Easement Reform and Employment Transfer of Forest Farmers: Evidence from China’s National Parks

Abstract

The easement reform of collective forest land (ERCFL) is an important part of national park system construction, which plays an important role in promoting the employment transfer of forest farmers. On the basis of survey data from forest farmers in Qianjiangyuan–Baishanzu National Park in Zhejiang Province, this paper uses the propensity score matching-difference in differences model to empirically analyze the impact of ERCFL on the transfer employment behavior and employment quality of forest farmers to provide experience for promoting the employment transfer of forest farmers and improving the ERCFL in national parks. This research shows the following: (1) The ERCFL in national parks has significantly promoted the employment transfer and improved the employment quality of forest farmers. This effect remains significant when controlling for possible endogeneity issues. (2) Three mechanisms of the ERCFL affect forest farmers’ transfer of employment, namely improving non-agricultural employment skills, expanding the scope of non-agricultural employment, and increasing non-agricultural employment opportunities. (3) Significant gender, age, and education differences exist in terms of the impact of the ERCFL on the employment transfer of forest farmers. The impact on men, middle-aged and elderly groups, and low-education groups is more significant. Finally, this paper proposed forward countermeasures and suggestions to promote the employment transfer of forest farmers.

1. Introduction

Easement is a legal system that originated in Ancient Rome. It refers to the right of the easement demander to use other people’s real estate, usually land, as agreed upon in a contract to enhance the value of their own real estate [1]. Other people’s real estate is the servitude land and one’s own real estate is the dominant land. Essentially, easement is a type of usufructuary right where the non-owner (the easement demander) has the right to use, control, and benefit from the immovable property owned by another (the easement provider). The United States inherited the relevant provisions from Roman law. In 1981, the U.S. Government enacted the Uniform Conservation Easement Act (UCEA), which established a legal system for protecting easements through legislation, clarifying public purposes such as the protection of natural resources and the ecological environment. China, in turn, drew on the UCEA and enacted the Property Law in 2007, which discussed easements and clarified the subjects and establishment rules of easements.

Collective forest land easement is an important component of easements. In the context of collective forest land easement, the servient estate refers to the collectively owned public welfare or commercial forest land, while the dominant estate refers to another’s public welfare or commercial forest land. The entity responsible for the servient estate is designated as the easement provider and the entity responsible for the dominant estate is designated as the easement demander. The easement demander and provider can establish a collective forest land easement according to contractual agreements. Once the collective forest land easement is established, the easement demander can utilize the easement provider’s forest land to enhance the value of his own forest land. In China, the establishment of collective forest land easements is closely related to the construction of national parks and is of great significance to enhancing the influence of China’s forestry in the world [2].

Beginning in 2015, China launched the system reform of the national parks and has achieved remarkable achievements in management system innovation, ecological protection, and community integration and development [3]. To promote the construction of ecological civilization, China formulated an overall plan for establishing a national park system in 2017. The plan pointed out that the construction of national parks should take into account the development needs of the surrounding economy and society and give priority to providing jobs for the local residents while protecting the ecological environment. In policy practice, Qianjiangyuan–Baishanzu National Park, Wuyishan National Park, Shennongjia National Park, and other national parks all have given priority to providing ecological management and social service public welfare positions such as patrolling, management, sanitation, and interpretation for the local residents, as well as employment support measures such as franchising and supporting characteristic industries [4].

The easement reform of collective forest land (ERCFL) is an important component of the national park system construction. Qianjiangyuan National Park, located in Kaihua County, Zhejiang Province, is the first national park to implement the ERCFL in China. The pilot program for Qianjiangyuan National Park was initiated in 2016. The Qianjiangyuan National Park Management Committee was established in 2017 and was restructured as the Qianjiangyuan National Park Administration in 2019. With the expansion of the pilot area, Qianjiangyuan–Baishanzu National Park was officially established in 2020. Regarding the ERCFL, in 2018, Zhejiang Province issued the first national easement certification for collective forest land in Qianjiangyuan National Park, which clearly defined the rights and obligations of both easement providers and demanders, marking the official launch of the ERCFL (

Table 1. Rights and obligations of participants in national park easement reform.

Participants	Rights	Obligations
Easement provider (Forest farmers)	1. Free visit and tour. 2. Priority for operating franchise projects related to ecological agriculture, ecological experiences, and recreation. 3. Free use of the national park branding.	1. Strict compliance with national park management regulations. 2. Assisting in scientific research and national park management. 3. Prohibition of environmental damage. 4. Monitoring activities that harm forest resources. 5. Increasing awareness of ecological conservation. 6. Cooperating in land easement registration.
Easement demander (National park administration)	1. Legally entitled to collective forest land easement rights. 2. Implementing management in accordance with relevant laws and regulations for national park construction. 3. Carrying out necessary protection and scientific monitoring. 4. Applying appropriate human intervention to trees and forest land according to conservation needs.	1. Conducting employment skills training. 2. Providing ecological conservation public welfare positions. 3. Assisting in community environmental improvement. 4. Disclosing relevant information about the national park. 5. Paying land easement compensation (CNY 48.2/mu per year) in accordance with the contract.

). In 2021, Zhejiang Province formulated the work plan for easement reform of public welfare forest land and collective forest land, which further clarifies the main tasks and work directions for the ERCFL.

Through a series of institutional arrangements, such as providing employment skills training services, providing public welfare positions for ecological management, and allowing residents to enjoy the priority of ecotourism franchises, the purposes of ERCLF in national parks are to effectively revitalize forest ecological resources and protect the ecological environment; promote the employment transfer and increase the income level of forest farmers; and to strengthen the village collective economy and realize the coordination of the interests of the government, the village collective, and forest farmers, thus achieving a win–win situation for the economy, ecology, and society. Through the ERCLF, the ecological environment in the pilot area has been significantly improved, the collective forest land has achieved unified and centralized management, and the average income of farmers has increased by more than CNY 4000.

The ERCFL is of great significance in promoting the employment transfer of forest farmers and realizing the transformation and development of forestry. Scholars have studied the impact of the collective forest tenure system reform on the employment transfer of forest farmers, but research has not drawn consistent conclusions and has not paid enough attention to easement reform. For example, studies have shown that the reform of the collective forest tenure system has promoted the employment transfer of forest farmers by clarifying property rights [5], providing forestry subsidies [6], and increasing non-agricultural employment opportunities [7]. However, some studies have found that the reform of the collective forest tenure system may also harm the employment transfer of forest farmers [8,9], mainly due to the lack of employment skills of forest farmers and incomplete employment guarantee policies.

An important reason for the differences above is that the existing research has not well-distinguished the different policies of the collective forest tenure system reform [10]. As an important part of the reform of the collective forest tenure system, the easement reform pays more attention to the transferred employment of forest farmers. From the perspective of sustainable livelihoods, forest farmers have a high level of acceptance in regard to the ERCFL [11]. With the practice and development of the ERCFL in national parks, the impact of this policy on the employment transfer of forest farmers needs to be explored further. In addition, the use of an oversimplified research method, such as simple linear regression, is another important reason for the discrepancies in the above results. With the improvement of research, methods such as difference-in-difference (DID), synthetic control method, regression discontinuity design, and propensity score matching (PSM) are widely used in agricultural policy evaluation, which can effectively improve the accuracy of ERCFL policy evaluation [12].

Qianjiangyuan–Baishanzu National Park in the Zhejiang Province is a typical example of the ERCFL in national parks in China. It is located in the collective forest area in southern China, and it involves complex forest ownership, high proportion of collective forests, and prominent problems in regard to forest fragmentation management [13]. By concluding an easement contract with forest farmers in 2018, this national park established a scientific and reasonable easement compensation mechanism and forest land co-management mechanism without changing the land ownership, then realized the unified, standardized, and efficient protection and management of natural resources. Furthermore, it has had a great impact on the employment transfer of forest farmers.

Therefore, this paper takes Qianjiangyuan–Baishanzu National Park in Zhejiang Province as its research object and theoretically and empirically analyzes the impact of the ERCFL on forest farmers’ employment transfer behavior and employment quality. The implementation of this study can provide experience for promoting the employment transfer of forest farmers and improving the ERCFL in national parks, offering a theoretical and practical basis for forest rights system reform in developing countries.

2. Theoretical Analysis

According to labor migration theory, labor transfer employment is affected by multiple factors, and institutional reform is an important factor [14]. The ERCFL in national parks has both direct and indirect effects on the employment transfer behavior and employment quality of forest farmers, as shown in Figure 1. The following impact will be analyzed below in terms of two aspects: direct and indirect impact.

2.1. Easement Reform and Employment Transfer of Forest Farmers

For traditional forestry production activities, forest land is the basic means of production for forest farmers to survive and maintain their livelihoods. A strong dependence of forest farmers on forest land corresponds to an obvious endowment effect and, thus, a small possibility of transferring employment. The ERCFL restricts some rights of forest farmers to use the forest land through system design. For example, farmers are not allowed to reclaim and excavate forest land, cut down and damage forest trees, transfer forest land, and perform other acts that may damage the ecological environment. To make up for the loss of income caused by the limitation of traditional forestry production activities, forest farmers have a strong willingness to transfer employment [15]. In addition, national parks, as the easement demanders, provide a series of assistance measures to promote the employment of local residents, such as increasing public welfare jobs and conducting non-agricultural employment skills training, which weakens forest farmers’ dependence on forest land, thus promoting the employment transfer of forest farmers [16]. Therefore, research hypothesis H1 is proposed.

H1: 

The ERCFL in national parks can help promote the employment transfer of forest farmers.

2.2. Easement Reform and Employment Quality of Forest Farmers

The ERCFL in national parks will not only promote the employment transfer of forest farmers but also affect the employment quality. National parks, as the easement demanders, need to provide cash compensation to the ecological environment provider, i.e., forest farmers, and are obliged to provide corresponding supporting measures such as employment, skill training, and industrial development. The ERCFL not only realizes ecological and environmental benefits but also promotes the diversified development of regional industries, broadens the employment scope of forest farmers, increases their income, and realizes the dual goals of ecological protection and income increases for forest farmers [17]. Fortunately, the positions provided by national parks, such as ecological protection, recreational services, and sanitation, are relatively stable and sustainable. Unlike in traditional forestry production, which has a high investment risk and low return rate [18], these positions have clear advantages in stabilizing employment, increasing income, and improving the employment quality of forest farmers [19,20]. Therefore, research hypothesis H2 is proposed.

H2: 

The ERCFL in national parks can improve the employment quality of forest farmers.

2.3. Indirect Effect of ERCFL

2.3.1. Improve Employment Skills of Forest Farmers

The implementation of employment skills training is a major obligation of national parks. It is an important means to promote employment transfer and help improve the quality of employment transfer of forest farmers [21]. By providing relatively complete employment skills training services, the ERCFL in national parks has alleviated the constraints of lack of knowledge and limited vocational skills in forest farmers under the traditional forestry production model [9,22], which is conducive to improving the livelihood ability of forest farmers, to better promote forest farmers to engage in non-agricultural work related or unrelated to forestry [23,24]. Therefore, research hypothesis H3 is proposed.

H3: 

Easement reform can help promote employment transfer of forest farmers by improving employment skills.

2.3.2. Broaden the Scope of Off-Farm Employment

The construction of national parks has promoted the development of ecological economies and optimized industrial structures in forest areas. Due to the establishment of the ERCFL in national parks, forest farmers are given operational priority in franchise projects such as ecological agriculture, ecological experience, and recreation. Those engaged in forestry-related work can find employment at their doorstep, thereby improving their sense of employment gain [25]. For those engaged in non-forestry work, the easement reform has weakened the dependence of forest farmers on forest land. Therefore, forest farmers can expand the scope of employment from rural areas to counties and provinces, and the level of non-agricultural income will also increase accordingly [26]. Therefore, research hypothesis 4 is proposed.

H4: 

Easement reform can help promote the employment transfer of forest farmers by expanding the scope of non-agricultural employment.

2.3.3. Increase Non-Agricultural Employment Opportunities

The non-agricultural employment opportunities is an important factor that affects the employment transfer of forest farmers. The ERCFL in national parks has provided forest farmers with direct public welfare positions such as patrolling and management. It has also indirectly promoted the development of forest tourism, homestay, catering, and other industries, thereby increasing non-agricultural employment opportunities [27,28]. Compared with traditional forestry production activities, the new non-agricultural jobs brought about by the ERCFL in national parks have greater economic resilience and more diversified and stable employment forms. Therefore, research hypothesis H5 is proposed.

H5: 

Easement reform can help promote the employment transfer of forest farmers by increasing non-agricultural employment opportunities.

3. Data and Models

3.1. Data Sources

The research data come from a survey of forest farmers in Qianjiangyuan–Baishanzu National Park in Zhejiang Province, China. Qianjiangyuan–Baishanzu National Park is composed of the Qianjiangyuan and Baishanzu areas. The Qianjiangyuan area is located in Kaihua County, with a total area of 252 square kilometers. The Baishanzu area is located in Lishui City, covering the three counties of Longquan, Qingyuan, and Jingning, with a total area of 505 square kilometers.

The survey sites are Kaihua County and Longquan City, which are key forest counties in Zhejiang Province. Both places are rich in forest resources, and the proportion of collective forests is as high as 80%. The survey took place from May to October 2022, and the survey objects include relevant persons in charge of county forestry bureaus and township forestry stations, as well as forest farmers. The survey is supported by forest farmers and approved by the Ethics Committee of ZAFU (No. Z2205015).

The survey of forest farmers adopts a stratified random sampling method. According to the implementation of the easement reform, we selected eight towns in Longquan City, six towns in Kaihua County, and three villages in each township, a total of forty-two villages. Among these villages, 22 villages participated in the easement reform and 20 villages did not. Then, we randomly selected 10 forest farmers in each village and conducted a questionnaire survey on a total of 420 forest farmers. The questionnaire survey was conducted in the form of face-to-face interviews. The survey content included questions about forest farmers’ employment status, family status, forest land production characteristics, participation in easement reform, and a cognitive evaluation. Considering that the easement reform started in 2018, the investigators asked in detail about the basic situation of forest farmers in 2017, before the easement reform, and in 2021, after the reform. To obtain accurate information regarding the employment transfer of forest farmers, we provided training to the investigators and made data corrections based on interviews with village cadres. The data can be considered relatively reliable. Excluding invalid questionnaires with missing key information, we finally obtained 381 valid sample questionnaires. Figure 2 shows the sample distribution.

3.2. Variable Selection

Explained variable: Employment transfer behavior and employment quality of forest farmers. Variables were quantified with reference to Salifu and Horlu [29] and Arranz et al. [30]. For the variable of employment transfer behavior, a value of 0 is assigned to those who are mainly engaged in agriculture and a value of 1 is assigned to those mainly engaged in non-agriculture. The variable of employment quality is represented by non-agricultural income and employment stability. Employment stability is measured by whether an employment contract is signed. The value is 1 if an employment contract is signed, and it is 0 otherwise.

Explanatory variable: Easement reform on collective forest land in national parks. It is represented by the product of two dummy variables, i.e., whether it is a reform village or not and the time of reform implementation. Villages with easement reforms are assigned a value of 1; those without are assigned a value of 0. The post-reform 2021 is assigned a value of 1, and the pre-reform 2017 is assigned a value of 0. The product of the two dummies is 1 for reformed villages in 2021 and 0 for others.

Mediator variables. According to the theoretical analysis framework, the indirect impact of the ERCFL on the employment transfer of forest farmers was tested by constructing the three following mediator variables: improving non-agricultural employment skills, broadening the scope of non-agricultural employment, and increasing non-agricultural employment opportunities. The proxy variables are participation in vocational skills training, employment location, and the number of people in the family engaged in non-agricultural employment.

Control variables. In reference to the studies of Deininger et al. [31] and Leight [32], household characteristics, forestry production characteristics, and county economic level were controlled.

Household characteristics include the gender, age, and education level of the household head; whether the household head has served as a village cadre; the total family population; and the population burden coefficient. Previous studies found that the gender, age, and education level of the household head are important factors that affect labor transfer. Young and well-educated male farmers are more willing and more likely to transfer. In addition, household heads with experience as village cadres usually have more social resources and are more likely to transfer [33]. However, family size and population burden will hinder their labor transfer [34].

The characteristics of forestry production include the area of forest land, the proportion of public welfare forest area, and the number of forest land plots. Previous studies found that land resource endowment, such as more forest area, public welfare forest area, and forest plots, will hinder labor transfer [35] because these forest farmers may gain more benefits from forestry than from non-agricultural employment.

The level of regional economic development is also an important factor affecting labor transfer [36]. The economic level of the county is controlled and represented by per capita GDP. Variables are detailed in

Table 2. Variable description and descriptive statistics.

Variables		Description	Unit/Assignment	Min	Max	Mean	Standard Deviation
Explained variables
Employment transfer behavior	y1	Type of employment	0 = Agricultural, 1 = Non-agricultural	0	1	0.741	0.438
Quality of transfer employment	y2	Non-agricultural income, logarithm	CNY 1000	0.770	2.334	1.678	0.302
	y3	Stability of employment	0 = No contract, 1 = With contract	0	1	0.160	0.367
Explanatory variable
ERCFL in national parks	did	Treat × Time	Treat = 1, if reform village; treat = 0, others.Time = 1, if after reform; time = 0, others	0	1	0.272	0.445
Mediator variables
Improving non-agricultural employment skills	m1	Whether participated in vocational skills training	0 = No, 1 = Yes	0	1	0.252	0.434
Broadening the scope of non-agricultural employment	m2	Employment location	1 = local village; 2 = local county; 3 = outside the county	1	3	1.764	0.735
Increasing non-agricultural employment opportunities	m3	Number of people in non-agricultural employment in the family	No.	0	5	1.904	1.231
Control variables
Household characteristics	x1	Gender	0 = Female, 1 = Male	0	1	0.609	0.488
	x2	Age	Years	26	89	50.73	9.398
	x3	Education	Years	0	20	8.297	3.636
	x4	Whether served as a village cadre	0 = No, 1 = Yes	0	1	0.324	0.468
	x5	Total family population	No.	1	11	4.400	1.485
	x6	Population burden coefficient	%	0	100	30.71	29.27
Forestry production characteristics	x7	Area of forest land	Mu	0	270	26.72	36.10
	x8	Proportion of public welfare forest area	%	0	100	46.91	40.82
	x9	Number of forest land plots	No.	1	60	4.528	5.555
County economic level	x10	GDP per capita	CNY 1000	41.27	58.26	49.238	6.161

Source: According to the survey data of forest farmers.

.

On the basis of descriptive statistics, the majority of forestry farmers are primarily engaged in non-agricultural employment, accounting for 74.1%, with an average income of CNY 5.35 thousand from non-agricultural jobs. However, the stability of non-agricultural employment is relatively poor, with only 16% having signed contracts. The proportion of forestry farmers participating in vocational skills training is 25.2% and the average number of people in non-agricultural employment in the family is 1.9 people. The employment scope is mainly within the county. With regard to individual characteristics, the average household size of forestry farmers is 4.4 people, and the population burden coefficient is 0.3. Household heads are predominantly male, accounting for 60.9%, with an average age of 50.7 years and an average education level of 8.3 years. In addition, 32.4% have experience as village cadres. In terms of forest land resources, the average forest land area per forestry household is 26.7 mu, with public welfare forest accounting for 46.9%, and the forest land is divided into 4.5 plots. The average per capita GDP in the county is CNY 49.2 thousand.

3.3. Model Specification

DID methods are widely used in policy evaluation [37]. This paper uses the DID method to study the impact of the ERCFL in national parks on the employment transfer of forest farmers. The general model is as follows:

$$Y_{it}={\beta }_0+{\beta }_{1}treat\times time+{\sum }_j{\gamma }_j{x}_{jit}+{\alpha }_i+{\mathsf{\delta }}_{\mathrm{t}}{+\epsilon }_{it}$$

(1)

In Equation (1), the explained variable $Y_{it}$ represents the employment transfer behavior and employment quality. $i$ represents forest farmers, and $t$ represents time. $did$ is the key explanatory variable, $did=treat\times time$, which is the interaction term between the village dummy variable (indicating whether the easement reform is implemented) and the time dummy variable (before and after the easement reform). $x$ is the control variables. ${\alpha }_i$ and ${\mathsf{\delta }}_{\mathrm{t}}$ are the individual and time fixed effects. $\beta ,\gamma$ are the parameters to be estimated, and ${\epsilon }_{it}$ is the residual item.

Considering the endogenous problems that may be caused by sample selection bias, this paper adopts the PSM-DID method proposed by Heckman et al. (1997) [38] for robustness testing. The basic idea of this method is that if the self-selection bias depends on the observable characteristics of forest farmers, then we can find those who have similar characteristics but did not participate in the easement reform as a “counterfactual” control group and then use the DID model to obtain the net impact. The key to this method is to select forest farmers whose characteristics are as similar as possible to the treatment group from the control group as a matching group and then compare them. The propensity score calculation equation is

$$P_i\left(\mathrm{X}\right)=P_r(Z_{it}=1|X_i)=\mathrm{l}\mathrm{o}\mathrm{g}\mathrm{i}\mathrm{t}(f\left(X_i\right))$$

(2)

In Equation (2), $X_i$ is the characteristic variable of forest farmer $i$ and $Z_{it}$ is the dummy variable of the treatment group. After calculating the propensity score for the forest farmers in the treatment group, an individual with a similar propensity score from the control group is identified as a match. At this time, the matched forest farmers and the corresponding treatment group have similar probabilities of participating in the easement reform. The treatment group and the control group meet a common trend. After matching, the matched treatment group and control group are used for DID estimation. The PSM-DID model is as follows:

$$Y_{it}^{psm}={\beta }_0+{\beta }_{1}treat\times time+{\sum }_j{\gamma }_j{x}_{jit}+{\alpha }_i+{\mathsf{\delta }}_{\mathrm{t}}+{\epsilon }_{it}$$

(3)

In Equation (3), ${\beta }_1$ is an important parameter to estimate. The meanings of other variables are the same as in Equation (1).

To further verify the indirect impact of the ERCFL in national parks on forest farmers’ transfer employment behavior and employment quality, according to the previous theoretical analysis and mediation effect analysis method, the model is set as follows:

$$M_{it}={\rho }_0+{\rho }_{1}treat\times time+{\sum }_j{\gamma }_j{x}_{jit}+{\alpha }_i+{\mathsf{\delta }}_{\mathrm{t}}{+\epsilon }_{it}$$

(4)

$$Y_{it}={\theta }_0+{\theta }_{1}treat\times time+{\sum }_j{\gamma }_j{x}_{jit}+{\theta }_2{M}_i+{\alpha }_i+{\mathsf{\delta }}_{\mathrm{t}}{+\epsilon }_{it}$$

(5)

In Equations (4) and (5), $M$ is the mediator variables, including improving non-agricultural employment skills, broadening the scope of non-agricultural employment, and increasing non-agricultural employment opportunities. $\rho$ and $\theta$ are the parameters to be estimated, and the meanings of other variables are the same as in Equation (1).

4. Empirical Results and Analysis

4.1. Benchmark Regression

Table 3. Benchmark regression results.

Variables	Model 1	Model 2	Model 3
did	0.1508 **	0.1861 ***	0.2736 ***
	(0.0632)	(0.0415)	(0.0564)
x1	0.0656 *	0.0506 **	0.0543 *
	(0.0355)	(0.0244)	(0.0314)
x2	−0.0057 ***	−0.0059 ***	−0.0077 ***
	(0.0018)	(0.0013)	(0.0016)
x3	0.0263 ***	0.0316 ***	0.0150 ***
	(0.0047)	(0.0029)	(0.0041)
x4	0.1337 ***	0.1473 ***	0.0977 ***
	(0.0273)	(0.0196)	(0.0333)
x5	0.0106	0.0198 ***	−0.0049
	(0.0092)	(0.0063)	(0.0071)
x6	−0.0019 ***	0.0001	0.0005
	(0.0005)	(0.0003)	(0.0004)
x7	−0.0008 *	−0.0000	−0.0001
	(0.0004)	(0.0002)	(0.0003)
x8	−0.0005	−0.0005 **	0.0005 *
	(0.0003)	(0.0002)	(0.0003)
x9	−0.0063 ***	−0.0029	0.0011
	(0.0024)	(0.0028)	(0.0022)
x10	0.0057	0.0298	−0.0370
	(0.0369)	(0.0249)	(0.0350)
Constant	0.3466	0.0267	2.0388
	(1.7679)	(1.1906)	(1.6687)
Individual fixed effects	Control	Control	Control
Time fixed effects	Control	Control	Control
R2	0.2880	0.4760	0.2400

Notes: *, ** and *** denote significance at 10%, 5% and 1% levels, respectively. Standard errors in parentheses.

shows the estimation results of the impact of the easement reform on the employment transfer of forest farmers. Models 1, 2, and 3 are the estimation results of forest farmers’ transfer employment behavior, non-agricultural income, and employment stability, respectively. The results show that, in the three models, the estimation results of the did variable are all significantly positive, indicating that the ERCFL in national parks has significantly promoted the employment transfer of forest farmers and improved the quality of employment. The estimation results are consistent with theoretical expectations. Thus, hypotheses H1 and H2 are verified.

Specifically, in model 1, after the ERCFL in national parks, the probability of non-agricultural employment of forest farmers increased by 15.08%, indicating that the easement reform significantly promoted the employment transfer behavior of forest farmers. In model 2, after the easement reform, the non-agricultural income of forest farmers increased by 18.61%, indicating that the easement reform significantly promoted the non-agricultural income of forest farmers, thereby promoting the employment quality. In model 3, after the easement reform, the stability of non-agricultural employment of forest farmers increased by 27.36%, indicating that the easement reform significantly promoted the stability of non-agricultural employment of forest farmers, thereby improving the employment quality.

Table 3 shows that education (X3) and village cadre experience (X4) have a significant positive impact on the employment transfer behavior of forest farmers, while age (X2) has a significant negative impact. Forest land resource endowment, such as forest land area (X7) and the proportion of public welfare forest area (X8), also has a significant negative impact on forest farmers’ employment transfer behavior.

4.2. Model Test

4.2.1. PSM-DID Estimation

According to Equations (2) and (3), the samples are matched first, and then the DID model is used to estimate the result. After the propensity score was calculated, a density function graph was further drawn to test the matched common support domain and thus ensure the matching quality, as shown in Figure 3. The figure shows that the propensity scores of the matched sample participating in easement reform and the non-participating sample have a large range of overlap, and most of the observed values are within the common value range.

Table 4. Balance test of explanatory variables before and after PSM.

Sample Size	Employment Transfer Behavior			Non-Agricultural Income			Employment Stability
	Pseudo R2	LR Statistics	Standardized Deviation (%)	Pseudo R2	LR Statistics	Standardized Deviation (%)	Pseudo R2	LR Statistics	Standardized Deviation (%)
Before matching	0.206	216.52	33.6	0.202	174.21	34.6	0.206	216.52	33.6
After matching	0.040	46.01	13.1	0.045	44.66	14.5	0.040	46.01	13.1

shows the balance test of covariates. The pseudo R², LR statistics, and standardized deviation values all decreased after matching, indicating that the propensity score matching method can effectively reduce the difference in the distribution of explanatory variables between the control group and the treatment group and eliminate estimation bias due to sample selection.

Table 5. Estimation results of the PSM-DID model.

Variables	Model 4	Model 5	Model 6
did	0.1430 **	0.1908 ***	0.2728 ***
	(0.0635)	(0.0415)	(0.0565)
Constant	0.05871	−0.2798	2.0495
	(1.7692)	(1.1940)	(1.6748)
Control variables	Control	Control	Control
Individual fixed effects	Control	Control	Control
Time fixed effects	Control	Control	Control
R2	0.2790	0.0475	0.2380

Notes: ** and *** denote significance at 5%, and 1% levels, respectively. Standard errors in parentheses.

shows the estimation results of the PSM-DID model. After PSM is used to control the endogenous problems that may be caused by sample selection bias, the impact of the ERCFL in national parks on forest farmers’ transfer employment behavior and employment quality is still significantly positive. The estimation results have strong robustness.

4.2.2. Placebo Test

The placebo test was carried out by randomizing the treatment group. Specifically, among all the samples, the same number of treatment groups was randomly selected as the pseudo-treatment group for the placebo test. It was used to generate interaction items with time dummy variables and then regressed. Random sampling was repeated 500 times. If the estimated coefficient is significant, then bias may exist in the original estimation results. Figure 4 shows that most of the regression coefficient estimates are concentrated around zero, thus suggesting that other random factors have little influence and the regression results are reliable.

4.3. Analysis of Impact Mechanism

According to Equations (1), (4) and (5), the mediation effect analysis method is adopted to test the indirect impact of the ERCFL in national parks on the employment transfer of forest farmers. The results validated hypotheses H3, H4, and H5.

4.3.1. Mediating Effect of Forest Farmers’ Employment Transfer Behavior

Table 6. Mediating effect test of employment transfer behavior.

Mediator Variable	Improving Non-Agricultural Employment Skills	Broadening the Scope of Non-Agricultural Employment	Increasing Non-Agricultural Employment Opportunities
Sobel test (Z statistic)	3.0990	5.9270	8.9550
Direct effect	0.2325	0.1765	0.0526
Mediating effect	0.0494	0.1054	0.2293
Total effect	0.2819	0.2819	0.2819
Mediation effect proportion	0.1753	0.3740	0.8136

shows the estimation results of the mediating effect of forest farmers’ employment transfer behavior. A significant intermediary effect occurs with the impact of the easement reform on the employment transfer behavior of forest farmers. The Z statistics in the Sobel test are 3.099, 5.927, and 8.955, all of which are greater than the critical value of 0.97 at the 5% significance level. Therefore, intermediary effects occur based on improving non-agricultural employment skills, broadening the scope of non-agricultural employment, and increasing non-agricultural employment opportunities. The proportion of each mediation effect is 17.53%, 37.40%, and 81.36%, respectively.

4.3.2. Mediating Effect of Forest Farmers’ Transfer Employment Quality

Similar to the results of the mediation effect analysis of employment transfer behavior, the easement reform also has a significant mediation effect on the quality of forest farmers’ transfer employment. In terms of non-agricultural income, the Z statistics in the Sobel test are 3.07, 8.63, and 6.74, respectively, all of which are greater than the critical value at the 5% significance level, indicating a significant mediating effect. The proportions of each mediation effect are 18.66%, 56.87%, and 47.35%, respectively (

Table 7. Mediating effect test of non-agricultural income.

Mediator Variable	Improving Non-Agricultural Employment Skills	Broadening the Scope of Non-Agricultural Employment	Increasing Non-Agricultural Employment Opportunities
Sobel test (Z statistic)	3.0700	8.6320	6.7470
Direct effect	0.1772	0.0940	0.1147
Mediating effect	0.0407	0.1239	0.1032
Total effect	0.2179	0.2179	0.2179
Mediation effect proportion	0.1866	0.5687	0.4735

). For employment stability, the Z statistics in the Sobel test are 3.160, 4.102, and 3.851, respectively, all of which are greater than the critical value at the 5% significance level. A significant mediating effect exists, and the proportions of each mediating effect are 22.52%, 26.60% and 32.16%, respectively (

Table 8. Mediating effect test of employment stability.

Mediator Variable	Improving Non-Agricultural Employment Skills	Broadening the Scope of Non-Agricultural Employment	Increasing Non-Agricultural Employment Opportunities
Sobel test (Z statistic)	3.1600	4.1020	3.8510
Direct effect	0.1488	0.1410	0.1303
Mediating effect	0.0432	0.0510	0.0617
Total effect	0.1920	0.1920	0.1920
Mediation effect proportion	0.2252	0.2660	0.3216

).

4.4. Heterogeneity Analysis

Considering that obvious gender, education, and age differences exist in the labor market [39,40], this paper mainly analyzes the heterogeneity of forest farmers’ gender, age, and education. The results are shown in

Table 9. Heterogeneity analysis results.

	Employment Transfer Behavior		Non-Agricultural Income		Employment Stability
Gender	Male	Female	Male	Female	Male	Female
did	0.3271 ***	0.2483 ***	0.0838 **	0.0734	0.3092 ***	0.0516
	(0.0574)	(0.0708)	(0.0353)	(0.0475)	(0.0452)	(0.0674)
Age	Elder	Young	Elder	Young	Elder	Young
did	0.4068 ***	0.1705 ***	0.0406	0.0413	0.2753 ***	0.1659 ***
	(0.0717)	(0.0566)	(0.0458)	(0.0369)	(0.0482)	(0.0539)
Education	High	Low	High	Low	High	Low
did	0.2109 ***	0.5410 ***	0.0112	0.0877	0.1907 ***	0.2160 ***
	(0.0483)	(0.0876)	(0.0333)	(0.0547)	(0.0510)	(0.0504)

Notes: **, and *** denote significance at 5%, and 1% levels, respectively. Standard errors in parentheses. Young (age < 55), elder (age ≥ 55). High education (≥8 years), low education (<8 years).

. Significant gender, age, and education differences are found in the impact of the easement reform on the employment transfer of forest farmers. The easement reform has a more significant effect on promoting employment and improving employment quality for men, as well as middle-aged, elderly, and low-education groups.

5. Conclusions and Discussion

This paper constructs a theoretical analysis framework for the impact of the ERCFL being employed in national parks on the employment transfer of forest farmers. On the basis of the micro-survey data of forest farmers in Qianjiangyuan–Baishanzu National Park in Zhejiang Province, PSM-DID is used to empirically analyze the impact of easement reform on the employment transfer behavior and employment quality of forest farmers. The main research conclusions are as follows:

First, the ERCFL in national parks has significantly promoted the employment transfer of forest farmers and improved the employment quality. This effect is still significant when considering possible endogeneity problems. Compared with the forest farmers who did not participate in the easement reform, for those participating in the easement reform, the non-agricultural employment probability increased by 14.3%, the non-agricultural income increased by 19.08%, and the employment stability increased by 27.28%.

Second, the ERCFL in national parks has a significant mediating effect on the employment transfer of forest farmers. Easement reform can promote employment transfer and improve the employment quality of forest farmers by improving forest farmers’ non-agricultural employment skills, broadening the scope of non-agricultural employment, and increasing non-agricultural employment opportunities.

Third, significant gender, age, and education differences exist in the impact of easement reform on the employment transfer of forest farmers. The easement reform has a more significant effect on promoting employment and improving the quality of employment for men and middle-aged, elderly, and low-education groups.

On the basis of the above analysis, the following suggestions are proposed to promote the employment transfer and improve the employment quality of forest farmers:

First, the participation of forest farmers in the easement reform of collective forest land in national parks needs to be enhanced. Forest farmers are the easement provider, and their effective participation is crucial for the smooth implementation of the easement reform. Considering the issue of relatively low compensation standards, moderately raising the compensation standards for easement reform is advisable. Actively guiding forest farmers to engage in ecotourism, forest wellness, and other industries and granting them the rights to use the national park brand logo can effectively achieve the integration of forest ecological conservation, rural revitalization in forest areas, and increased income for forest farmers.

Second, formulating and improving policies to promote the transfer of employment for forest farmers is important. In the design of the collective forest land easement system in national parks, it is necessary to enhance policies for the transformation of forest farmers by continuing to increase support for their entrepreneurship and non-agricultural employment. Additionally, differentiated fiscal, financial, and tax support policies should be provided based on the different characteristics of forest farmers to offer more non-agricultural jobs and opportunities.. In addition, establishing a robust employment information platform, improving mechanisms for information sharing and dissemination, and further strengthening non-agricultural employment skills training for forest farmers are also necessary. These steps can assist forest farmers in acquiring the necessary skills and knowledge for non-agricultural work, enhancing their competitiveness in the labor market, and improving the quality of their employment transition.

Existing research primarily focuses on the impact of establishing conservation easements on forest protection, the ecological environment, and biodiversity [41,42]; however, insufficient attention has been given to the socio-economic effects these easements. Some studies, such as that of Wang and Wang (2023), discussed the impact of the establishment of natural reserves on agricultural labor transfer and found that natural reserve construction significantly promoted the transfer of agricultural labor to non-agricultural industries, which is similar to the present study [35]. He and Wei (2023) focused on forest farmers, explored the influence of livelihood assets on forestry farmers’ participation in conservation easements, and proposed that further attention should be paid to the impact of easement reform policies on forest farmers [11]. This study fills this gap by analyzing the impact of easement reforms on the employment transfer of forestry farmers, expands the theory of labor transfer employment, and provides a theoretical basis for the impact of institutional reform on labor transfer employment.

In contrast to existing studies, this paper focuses mainly on the study of socio-economic impacts rather than impacts on the natural environment, such as forest protection, ecology, and biodiversity [41,42]. This study is similar to that of Wang and Wang (2023), who discussed the impact of the establishment of natural reserves on agricultural labor transfer and found that natural reserve construction significantly promoted the transfer of agricultural labor to non-agricultural industries, which is similar to the present study [35]. Focusing on forest farmers, He and Wei (2023) explored the impact of livelihood assets on forest farmers’ participation in conservation easements and suggested that further attention be paid to the impact of easement reform policies on forest farmers [11]. This study fills this gap by analyzing the impact of easement reforms on the employment transfer of forestry farmers, expanding the theory of labor transfer employment, and providing a theoretical basis for the impact of institutional reform on labor transfer employment.

It should be noted that, due to different ownership of forest land, most of the forest lands in national parks in other countries are state owned, and few studies have been conducted on easement reform related to forest farmers. However, in southern China, forest lands are primarily owned by village collectives, while forest farmers only have the right to use the land. This condition limits the comparability of different research results. Furthermore, the ERCFL is still in the exploratory and pilot stage in China, which is why we currently have limited access to more extensive data, thus restricting the applicability of the results. In the future, given favorable conditions, further research can be conducted to deepen our understanding of the ERCFL. Easement right is essentially a type of usufructuary right, representing the use right of forest land granted by forest farmers to the national park management department. This study can provide valuable insights into the complex issue of easement reform in relation to forest land ownership.