioral principles. Such internalized norms activate the environmental identity mechanism, drive farmers to pursue consistency between self-identity and behavioral practice, and encourage them to select livelihood strategies compatible with pro-environmental values³⁹. Existing studies show that pro-environmental beliefs and environmental identity have been found to have an indirect effect on the acceptability of policies such as the reduction of renewable energy consumption (Almudí & Sánchez, 2011), and have also been found to have an indirect effect on the acceptability of policies that limit energy consumption, as well as policies that reduce the amount of fertilizers^4,11. This suggests that individuals with stronger identification have more positive attitudes toward coercive policies and are more able to increase the acceptability of more coercive policies compared to individuals with weaker environmental identification⁴. Second, different dimensions of environmental identity are differently effective in predicting pro-environmental behavior. It has been shown that stronger EIT is more able to demonstrate individuals’ attitudes and behavioral intentions to actively support the environment and actively participate in specific sustainable behaviors^40,41. However, EIA, with its likelihood of activation in an individual’s thinking, is a central driver of the occurrence of motivation for identity-consistent behavior^42,43. Theory suggests that EIA and EIT are related, but the functions and effects of the two are structurally distinct (Reed et al., 2004). Applied research has shown that when individuals engage in environmentally relevant behaviors, individuals whose EI is frequently activated are more likely to engage in pro-environmental behaviors, and that EIA is superior in predicting pro-environmental behaviors compared to eit²⁴.

In summary, activating subjects’ environmental identity is conducive to increasing the acceptability of environmental policies, especially increasing the acceptability of more coercive policies, as well as stimulating subjects’ pro-environmental behaviors and even making behavioral decisions that may be costly. That is to say, when faced with environmental regulation perturbations, the more smallholders with a high degree of identification with environmental regulation tend choose the positive utilization livelihood strategy. At the same time, the impact of environmental regulation on smallholders’ livelihood strategies may also differ significantly depending on individual eit and EIA. Therefore, this paper proposes the following hypotheses:

H2:

The relationship between environmental regulation and smallholder livelihood strategies may be mediated by environmental identity, and this mediating effect can be heterogeneous across dimensions of environmental identity.

Mediating effects of self-efficacy

While there may be a mediating effect of environmental identity, Self-efficacy (SE) may also be a mediator in explaining the path from environmental regulation to smallholder livelihood strategies. Individuals’ behavioral choices are not only embedded in environmental identity, but are also embedded in the practice environment to reshape practice behaviors, which does not just mean repeating previous behaviors, but acting in a way that creates self-perceived congruence⁴⁴, i.e., believing that one’s own behaviors can have an impact. Farmers’ self-efficacy is a farmer’s judgment of the ability to organize and execute a set of actions to achieve the desired agricultural performance referred to³⁷. The higher the level of self-efficacy, the more confident people are in adhering to new behaviors⁴⁵. It has been shown that individuals responding to new demands and challenges at work will actively seek and implement changes to their work practices if they perceive the ability to find solutions and take control of these problems³². Roy³⁷, states that farmers with high self-efficacy believe they are more competitive, more challenging and adaptable, able to perceive environmental uncertainty, motivated to learn new skills, and easily understand the needs of society. That is, if individuals feel that their efforts are successful or that they have the ability to create desired effects, individuals will be more likely to adopt new behavioral decisions and abandon following conventions, and vice versa. Indeed, self-efficacy influences the degree of effort expended and perseverance on a task, directing individuals to seek to maintain problem-solving abilities and achieve the satisfaction of control⁴⁴.

The above analysis suggests that smallholder farmers who are able to be confident about the possible challenges and environmental consequences of environmental regulation will increase their sense of self-efficacy, pressures from environmental regulation will be mitigated or eliminated, and the flexibility of their livelihoods to adapt to environmental regulation will be increased, and then they will be more inclined to adopt pro-environmental livelihood strategies in their choices. Therefore, we argue that self-efficacy will increase the confidence of smallholder farmers to cope with environmental regulatory constraints and perturbations, and will adopt more active and pro-environmental livelihood strategies under the constraints and guidance of environmental regulations. This paper proposes the following hypothesis:

H3:

Self-efficacy mediates the relationship between environmental regulation and smallholder livelihood strategies.

Chained multiple mediating effects of environmental identity and self-efficacy

As hypothesized above, environmental identity and self-efficacy mediate the relationship between environmental regulation and smallholder livelihood strategies. However, when multiple mediators exist, is there a parallel or chain relationship between the two mediating paths of environmental identity and self-efficacy? Which mediating pathway plays a more critical role? According to EI theory, environmental identity and self-efficacy in promoting sustainable behaviors are intertwined and interdependent, i.e., environmental identity guides and enhances self-efficacy, and self-efficacy reinforces the construct of environmental identity³². Whitmarsh and O’Neill²³, argue that Contextual factors are also important drivers and barriers to pro-environmental behavior, and individuals are also unable to translate their self-identity into consistent pro-environmental behavior if they lack the ability to cope with behaviors in a given context. Mills¹⁸, also noted that higher willingness to participate in environmental management activities did not translate into actual behavior due to farmers’ lack of practical skills. It has also been demonstrated that increased self-efficacy further extends social identity, defining one’s self at the group level⁴⁶. Do et al.⁴⁷, argued that the interplay between self-efficacy and identity is inconsistent across all behaviors, and that the inclusion of personal variables (e.g., self-efficacy and social norms) to multidimensionally measure environmental behaviors is essential. Considering theoretical and empirical evidence^32,44, environmental identity and self-efficacy may chain or parallel mediate the relationship between environmental regulation and livelihood strategies. Therefore, this study hypothesized:

H4:

There is parallel or chained multiple mediating effects of environmental identity and self-efficacy in the impact of environmental regulation on smallholder livelihood strategies.

In summary, although the relationship between the variables of environmental regulation, environmental identity, self-efficacy and livelihood strategies has been independently studied in the literature, the role of environmental identity and self-efficacy in the impact of environmental regulation on livelihood strategies has not been thoroughly tested. Therefore, we tested the relationship between heterogeneous environmental regulation and smallholder livelihood strategies, as well as the mediating effects of environmental identity and self-efficacy, within the framework of environmental identity theory. Based on previous research, this study proposes four main hypothesized models (See Fig. 1):

Fig. 1.

Mechanism of multiple mediating effects.

Empirical model design

Baseline modeling

1

2

3

In the above equation, EI denotes environmental identity (including Strength of environmental identity and Environmental identity salience ), denotes environmental regulation (including Formal environmental regulation and Informal environmental regulation ), denotes a set of control variables, denotes the Perceived self-efficacy of smallholders, and denotes the Livelihood strategy of smallholders. and are all standardized regression coefficients. and are the residuals which are assumed to be normally distributed and independent of each other.

Description of data sources

The data used in this study were derived from a household survey conducted by the research team of China West Normal University in 2025. The purpose of this questionnaire survey was to collect smallholder farmers’ livelihood strategy choices when facing perturbations from environmental regulation. In this study, smallholder farmers refer to traditional farmers who are distinct from new agricultural operators (e.g., family farms, agricultural cooperatives, and agricultural enterprises).

Based on the overall level of economic development and agricultural development foundation, China was divided into four major regions: Eastern China, Central China, Western China, and Northeast China. A stratified random sampling method was adopted to randomly select one province from each region to represent the four regions, ultimately selecting four major agricultural provinces: Jiangxi Province, Hunan Province, Sichuan Province, and Heilongjiang Province. Subsequently, 40 towns were randomly selected, and 100 administrative villages were further sampled from these towns, with at least 8 smallholder households randomly interviewed in each administrative village.

Meanwhile, more than 150 interviewers were recruited from four universities to participate in this survey, namely Jiangxi Normal University in Jiangxi Province, Hunan University of Science and Engineering in Hunan Province, China West Normal University in Sichuan Province, and Heilongjiang Bayi Agricultural University in Heilongjiang Province. All interviewers were undergraduates and postgraduates majoring in agricultural economics or related economic disciplines. The recruited interviewers were divided into four groups, conducting one-on-one interviews with each sample household in the four sample provinces respectively.

The questionnaire mainly consisted of four parts: (1) the socioeconomic characteristics of farm households; (2) farm household welfare and the characteristics of their respective villages; (3) farmers’ attitudes toward different types of environmental regulation; (4) the evaluation of farmers’ environmental identity and self-efficacy. Before the formal household survey, more than 150 interviewers received formal online or offline research training and simulated interview training from the research team to ensure that each interviewer had an accurate understanding of the structured questionnaire. To further improve the validity of the questionnaire, each respondent voluntarily agreed to complete all items of the questionnaire within 20 min.

A total of 850 questionnaires were distributed. After screening and classifying the questionnaire responses, 804 valid questionnaires were obtained, with an effective rate of 94.6%. This dataset has good representativeness of agricultural households in major agricultural regions and provides a solid foundation for the empirical analysis of this study.

Selection of indicators

Smallholder livelihood strategies (cl)

Drawing on the classification of livelihood strategies proposed by Tsvegemed et al.¹⁶, and Ma Cheng et al.¹⁵, this study categorizes smallholders’ livelihood strategies into three types: passive adaptation, proactive adjustment, and positive utilization. Respondents selected the strategy that best aligned with their attitudes, perceptions, and behavioral tendencies through six questions designed for each strategy type. The answers were coded as follows: 1 represents passive adaptation strategy, 2 represents proactive adjustment strategy and 3 represents positive utilization strategy.

(1) Passive adaptation livelihood strategy includes the following items: I cannot resist environmental policies such as restrictions on pesticide and chemical fertilizer application, as well as the resource utilization of livestock and poultry manure, and can only do what I am capable of; In accordance with the requirements of agricultural environmental policies, we have done all that we can think of, and we have no other options; I do not know how to adapt to the provisions of agricultural environmental policies, nor do I know what adverse impacts the implementation of these policies will have on me; In the process of adapting to agricultural environmental policies, it is enough to do my best, and there is no need to worry about other things; I am willing to change traditional farming and breeding methods to reduce agricultural environmental pollution; Appropriately reduce the application of pesticides and chemical fertilizers, and try to treat pollutants to avoid violations.

(2) Proactive adjustment livelihood strategy includes the following items: To avoid violating the provisions of agricultural environmental policies, I will rent my land to others for cultivation; The environmental requirements for livestock and poultry breeding are too high, so I will give up livestock breeding; I will manage the land myself, but entrust professional enterprises to handle crop planting, fertilization, pesticide spraying, and straw treatment; I will breed livestock and poultry myself, but entrust professional enterprises to carry out resource utilization of feces and other wastes; Traditional planting and breeding industries are subject to many environmental restrictions, so I will switch to the non-agricultural sector; Give up crop planting and livestock breeding, and obtain income by taking temporary jobs or finding fixed jobs in the non-agricultural sector.

(3) Positive utilization livelihood strategy includes the following items: For me, it is easy to make changes to adapt to agricultural environmental policies, and I can adapt; Using existing resources and technologies, I think I can meet the relevant requirements in agricultural environmental policies; I am a very confident person who likes to face challenges and can achieve success; Improving crop and livestock varieties and introducing new technologies for comprehensive waste treatment can reduce agricultural environmental pollution; I will adjust the planting structure and optimize the livestock and poultry breeding structure to reduce agricultural environmental pollution; I will actively expand income channels other than planting and breeding industries (e.g., non-agricultural employment).

Environmental regulation (GZ)

Synthesize the categorization of different types of environmental regulation and the design of the scale from the studies of Li et al.⁷, Xia et al.⁶, and Wang et al.⁵. (1) Formal environmental regulation (gzz). Considering mainly coercive tools such as administrative penalties or economic incentives for environmental regulation, this study developed a set of 5-item, 5-point composite rating scale to assess the level of constraints of smallholder farmers to formal environmental regulation. Specifically, respondents were asked the following five questions: strong government regulation in the rational use of straw, restriction of pesticide and chemical fertilizer application, and resource utilization of livestock and poultry manure; well-established laws and regulations on agricultural pollution prevention and control; financial compensation from the government for the reduction of agricultural pollution emissions; appropriate policy concessions from the government for the protection of the environment in agriculture; and financial compensation from the government for the supportive facilities and equipments for the treatment of agricultural pollutants construction to give financial compensation. The responses of 1 represents disagree, 2 represents somewhat disagree, 3 represents generally, 4 represents somewhat agree, and 5 represents agree. Finally, the scores of the five items were summed and averaged, with the higher the score, the higher the constraint level of formal environmental regulation. The Cronbach’sα value of the scale is 0.890 (see Table 2), and the scale shows very good reliability. (2) Informal environmental regulation (gzf). Considering mainly non-coercive tools such as self-awareness, village rules and folklore, this study measured it through four questions: if someone pollutes the agricultural environment, I will blame him; I will take the initiative to adopt protective behaviors to reduce pollution in agricultural production and livestock breeding activities; pollution of the agricultural environment affects my own health as well as that of other people; and pollution of the agricultural environment affects food safety. Similarly, the response options ranged from disagree (1) to agree (5), and the four questions were summed to calculate the final score by taking the mean of the four questions; the higher the score, the higher the level of constraints of informal environmental regulation. The Cronbach’sα value for the scale was 0.776 (see Table 2), showing good reliability.

Table 2.

Descriptive statistics, correlation and multicollinearity tests for the main variables (N = 804).

Vare	Aver	SD	α	VIF	cl	gzz	gzf	eit	eia	gnl
cl	2.218	0.852	–	–	1	0.111***	0.156***	0.188***	0.164***	0.173***
gzz	3.744	1.052	0.890	1.280		1	0.401***	0.305***	0.189***	0.364***
gzf	4.270	0.788	0.776	1.510			1	0.511***	0.258***	0.346***
eit	4.344	0.705	0.813	1.620				1	0.264***	0.499***
eia	4.020	0.737	0.640	1.140					1	0.304***
gnl	4.054	0.756	0.730	1.480						1

Note: ***, **, and * indicate significant at the 1%, 5%, and 10% levels, respectively.

Environmental identity (EI)

We specifically consider two dimensions of environmental identity, identity strength and identity salience, in our study. (1) Environmental identity strength (eit). A more concise indicator of environmental identity strength has been shown to be positively correlated with the acceptability of a range of environmental regulatory policies²³. eit is measured using a four-item, five-point scale developed in the study: i am very concerned about pollution of the agro-environment; for those who live in the countryside, everyone should take care of our living production environment; I feel worried when I find that my living and production environment is polluted; Legislation is necessary for major agro-environmental pollution. Responses were 1 = disagree, 2 = somewhat disagree, 3 = generally, 4 = somewhat agree, and 5 = agree. Finally, the scores of the four items were summed and averaged, the higher the score the higher the level of intensity of agreement. The Cronbach’sα value of the scale is 0.813 (see Table 2), showing very good reliability. (2) Environmental identity salience (eia). The core of environmental identity salience is to measure the frequency with which environmental identity is activated in smallholders’ daily cognition and decision-making. Specifically, it refers to smallholders’ psychological tendency to actively incorporate environmental factors into their cognitive framework and prioritize environmental impacts in production and living scenarios. To avoid behavioral overlap with the dependent variable (livelihood strategies), this study draws on the design idea of the environmental identity scale proposed by Rahmani et al.²⁴,, and designs four items combined with the cognitive characteristics of smallholders, focusing on the activation state of identity at the psychological cognitive level. The specific items are as follows: When planning agricultural production arrangements, I often give priority to considering the impact on the environment; In my daily production and life, I often pay attention to matters related to agricultural environmental protection; When communicating with others about agricultural production topics, I often mention the importance of environmental friendliness; When making production-related decisions, I often think of the environmental protection responsibilities I should bear. Similarly, the answer options range from Disagree (1) to Agree (5). The final score is calculated by taking the average of the four questions; a higher score indicates that smallholders’ environmental identity is activated more frequently in daily cognition and decision-making, and the level of environmental identity salience is stronger. The Cronbach’sα value for the scale was 0.640 (see Table 2), indicating good reliability.

Self-efficacy (gnl)

Based on the definition of the concept by Bandura³⁵ and the scale developed by Roy³⁷, a 3-item 5-point composite scale was developed in this study to assess the self-efficacy of smallholder farmers. The were three questions specified: (1) I can adapt if it is easy for me to make changes to adapt to agri-environmental policies; (2) I think I can fulfill the requirements associated with agri-environmental policies with the available resources and technology; (3) I am a very self-confident person, I like to face challenges and I can be victorious. Similarly, the response options ranged from disagree one through five, which were summed to get their final score, with the higher the score the higher the self-efficacy. The Cronbach’s α value for the scale was 0.730 (see Table 2), showing good reliability.

Control variables (X)

Three aspects of control variables that may be related to environmental identity, self-efficacy, and livelihood strategies were considered. Household socio-economic characteristics include age of the head of household (ss, years), experience in the field (ex, years), household income (sr, yuan), land scale (gd, mu) and asset ownership status(zc. which characterized by the number of asset types owned by the household, including 9 categories: cars, agricultural tractors, agricultural machinery such as sowing and harvesting equipment, agricultural tricycles, refrigerators, color TVs, grain warehouses, brick-and-tile houses with ownership rights, and well-equipped livestock and poultry breeding sheds.); aspects of farm household well-being include health (jk, 0 = unhealthy, 1 = healthy), and years of education (jy, years); and rural accessibility reflecting rural characteristics (xs, hours), which is defined in terms of the time it takes to get to the township government by adopting the most commonly used means of transportation. The basic statistical description of the control variables is shown in Table 1.

Table 1.

Statistical description of control variables.

Var	Aver	SD	Min	25th	Med	75th	Max	Sample
ss	47.857	17.081	18	40	48	60.	80	804
jk	0.598	0.491	0	0	1	1	1	804
jy	7.874	4.466	0	5	8	11	19	804
ex	19.177	16.534	0	2	20	30	60	804
gd	15.919	37.820	0	0.200	5	15	450	804
zc	4.151	1.697	0	3	4	5	9	804
sr	5.812	5.856	0.160	1.700	4.170	7.515	25.200	804
xs	0.629	0.735	0	0.250	0.500	0.675	6.600	804

Model estimation procedure

Compared to structural equation modeling estimation, the use of estimating the parameters of each equation independently allows for greater flexibility and control in the analysis, allows for a more effective examination of the relationship between the main variables rather than considering all equations at once⁴⁸, and estimating the parameters independently allows for a more effective treatment of complex and nonlinear relationships⁴⁹. This study used the chained multiple mediator model developed by the SPSS macro PROCESS 3.5 Model 6⁵⁰ in order to test the chained path between the two mediators of environmental identity and self-efficacy and the mediating effects of each mediator variable independently, specifically based on 5,000 bootstrap samples to compute whether the mediating effects were significant at 95%. The study mainly used SPSS 25.0 and macro PROCESS 3.5 for data analysis and processing.

Analysis of results

Descriptive statistics, correlation analysis and multicollinearity tests

As shown in Table 2, the more positive the attitude of smallholder farmers towards environmental regulation (formal and informal), the higher the strength and significance of environmental identification, the stronger the sense of self-efficacy, and the more active smallholder farmers will be in adopting pro-environmental livelihood strategies. Higher environmental identification and stronger self-efficacy were associated with more active livelihood strategies. Stronger self-efficacy is associated with more active livelihood strategies. There is a significant positive correlation between the main variables, which also provides a good basis for subsequent hypothesis testing.

In order to eliminate the problem of estimation bias due to multicollinearity among the main variables, the study mainly uses two ways to conduct the test. One is the variance inflation factor test. As can be seen from Table 2, the VIF coefficients of the main variables are all close to 1, which is much smaller than the threshold criterion of 3, indicating that there is no problem of multicollinearity between the main variables. Second, correlation. In order to judge whether there is multicollinearity between the main variables, we draw on the research idea of Ansari et al.⁴⁸, by testing the correlation coefficients between the six main variables. From the correlation coefficients presented in Table 2, the maximum correlation coefficient between gzf and eit is 0.511, and the correlation coefficients between the other variables do not exceed 1, and all of them are lower than the threshold of 0.75 (Kline, 2023), which shows that each variable is different from the others, and the hypothesis of multiple covariance between the six main variables does not exist.

Tests of the mediating effect of environmental identity

First, the results in columns (1) and (2) of Table 3 show that the positive effects of both gzz and gzf on eit are significant at the 1% level. Meanwhile, columns (5) and (6) show a significant positive correlation between eit and cl, while the effects of gzz and gzf on cl are not significant. The results indicate that heterogeneous environmental regulation can indirectly optimize smallholders’ livelihood strategies by enhancing the intensity of their environmental identity, presenting a full mediating effect (see Fig. 2). This implies that environmental regulation does not directly intervene in smallholders’ livelihoods; instead, it reshapes their environmental identity to guide them in proactively adjusting their livelihood choices. In practice, promoting environmental regulation should not only focus on mandatory enforcement but also strengthen smallholders’ environmental identity through publicity and guidance, as well as participatory governance, enabling environmental protection concepts to be integrated into their livelihood decision-making. Hypothesis 2 of this study is confirmed.

Table 3.

Results of Multiple Chained Mediation Effect Tests for Strength of Environmental Identity and Perceived Self-Efficacy.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	eit		gnl		cl		Total Effect
gzz	0.302*** (8.999)		0.361*** (10.968)		0.033 (0.874)		0.105*** (2.974)
gzf		0.505*** (16.512)		0.343*** (10.274)		0.072 (1.779)		0.156*** (4.429)
eit			0.425*** (13.548)	0.433*** (12.166)	0.131*** (3.209)	0.103** (2.339)
gnl					0.092** (2.211)	0.092** (2.275)
ss	0.002 (0.543)	0.000 (0.088)	0.004 (1.234)	0.003 (0.863)	0.002 (0.432)	0.001 (0.362)	0.002 (0.592)	0.002 (0.441)
jk	−0.008 (−0.112)	−0.050 (−0.778)	−0.038 (−0.540)	−0.060 (−0.855)	−0.018 (−0.244)	−0.025 (−0.334)	−0.023 (−0.302)	−0.035 (−0.475)
jy	0.0211** (2.430)	0.019** (2.375)	0.015* (1.767)	0.014* (1.659)	0.003 (0.358)	0.003 (0.341)	0.007 (0.795)	0.006 (0.696)
ex	−0.001 (−0.150)	0.000 (−0.127)	−0.003 (−0.829)	−0.003 (−0.900)	−0.006 (−1.613)	−0.006 (−1.602)	−0.006 (−1.677)	−0.006 (−1.677)
gd	−0.001 (−1.418)	−0.001 (−0.569)	0.001 (1.172)	0.002* (1.864)	0.001 (0.529)	0.001 (0.621)	0.000 (0.450)	0.001 (0.729)
zc	0.047** (2.333)	0.039** (2.130)	0.030 (1.542)	0.028 (1.383)	0.003 (0.131)	0.002 (0.111)	0.012 (0.544)	0.009 (0.421)
sr	−0.008 (−1.365)	0.000 (0.006)	−0.008 (−1.468)	−0.003 (−0.576)	0.003 (0.467)	0.004 (0.607)	0.001 (0.161)	0.003 (0.551)
xs	−0.069 (−1.508)	0.000 (0.088)	−0.050 (−1.123)	0.003 (0.863)	0.025 (0.534)	0.001 (0.362)	0.105*** (2.974)	0.002 (0.441)
_cons	−0.328* (−1.659)	−0.277 (−1.577)	−0.309 (−1.590)	−0.282 (−1.471)	−0.029 (−0.143)	−0.001 (−0.006)	−0.101 (−0.208)	−0.056 (−0.275)
R2	0.338	0.521	0.383	0.364	0.228	0.233	0.143	0.183
F	11.341	37.081	15.179	15.139	3.937	4.559	1.842	3.427
P	0.000	0.000	0.000	0.000	0.000	0.000	0.058	0.001

Note: ① ***, **, and * denote significant at the 1%, 5%, and 10% levels, respectively; t-values in parentheses are t-values after taking robustness standard errors; coefficients are standardized coefficients. ② Columns (1), (3), (5), and (7) in the above table are the results of operations with formal environmental regulation as the dependent variable, and columns (2), (4), (6), and (8) are the results of operations with informal environmental regulation as the dependent variable.

Fig. 2.

Multiple mediation pathways mediated by environmental identity strength. Note: The dashed line represents the path with informal environmental regulation as the independent variable, while the solid line represents the path with formal environmental regulation as the independent variable.

Second, from columns (1) and (2) in Table 4, both gzz and gzf have a positive effect on eia at the 1% significance level, with coefficients of 0.159 and 0.269, respectively. from the results in columns (5) and (6), it can be seen that eia also has a significant positive effect on cl, and both gzz and gzf have a positive effect on cl, the latter passing the 5% significance level test while the former was not significant. The results indicate that environmental identity salience fully mediates between formal environmental regulation and livelihood strategies, while it partially mediates between informal environmental regulation and livelihood strategies (see Fig. 2). The possible reason is that formal environmental regulation relies on institutional constraints to form external pressure, and its impact on smallholders’ livelihoods is entirely dependent on psychological transmission. In contrast, informal environmental regulation (such as community norms and neighborhood influence) has both restrictive and guiding functions: it can indirectly act through the salience of environmental identity, and also directly affect smallholders’ livelihood choices through social norms. In practice, informal environmental regulation can serve as a supplement to formal regulation, activating grassroots environmental consensus and promoting the transformation of smallholders’ livelihoods towards green development. The results further confirm that Hypothesis 2 of this study is valid.

Table 4.

Results of multiple chained mediation effect tests for environmental identity salience and perceived self-Efficacy.

Variable	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
	eia		gnl		cl		Total Effect
gzz	0.195*** (5.606)		0.313*** (9.651)		0.039 (1.029)		0.104*** (2.962)
gzf		0.269*** (7.820)		0.280*** (8.326)		0.091** (2.408)		0.156*** (4.426)
eia			0.246*** (7.606)	0.233*** (6.970)	0.125*** (3.424)	0.113*** (3.053)
gnl					0.115*** (2.963)	0.102*** (2.669)
ss	−0.001 (−0.141)	−0.002 (−0.414)	0.004 (1.316)	0.003 (0.967)	0.002 (0.491)	0.002 (0.421)	0.002 (0.599)	0.002 (0.452)
jk	−0.080 (−1.083)	−0.101 (−1.395)	−0.018 (−0.267)	−0.037 (−0.537)	−0.008 (−0.110)	−0.018 (−0.240)	−0.022 (−0.301)	−0.035 (−0.473)
jy	0.000 (−0.040)	−0.002 (−0.204)	0.015* (1.840)	0.014* (1.709)	0.006 (0.632)	0.005 (0.589)	0.007 (0.808)	0.007 (0.715)
ex	0.003 (0.768)	0.003 (0.788)	−0.004 (−1.065)	−0.004 (−1.108)	−0.006* (−1.711)	−0.006* (−1.702)	−0.006* (−1.680)	−0.006* (−1.682)
gd	0.000 (−0.166)	0.000 (0.326)	0.001 (1.258)	0.002* (1.845)	0.000 (0.347)	0.001 (0.514)	0.000 (0.447)	0.001 (0.725)
zc	0.014 (0.651)	0.009 (0.458)	0.027 (1.420)	0.025 (1.282)	0.007 (0.311)	0.005 (0.255)	0.012 (0.552)	0.009 (0.433)
sr	0.008 (1.349)	0.012** (2.054)	−0.010* (−1.882)	−0.006 (−1.113)	0.001 (0.158)	0.002 (0.392)	0.001 (0.165)	0.003 (0.558)
xs	−0.039 (−0.816)	−0.029 (−0.632)	−0.041 (−0.942)	−0.026 (−0.592)	0.022 (0.472)	0.023 (0.494)	0.012 (0.243)	0.017 (0.349)
_cons	−0.055 (−0.269)	−0.005 (−0.022)	−0.295 (−1.573)	−0.252 (−1.322)	−0.058 (−0.284)	−0.044 (−0.217)	−0.101 (−0.484)	−0.071 (−0.341)
R2	0.210	0.278	0.453	0.427	0.231	0.243	0.143	0.183
F	4.080	7.403	20.424	17.743	4.071	4.526	1.842	3.056
P	0.000	0.000	0.000	0.000	0.000	0.000	0.058	0.001

Note: ① ***, **, and * denote significant at the 1%, 5%, and 10% levels, respectively; t-values in parentheses are t-values after taking robustness standard errors; coefficients are standardized coefficients. ② Columns (1), (3), (5), and (7) in the above table are the results of operations with formal environmental regulation as the dependent variable, and columns (2), (4), (6), and (8) are the results of operations with informal environmental regulation as the dependent variable.

In summary, both dimensions of environmental identity can transmit the impact of environmental regulation, which verifies the bridging role of environmental cognition. This provides a psychological entry point for optimizing regulation and improving smallholders’ livelihoods. Meanwhile, it partially verifies Hypothesis 1, that is, the positive impact of heterogeneous environmental regulation on smallholders’ livelihood strategies is mainly achieved through the mediating effect of environmental identity and self-efficacy.

Tests of the mediating effect of self-efficacy

On the one hand, the results in columns (3) and (4) of Table 3 show that there is a positive correlation between gzz, gzf and gnl. Meanwhile, columns (5) and (6) show that there is also a positive correlation between gnl and cl, while the positive correlation between gzz and gzf and cl is not significant. The results suggest that self-efficacy plays a fully mediating role between environmental regulation and livelihood strategies (see Fig. 3). On the other hand, as shown by the results in columns (3) and (4) of Table 4, both gzz and gzf positively influence. gnl also positively influences cl significantly, however, the positive relationship between gzz and cl was not significant and the positive correlation between gzf and cl was significant. The results suggest that self-efficacy plays a crucial mediating role between environmental regulation and livelihood strategies, with distinct mediating types observed in formal and informal environmental regulation: full mediation in formal regulation and partial mediation in informal regulation. This finding implies that higher self-efficacy enhances smallholders’ willingness and ability to respond to environmental changes and adopt proactive livelihood behaviors. Environmental regulation reshapes the livelihood environment of smallholders, and improving self-efficacy can help them adapt to such changes and explore green livelihood models. In practice, promoting environmental regulation should emphasize enhancing smallholders’ self-efficacy. Specific measures such as skill training and policy support can strengthen their green production capabilities. In particular, since informal regulation has limited direct impacts, it is even more necessary to strengthen its mediating role. Hypothesis 3 of this study was confirmed.

Fig. 3.

Multiple mediation pathways mediated by strength of environmental identity. Note: The dashed line represents the path with informal environmental regulation as the independent variable, while the solid line represents the path with formal environmental regulation as the independent variable.

Tests for multiple intermediation effects

Multiple mediation effect test with environmental identity strength as a mediating variable

As can be seen in the left half of Table 5, the total and mediating effects of formal environmental regulation on the livelihood strategies of smallholder farmers are significant, whereas the direct effect in the 95% confidence interval [−0.030, 0.094] contains 0 and is not significant. The test results indicate that the total effect of formal environmental regulation on smallholders’ livelihood strategies is mainly achieved through three pathways: the parallel mediation of environmental identity intensity, the parallel mediation of self-efficacy, and the chain mediation of environmental identity intensity→ self-efficacy. All three pathways can effectively transmit the impact of formal environmental regulation, with the parallel mediation pathways accounting for a higher proportion. This finding suggests that formal environmental regulation can promote the optimization of smallholders’ livelihood strategies by simultaneously activating their environmental identity and self-efficacy. When formulating formal environmental regulation policies, government departments should take into account the psychological guidance of both factors, strengthen smallholders’ environmental identity and self-efficacy, and construct a complete transmission chain of policy guidance→ psychological activation→ behavioral optimization.

Table 5.

Path test for multiple chain mediation model mediated by strength of environmental identity.

Type	Value	Boot SE	lower limit	limit	Proportion	Type	Value	Boot SE	lower limit	limit	Proportion
Total	0.104	0.035	0.046	0.163	100	Total	0.156	0.035	0.098	0.214	100
Direct	0.032	0.038	−0.030	0.094	30.77	Direct	0.072	0.041	0.005	0.139	46.12
Mediated	0.073	0.016	0.047	0.101	70.19	Mediated	0.084	0.023	0.048	0.121	53.88
d1:gzz→eit→cl	0.040	0.014	0.018	0.063	38.46	d1:gzf→eit→cl	0.053	0.023	0.014	0.091	33.68
d2:gzz→gnl→cl	0.021	0.011	0.005	0.041	20.19	d2:gzf→gnl→cl	0.012	0.007	0.002	0.025	7.38
d3:gzz→eit→gnl→cl	0.012	0.006	0.003	0.021	11.54	d3:gzf→eit→gnl→cl	0.020	0.009	0.006	0.036	12.89
(C1)	0.018	0.021	−0.016	0.053		(C1)	0.041	0.026	−0.002	0.084
(C2)	0.028	0.016	0.003	0.056		(C2)	0.032	0.028	−0.013	0.078
(C3)	0.010	0.007	0.001	0.022		(C3)	−0.009	0.007	−0.022	0.001

Note: ① Utility values are fully standardized estimates. ② (C1), (C2), and (C3) are the contrast mediating effects, where (C1) is the contrast mediating effect of path d1-d2, (C2) is the contrast mediating effect of path d1-d3, and (C3) is the contrast mediating effect of path d2-d3.

The results presented in the right half of Table 5 show that the total effect, direct effect, and mediating effect of informal environmental regulation on smallholders’ livelihood strategies are all significant. These results confirm that the multiple mediating pathways of informal environmental regulation are similar to those of formal environmental regulation: all three pathways are significant, with no obvious differences in the importance of each pathway (see Fig. 2). This indicates that as an important form of grassroots governance, the impact of informal environmental regulation on smallholders’ livelihoods also relies on the dual psychological transmission of environmental identity intensity and self-efficacy. Compared with formal environmental regulation, informal environmental regulation is more flexible and closer to farmers. It can rely on community networks to simultaneously enhance smallholders’ environmental identity and self-efficacy through publicity and education, typical demonstrations, and other means, giving play to the advantages of grassroots governance and promoting livelihood transformation.

In conclusion, both environmental identity intensity and self-efficacy exert parallel mediating and chain mediating effects between heterogeneous environmental regulation and smallholders’ livelihood strategies, with no significant differences in the importance of the two parallel mediating pathways.

Multiple mediation effect test with environmental identity significance as mediating variable

As can be seen in the left half of Table 6, the mediation effect of the formal environmental regulation→ environmental identity salience→ smallholder livelihood strategies and formal environmental regulation→ self-efficacy→ smallholder livelihood strategies paths had significant mediation effects, with the mediation effect accounting for 19.50% and 18.28% of the total effect, respectively, indicating that environmental identity salience and self-efficacy mediated the relationship between formal environmental regulation and smallholder livelihood strategies. Similarly, the chain mediation effect of formal environmental regulation→ environmental identity salience→ self-efficacy→ smallholder livelihood strategies was significant, with a mediation effect of 4.11% of the total effect. In formal environmental regulation, the parallel mediating pathways associated with environmental identity salience and self-efficacy are both statistically significant and of comparable importance, whereas the chain mediating pathway exerts a weak effect (see Fig. 3). This finding demonstrates that the impact of formal environmental regulation on smallholders’ livelihood strategies is primarily realized through the direct activation of environmental identity salience and self-efficacy, rather than relying on the chain transmission between these two psychological factors.

Table 6.

Path test for multiple chained mediation model mediated by environmental identity salience.

Type	Value	Boot SE	lower limit	limit	Proportion	Type	Value	Boot SE	lower limit	limit	Proportion
Total	0.104	0.035	0.046	0.163	100	Total	0.156	0.035	0.098	0.214	100
Direct	0.039	0.037	−0.023	0.100	58.18	Direct	0.091	0.038	0.029	0.153	58.33
Mediated	0.066	0.016	0.041	0.092	41.82	Mediated	0.065	0.017	0.039	0.095	41.67
d1:gzz→eia→cl	0.025	0.009	0.011	0.040	19.50	d1:gzf→eia→cl	0.030	0.012	0.013	0.051	19.23
d2:gzz→gnl→cl	0.036	0.013	0.016	0.057	18.28	d2:gzf→gnl→cl	0.029	0.012	0.010	0.050	18.59
d3:gzz→eia→gnl→cl	0.006	0.002	0.002	0.010	4.11	d3:gzf→eia→gnl→cl	0.006	0.003	0.002	0.012	3.85
(C1)	−0.011	0.017	−0.039	0.017		(C1)	0.002	0.018	−0.028	0.032
(C2)	0.019	0.009	0.005	0.035		(C2)	0.024	0.012	0.005	0.045
(C3)	0.030	0.011	0.013	0.050		(C3)	0.022	0.010	0.008	0.040

Note: (1) Utility values are fully standardized estimates. (2) (C1), (C2), and (C3) are the contrast mediating effects, where (C1) is the contrast mediating effect of path d1-d2, (C2) is the contrast mediating effect of path d1-d3, and (C3) is the contrast mediating effect of path d2-d3.

Theoretically, this phenomenon reflects the mandatory nature of formal environmental regulation. Unlike informal regulation that relies on voluntary compliance, formal environmental regulation is backed by institutional constraints and mandatory enforcement mechanisms, which allows it to directly shape smallholders’ psychological cognition and behavioral tendencies without the need for multiple intermediate psychological transmission links. In practical terms, the promotion of formal environmental regulation should prioritize the enhancement of smallholders’ environmental identity salience and self-efficacy. By streamlining redundant psychological transmission links, the efficiency of policy implementation can be improved, thereby better facilitating the optimization of smallholders’ livelihood strategies.

The results presented in the right half of Table 6 show that the 95% confidence intervals for the three paths of informal environmental regulation→ significance of environmental identity→ smallholder livelihood strategy, informal environmental regulation→ self-efficacy→ smallholder livelihood strategy and informal environmental regulation→ significance of environmental identity→self-efficacy→smallholder livelihood strategy do not contain 0, accounting for 19.08% of the total effect, respectively. The 95% confidence intervals of the three paths of regulation→environmental identity salience→self-efficacy→smallholder livelihood strategy do not contain 0, and the mediation effects are all significant, accounting for 19.23%, 18.59% and 3.85% of the total effect, respectively. In terms of the comparative mediation effect, the 95% confidence interval of (C1) contains 0, the comparative mediation effect is not significant, and the two paths of informal environmental regulation→ environmental identity salience→ smallholder livelihood strategy and informal environmental regulation→ self-efficacy→ smallholder livelihood strategy are parallel. In informal environmental regulation, all three mediating pathways are statistically significant, and the parallel mediating effect is significantly higher than the chain mediating effect, which is consistent with the results of formal environmental regulation (see Fig. 3). The difference lies in that, in informal environmental regulation, the parallel mediating pathways of environmental identity salience and self-efficacy account for the same proportion, indicating that both play equally important roles in transmitting the impact of informal environmental regulation. This finding suggests that informal environmental regulation can rely on community interaction to simultaneously strengthen smallholders’ environmental identity salience and self-efficacy. Through two-way psychological guidance, it can promote smallholders to proactively adjust their livelihood strategies, achieving a win-win situation between environmental protection and livelihood improvement.

In summary, environmental identity (intensity and salience) and self-efficacy form a multiple mediating system of “parallel + chain” between heterogeneous environmental regulation and smallholders’ livelihood strategies, with the parallel mediating pathways exerting a more prominent role, and the results support Hypothesis 4 of this study.

Discussion and limitations

Discussion

Within the framework of environmental identity theory, this study attempts to answer the following pressing but realistic questions: are there differences in the impacts of heterogeneous environmental regulations on smallholder livelihood strategies? Which type of environmental regulation is more effective in promoting more pro-environmental livelihood strategies among smallholder farmers? What are the social-psychological mechanisms underlying the effects of environmental regulation on smallholder livelihood strategies? To this end, this paper utilizes 804 smallholder data from four provinces and analyzes the data using the macro PROCESS 3.5, discusses and explains the relationship between heterogeneous environmental regulation and smallholder livelihood strategies, and empirically tests the multiple mediating effects of environmental identity strength, salience, and self-efficacy in the relationship, with the following conclusions:

(1) Regardless of the model test with either eit or eia as the mediating variable, the effect of gzz on cl is not significant, while only when eia is used as the mediating variable gzf has a significant positive effect on cl. First, it has been shown in the literature that both formal and informal environmental regulation can significantly and positively influence the willingness of large-scale pig farmers to utilize resources⁶, and the behavior of farmers’ film recycling⁵. Unfortunately, however, this study did not confirm that formal environmental regulation has a significant positive impact on smallholder farmers’ livelihood strategies. The empirical results suggest that neither the penalizing nor incentivizing instruments of formal environmental regulation directly lead to a shift in smallholder livelihoods towards more pro-environmental livelihood strategies. Possible explanations are: first, smallholder more pro-environmental livelihood need to be integrated with the geography, resource endowment, and industrial development status of the local countryside. Due to the vastness of China’s rural areas, the imbalance of natural endowments, industrial development and environmental technology levels in the countryside has had a heterogeneous impact on the spatial pattern of rural livelihoods, and their development capacity⁹, resulting in the fact that positive utilization livelihood strategies have not become a common choice for smallholder farmers. Secondly, formal environmental regulation has experienced the phenomenon of relative institutional failure and command but no control^7,8. Formal environmental regulation uses instruments such as penalties and economic subsidies to constrain smallholders’ bad behaviors, and does not directly create more opportunities for livelihood activities or is inconsistent with environmental strategy objectives, resulting in smallholder livelihood strategy choices that do not achieve the expected effects of the policy. Second, research has shown that informal environmental regulation is a self-regulatory ideology formed through farmers’ self-awareness and village rules and customs, and that more stable village rules and customs formed over a long period of time can greatly enhance the salience of environmental identity and stimulate the spillover effects of smallholder behavior (Frezza³²,, which will lead to a more active choice of positive utilization livelihood strategies.

(2) Environmental identity mediates the relationship between heterogeneous environmental regulation and smallholder livelihood strategies, and this mediating effect is heterogeneous depending on the dimensions of environmental identity. Specifically, environmental identity strength fully mediated the relationship between heterogeneous environmental regulation and livelihood strategies; environmental identity salience fully mediated the relationship between formal environmental regulation and livelihood strategies, while environmental identity salience partially mediated the relationship between informal environmental regulation and livelihood strategies. The results are not difficult to understand that formal environmental regulation is mostly devoted to constraining or rewarding smallholders’ environmental behaviors by means of policies or economic compensation⁷, and the lack of sufficient guidance and publicity makes formal environmental regulation less acceptable to smallholders³. ei Theory suggests that environmental identity is a part of an individual’s self-definition of the environment in relation to a sense of awareness related to the natural environment²², and that smallholder farmers’ livelihood strategy choices are not made in isolation, but embed environmental identity in preference structures and moral systems, which provide them with norms and values (Almudí & Sánchez, 2011²⁶;, limiting their undesirable behaviors and stimulating a shift in livelihood strategies towards more positive and pro-environmental ones. Therefore, through the mediating role of environmental identity, it can enhance smallholders’ accurate knowledge of the objectives and regulatory level of formal environmental regulation, avoid the information asymmetry in the implementation of environmental regulation that reduces the identification and understanding of formal environmental regulation, and circumvent the emergence of smallholders’ livelihood strategies of adverse selection(Huang⁵¹, and the environmental regulation of adverse choice and the boomerang effect of environmental regulation³¹. In addition, EI theory suggests that environmental identity salience is related to the likelihood of its activation in an individual’s thinking⁴³, and that environmental identity salience is more advantageous than environmental identity strength in predicting environmentally relevant behaviors²⁴. Indeed, informal environmental regulation is formed in more stable folkways, where smallholders’ environmental identity is more likely to be activated and more likely to trigger consistent behavior. Thus, environmental identity salience plays a significant mediating role between informal environmental regulation and livelihood strategies.

(3) Regardless of whether eit or eia was used as the mediating variable, self-efficacy fully mediated the relationship between formal environmental regulation and smallholder livelihood strategies, while self-efficacy partially mediated the relationship between informal environmental regulation and livelihood strategies. Self-efficacy is a person’s judgment and perception of their ability to achieve or accomplish an action³⁷, smallholder farmers facing environmental regulatory constraints will be more active and proactive in their livelihood strategy choices if they are able to utilize their existing resources and technologies to meet the goals and requirements of the environmental regulations, and if they have the confidence and ability to adapt to agri-environmental policies to make changes⁴⁴. In order to further enhance the self-efficacy of smallholder farmers, the construction of the environmental governance system should emphasize, firstly, increasing the implementation of social information programs, clarifying the objectives and possible environmental outcomes of environmental regulation policies, and eliminating the negative behaviors of smallholder farmers. Studies have shown that effective environmental regulatory policy information dissemination and increased environmental information disclosure can maximize the impact on smallholder farmers’ expected behavior, reduce the environmental regulatory compliance costs of smallholder farmers’ livelihood strategies and improve the effectiveness of environmental regulation, and further enhance the synergistic effects of environmental regulatory interventions in their livelihood strategies (Huang^515,32,;. Secondly, the training of smallholders in various new technologies should be strengthened. The government or research institutions and other organizations should go deep into the fields of small farmers to effectively understand the technical needs of small farmers in production, and promote the adoption of new environmental technologies with easy-to-understand training methods. The adoption of new technologies can significantly improve the self-efficacy of small farmers. Undoubtedly, smallholders with high self-efficacy believe that they are more challenging, perceive environmental uncertainty, are more motivated to learn and adopt new technologies, and are more able to adopt pro-environmental livelihood strategies³⁷.

(4) Environmental identity and self-efficacy played parallel and chained multiple mediating roles between heterogeneous environmental regulation and smallholder livelihood strategies, as shown by the fact that the mediating paths of environmental identity strength, salience, and self-efficacy were more important than the chained mediating paths in the impact of heterogeneous environmental regulation on smallholder livelihood strategies. The results suggest that synergistically enhancing smallholder environmental identity and self-efficacy can both increase the acceptability of heterogeneous environmental regulation and facilitate the choice of smallholder more pro-environmental livelihood strategies in agricultural practices. On the one hand, enhancing smallholders’ value recognition and self-identity of formal and informal environmental regulatory tools and strengthening the acceptability of environmental regulatory policies are effective paths for environmental regulation to achieve desired implementation effects³. On the other hand, giving full play to the effect of learning by doing, which is an effective way to enhance the self-efficacy of smallholders. Environmental regulation not only imposes a certain degree of constraints on the livelihood behavior of smallholders, but also brings dynamic opportunities for them to choose more pro-environmental livelihood strategies. The promotion of appropriate new technologies, skills and information on off-farm employment in agricultural practices, and the ability and confidence of smallholders to engage more pro-environmental livelihood by learning new skills in agricultural practices, are key to seizing these opportunities. Finally, mutual substitution and transformation between formal and informal environmental regulation⁵. Integrating formal environmental regulation with local folk rules and customs⁶, synergistically exerting the punishment and incentive mechanism of formal environmental regulation as well as the value orientation and transmission and internalization mechanism of informal environmental regulation⁷, to build a good environmental regulation system for smallholder positive utilization livelihood strategies.

In summary, the theoretical framework of the model suggests that heterogeneous environmental regulation indirectly influences smallholder livelihood strategy behavior through environmental identity and self-efficacy. This is because environmental identity and self-efficacy increase the acceptability of environmental regulation, confidence and ability to comply with environmental regulation, and thus promote smallholder farmers to choose more positive and pro-environmental livelihood strategies.

Limitations and future study

First, the use of one-period research data for this study limits further exploration of smallholder livelihood dynamics. Future research could consider using panel data from multiple periods to test for multiple mediation effects. Second, the main variables were collected by smallholders answering questionnaires, which is limited by time, place and scenario and may have some impact on the validity of this study. Thirdly, the findings of this study may be subject to some changes due to the heterogeneity of the smallholder farmers themselves and the culture of the geographic area in which they are located. Future studies may consider adding more smallholders’ own characteristic variables as well as different cultural contexts to further test the heterogeneity of the findings across different types of smallholders and cultural contexts.

Despite these limitations, the present study achieved some unique insights in terms of theory and practice. Theoretically, this study clarifies the relationship between heterogeneous environmental regulation and smallholders’ livelihood strategies, which enriches empirical evidence for research on pro-environmental behavior. By decomposing environmental identity into two dimensions—intensity and salience—and incorporating self-efficacy, this research refines the integrated application of relevant theories, deepens the understanding of the micro-level mechanisms of environmental regulation, and offers a replicable empirical framework for related studies. Practically, it provides targeted pathways for optimizing the portfolio of environmental regulation instruments, promoting the diversified transformation of smallholders’ livelihoods, and building an environmental governance system suitable for small-scale farmers. These implications help achieve a win-win outcome between rural ecological protection and income growth for smallholders, thereby supporting the comprehensive advancement of rural revitalization.

Conclusion

Theoretically, this study clarifies the relationship between heterogeneous environmental regulation and smallholders’ livelihood strategies, which enriches empirical evidence for research on pro-environmental behavior. By decomposing environmental identity into two dimensions—intensity and salience—and incorporating self-efficacy, this research refines the integrated application of relevant theories, deepens the understanding of the micro-level mechanisms of environmental regulation, and offers a replicable empirical framework for related studies. Practically, it provides targeted pathways for optimizing the portfolio of environmental regulation instruments, promoting the diversified transformation of smallholders’ livelihoods, and building an environmental governance system suitable for small-scale farmers. These implications help achieve a win-win outcome between rural ecological protection and income growth for smallholders, thereby supporting the comprehensive advancement of rural revitalization.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

Conceptualization, Z.Y.; Data curation, X.L.; Funding acquisition, Z.Y.; Investigation, X.L.; Methodology, Z.Y.; Software, Z.Y. and X.L.; Supervision, Z.Y.; Writing – original draft, Z.Y. and X.L.; Writing – review & editing, Z.Y. All authors reviewed the manuscript.

Funding

This work was supported by the National Social Science Fund of China (Grant No. 20BJY129).

Data availability

The data used in this study were derived from a survey of 804 farmers conducted in four provinces of China, namely Heilongjiang, Hunan, Jiangxi, and Sichuan. The survey data were collected through face-to-face interviews and structured questionnaires. Due to privacy and ethical considerations, the raw data are not publicly available. However, aggregated and anonymized data supporting the findings of this study can be made available upon reasonable request to the corresponding author.

Declarations

Competing interests

The authors declare no competing interests.