1. Introduction
In recent years, China’s sustainable rural development has faced many challenges, some of which involve farmers’ land decisions such land-responsibility behaviour intention. For example, farmers are confronted with decisions on non-point source pollution caused by the overuse of pesticides, the loss of traditional farming culture caused by the abandonment of peasants, as well as vegetable and food safety problems caused by pesticide residues. Farmers are the main subjects of the countryside and the principal force promoting the revitalisation of China’s rural areas. Therefore, understanding farmers’ land behaviour could help to clarify the role of policies in development of rural areas, and to aid the integrated and coordinated development of multiple industries, such as agriculture, forestry, and tourism, as well as encourage the protective use of environmental and cultural factors in land-planning.
Researchers have primarily used the research framework of the theory of planned behaviour (TPB) [
1] to investigate farmers’ land behaviour and policy interactions. The Theory of Planned Behaviour (TPB) is an extension of the theory of reasoned action [
2]. The theory uses factors such as attitudes, social norms, and personal abilities to predict individual farmers’ behavioural intentions and directions [
3,
4,
5,
6]. Farmers’ decision making in agricultural practice can be rooted in certain social and cultural backgrounds, and many behaviours cannot be explained by the theoretical framework of rational behaviour alone. Indeed, the included variables in the TPB cannot fully explain the large variances in individual behavioural intentions [
7,
8]. Therefore, increasingly, researchers have attempted to combine the rational logic of TPB with farmers’ personal emotions, feelings, and other psychological factors to more accurately predict and manage farmers’ land behaviours and policy responses [
9,
10]. To enhance the predictive power of TBP in farmers’ decision making, social psychologists have renewed their interests recently in articulating how self-identity can play an important role [
10,
11]. Self-identity has a strong correlation with behavioural intention across a wide range of public health areas [
12,
13,
14], consumer behaviour [
15,
16], and environmental behaviour [
17,
18,
19]
The role of self-identity to farmers’ behavioural intentions has been illustrated empirically in many studies [
20,
21]. Based on Self-Identity Theory, self is envisaged as a social construct in which a distinctive self-component represents each of the roles we occupy in different social settings, internally generated role-expectation [
22]. It views self as a societal role and incorporates the meanings and expectations associated with that role. For farmers, when their self-perception positively matches with the behavioural outcomes, they would have more intention to undertake the action. To summarise, self-identity is one of the most important non-rational factors [
23] for predicting farmers’ land-responsibility behaviour.
At present, the process of urbanisation in China is quickly advancing to the countryside, and rural tourism around large cities is developing rapidly. This rapid development can be expected to affect the thinking, especially in terms of how to re-recognise themselves, agriculture and land. The development of rural tourism has changed the way that rural residents and urban populations interact, forming a new social interaction context. On the one hand, more and more Chinese urban residents flock to the countryside to experience rural life. This may cause farmers to re-examine their environment and self-identity, and thus generate a new self-identity. On the other hand, in these rural areas, part of the land is dedicated to rural tourism, and agriculture has gradually developed from having a single function (food and vegetable production) to multiple functions, such as natural landscape provision, cultural atmosphere creation, and education; furthermore, these non-economic functions are linked to economic functions through the tourism industry. At the same time, farmers’ land decisions such as land-responsibility behaviour can also be affected by farmers’ perception of internal self and external environment.
Self-Identity Theory emphasises that self-identity affects the cognitive style of individuals and can predict the direction of individual perception and cognitive process. Lee et al. [
24] used it within the context of the Technology Acceptance Model (TAM), introducing an extension of a validated framework, and proved that self-identity influences the adoption behaviour of WebCT through perceived usefulness. From this perspective, how does farmers’ self-identity affect their cognition of agricultural function? Additionally, whether it determines their land responsibility behaviour? This has not been discussed in the literature before.
The environmental protection behaviour of farmers and the issue of inheritance of traditional culture are related to future rural sustainable development. The study defined the land decision-making behaviours that are conducive to sustainable rural development as land responsibility behaviour intention. In rural tourism text, farmers’ land-responsibility behaviour intention refers to their intention of behaviour aimed the environmental protection, social profit, and inheritance of traditional culture that are related to future rural sustainable development. Land-responsibility behaviour is the most important factor related to sustainable development of rural areas. The aim of sustainable development is to achieve a balance between the complementary goals of providing environmental, economic, and social opportunities for the benefit of present and future generations, while also maintaining and enhancing the quality of the land resource [
25].
From a farm level, contributing to the preservation of the landscape’s character, strengthening the landscape’s quality, and sustainable development, the study presents the relations of farmers self-identity, agricultural multifunction perception (including agricultural non-economic function perception and agricultural economic function perception), and their land responsibility behaviour. This empirical study sought to answer how farmers’ self-identity affects their perception of agricultural economic function, non-economic function and land responsibility behaviour. The findings could assist policymakers and land use planners in decision making related to sustainable rural tourism in China.
3. Methodology
3.1. Research Area
Over nearly 10 years of development, a more mature suburban rural tourism industry cluster has formed around Chengdu, Sichuan Province, which is a region in China that well reflects multifunctional agriculture. In this study, four rural tourist sites around Chengdu were selected as our survey sites (Nongke Village in Pi County, Mingyue Village in Pujiang County, Taohuagou in Longquanyi District, and Sansheng Township in Jinjiang District).
Rural tourism has been developing in Sansheng Township, Jinjiang District, since the early 1990s. The “Five Golden Flowers”, which has an area of about 12 km2, is a typical representative area of rural tourism; it is a tourism and leisure area that integrates business, leisure, vacation, culture, and creativity. Taohuagou, Longquanyi District, is a representative area of multifunctional agriculture; it is an agricultural tourist attraction with peach and pear trees as the main rural landscape resources. Compared with Sansheng Township, the degree of government participation is relatively low, and most of initiatives are farmers’ independent development. Nongke Village, which is located in Youai Town, Pidu District, is a national agricultural tourism site; its main tourism interest is agritainment, which originated in the 1980s, whereby farmers attract citizens by opening their Sichuan bonsai nurseries for tourists to visit. The tourism industry started later in Mingyue Village, Pujiang County, which relies on local resources such as Phyllostachys praecox, ecological tea gardens, and ancient kilns, with the theme of pottery culture and home to the Mingyue International Ceramic Art Industry Cultural and Creative Park. With the creation of a humanistic ecological resort that integrates ceramic art production and sales, cultural display, creative experience, leisure sports, and rural vacations, this area is now a well-known rural tourism destination and a prime example of rural construction in Chengdu.
3.2. Sampling Procedures
From March to June 2018, random sampling method was used to select farmers from village household lists provided by local government authorities and institutions in the selected four rural tourist sites, resulting in a total of 393 famers overall.
After 46 farmers were removed from the dataset because of missing and inconsistent answers, a total of 347 valid questionnaires were obtained. Of these, 92 were from Mingyue Village, 88 from Taohuagou Village, 91 from Nongke Village, and 76 from Sansheng Township. The socioeconomic characteristics of the study sample are shown in
Table 1.
3.3. Measures
3.3.1. Farmers’ Self-Identity
The three items of farmers’ self-identity were measured using the self-identity scale by Lee et al. [
24]. Amendments to this scale were made with consideration to the characteristics of farmer identity, mainly from the perspective of professional identity, such as perception of farmers’ job independence, pride (Christensen & P., 2004; Key, 2005), and lifestyles (Howley, 2014). Based on previous interviews and the identity characteristics of Chinese farmers, three measurement items were used, including “I enjoy the lifestyle of being a farmer” (F1), “Being a farmer is an honest profession” (F2), and “I have freedom and independence as a farmer” (F3). All variables were scored on a Likert-type scale that ranged from totally disagree (1) to totally agree (7).
3.3.2. Multifunctional Agriculture Perception: Agricultural Non-Economic Function Perception and Agricultural Economic Function Perception
The multifunctional agriculture perception scale was developed by Kvakkestad et al. [
39]. In that original study, the authors designed a 16 item land multifunctional perception survey that was completed by farmers in Norwegian agricultural cultural heritage sites. In this study, to ensure the validity of the measurement items, we randomly selected 15 farmers in the surveyed area and conducted one-to-one in-depth interviews. The main question of the interview was “As a farmer, what function of the agriculture is important to you?” According to the interview results and specific items of multifunctional agriculture from Kvakkestad et al. [
39], nine items were put forward by the farmers (e.g., keeping the land or countryside tidy), seven items were not suitable for the research situation (e.g., securing the workplace of myself and my family). Finally, nine items were used to measure multifunctional agriculture perception including agricultural non-economic function perception and agricultural economic function perception in this study [
48].
In these items, agricultural non-economic function perceptions are as follows: carrying forward production knowledge and a lifestyle that had been passed down from ancestors (NEF1); keeping the land or countryside tidy (NEF2); maintaining the rural cultural landscape (NEF3); conserving the natural environment (e.g., by minimising pollution) (NEF4); taking care of the land and other resources left by seniors (NEF5). Farmers were asked to rate their importance on the five items, ranging from 1 to 7, representing the importance from not being important at all to being very important.
Agricultural economic function perceptions are as follows: receiving a higher income through agriculture on the basis of constant land area (EF1); obtaining the maximum economic benefits (EF2); receiving a satisfactory income (EF3); ensuring that there are sufficient food and vegetable supplies in the event of an emergency (such as a natural disaster) (EF4). Farmers were asked to rate their importance on the four items, ranging from 1 to 7, representing the importance from not being important at all to being very important.
3.3.3. Land-Responsibility Behaviour Intention
Considering three typical land decision-making behaviours of farmers that affect sustainable rural tourism, such as food production, farmland landscape and cultural inheritance [
11,
49,
50], farmers’ land-responsibility behaviour intention was measured by the degree of attention paid by farmers to the three following aspects during land disposal: social benefits (such as food security and reducing pesticide use) (P1), environmental benefits (such as reducing pollution and protection of farmland landscape) (P2), and cultural conservation benefits (such as teaching agricultural knowledge to future generations) (P3). All variables were scored on a Likert-type scale that ranged from totally disagree (1) to totally agree (7).
3.4. Pre-Test
To ensure the validity of the questionnaire, five farmers in Sansheng Township were asked to complete a pre-interview, and the words and expressions that appeared in the pre-interview were used to form a pre-test questionnaire. The pre-test questionnaire was then used in a small sample survey of 40 farmers. Analysis of pre-test sample scores was performed to identify variables that passed the reliability and validity tests. The corrected item-total correlation and the internal consistency reliability index (Cronbach’s α coefficient) were used to test the reliability of the four variables measured by the questionnaire, and SPSS 25.0 was used. The corrected item-total correlation values of all items retained exceeded 0.6 [
51], and the Cronbach’s α reliability coefficients of all variables exceeded the recommended value of 0.6. After deleting any item, the overall Cronbach’s α reliability coefficient of each variable did not increase significantly. Thus, we concluded that the scale had good internal consistency, reliability, and stability, and ideal internal reliability. Furthermore, the validity of the construction of the measurement scale was tested using factor analysis.
As shown in
Table 2, the Kaiser–Meyer–Olkin values of the four latent variables (self-identity, non-economic function perception, economic function perception, and land-responsibility behaviour intention) all exceeded 0.6, which was greater than the recommended value of 0.5. The significance of the Bartlett’s test of sphericity was 0.000. The original hypothesis of the Bartlett’s test of sphericity was rejected, and so the questionnaire measurement scale and the construct validity of each variable could be considered as good.
4. Results and Discussion
4.1. Measurement Model Estimation
We used Mplus8.0 to perform confirmatory factor analysis to detect the structural validity of the measurement scale, including factors self-identity, agricultural non-economic function perception, agricultural non-economic function perception, and land-responsibility behaviour intention. The various inspection indicators after deleting “taking care of the land and other resources left by seniors is important to me.” (NEF5) for correction are shown in
Table 3; the revised scale had a better composition reliability and structural validity.
Table 3 presents all constructs’ factor loadings, Construct Reliability (CR), and Average Variance Extracted (AVE), and
Table 4 presents the relationships between the constructs.
The results revealed that the model fit was χ2 = 212.604, df = 71, χ2/df = 2.9 (less than the recommended value of 3), CFI = 0.942, TLI = 0.926, RMSEA = 0.068, and SRMR = 0.046. The factor loadings of most items were >0.7, the composition reliability of each factor was >0.8, the convergence validity was >0.5, and the fit degree of the measurement model reached an ideal value
4.2. Structural Model Estimation
4.2.1. Path Analysis
First, we adopted Mplus8.0 to estimate the regression coefficient between variables. Given that the data were non-normally distributed, we used the maximum likelihood estimation method provided by Mplus8.0 to verify the relationships between the variables, and the standard error and mean-variance corrected chi-square test (MLMV) as the estimation method. The results revealed that RMSEA = 0.076, SRMR = 0.048 (recommended value < 0.08), CFI = 0.942, TLI = 0.915 (recommended value > 0.9), and χ2 (162) = 411; these results show that the data fit the model well. We further tested the hypotheses. Self-identity had a significant positive influence on land-responsibility behaviour intention (β = 0.224,
p < 0.01), agricultural non-economy function perception (β = 0.128,
p < 0.01) and agricultural economy function perception (β = 0.319,
p < 0.001), thus supporting H1, H2a and H2b. Agricultural non-economic function perception positively and significantly affected agricultural economy function perception (β = 0.557,
p < 0.001), supporting H4, but agricultural non-economic function perception negatively and significantly affected land-responsibility behaviour intention (β = −0.319,
p = 0.016), not supporting H3a. Similarly, agricultural economy function perception was found to significantly influence land-responsibility behaviour intention (β = 0.866,
p < 0.001), which supported H3b. Results of the hypotheses tests are summarised in
Table 5.
4.2.2. Mediating Effects Estimation
Then, we used Mplus8.0 bootstrapping to test the mediating role of ANEFP and AEFP in the relationship between SI and farmers’ LRBI. The bootstrapping method has more advantages than the traditional mediation analysis method because it can statistically calculate the significance of direct effects, indirect effects, and total effects within a certain confidence interval (CI) [
52,
53]. The results are shown in the
Table 6.
We found that farmers’ self-identity had a significant positive impact on farmers’ land-responsibility behaviour intention (total effect: β = 0.521, 95% CI = (0.335, 0.743)). The total indirect effect was β = 0.297, 95% CI = (0.169, 0.467). The ratio of total indirect effects to total effects was 0.297/0.521 = 0.692. In other words, 69.2% of the impact of farmers’ self-identity on land-responsibility behaviour intention was affected by agricultural non-economic function perception and agricultural economic function perception. Results also indicated that self-identity directly or through intermediary conditions affected land-responsibility behaviour intention.
In addition, the mediation test of agricultural economic function perception revealed that the significant impact of self-identity on farmers’ land-responsibility behaviour intention was mediated by agricultural economic function perception (β = 0.224, 95% confidence interval (CI) = (0.152, 0.427)). However, the mediation test for agricultural non-economic function perception was not supported. This means that farmers’ agricultural non-economic function perception mediates the relationship between farmers’ self-identity and land-responsibility behaviour intention (H2a) was not supported. The results of each hypothesis test are shown in
Table 6.
4.3. Discussion
This study investigated the impact of farmers’ self-identity on land-responsibility behaviour intention from individual perspective of local farmers and examined the mediating effect of agricultural multifunctional perception on the relationship between these two variables. Two main research conclusions were obtained.
First, in rural tourism destinations in suburban districts of China, farmers’ self-identity is an important variable that affects farmers’ land-responsibility behaviour intention, whereby the higher the level of farmers’ self-identity, the more likely they are to adopt land-responsibility behaviours. On the one hand, farmers’ self-identity can directly affect farmers’ land-responsibility behaviour intention. On the other hand, self-identity can also positively influence farmers’ land-responsibility behaviour intention through the mediating effect of agricultural multifunctional perception, which means that farmers’ self-identity will further initiate rational behaviour decision making through the functional evaluation of agriculture. That is, self-identity support land-responsibility behaviour, directly and indirectly supporting the perceived economic function of agriculture.
Second, agricultural non-economic function perception negatively and significantly affected land-responsibility behaviour intention. It was found that some non-economic benefits of agriculture, such as environmental protection and social culture, may come at the cost of individual farmers’ interests, which supports previous findings [
54,
55]. Therefore, although farmers know that adopting certain technologies or programs can improve the non-economic functions of agriculture, they may not adopt corresponding land behaviours [
56,
57]. In other words, although farmers can envision the non-economic functions that their land-responsibility behaviour may bring about, they may not adopt it, even oppose this kind of behaviour. The main reason for this could be that the farmers bear the additional economic costs for land-responsibility behaviour.
Third, the analysis of the mediating effect shows that the perception of the importance of agricultural multifunction is an important mediating condition for the influence of farmers’ self-identity on farmers’ land-responsibility behaviour intention. However, the utility of importance perception of agricultural economic function and that of non-economic function are different. Self-identity can influence farmers’ land-responsibility behaviour intention through the mediating role of agricultural economic function perception. However, agricultural non-economic function perception does not have a direct mediating effect; that is, although farmers perceive the non-economic functions that their land-responsibility behaviour may bring, this does not directly stimulate them to generate land-responsibility behaviour intention. Farmers’ agricultural non-economic functions perception can only significantly influence their land-responsibility behaviour intention through the intermediary effect of agricultural economic functions of land. That is, only when farmers perceive that non-economic function is positively related to economic function does the corresponding land-responsibility behaviour intention occur. Furthermore, the likelihood of adopting land-responsibility behaviours will increase if farmers feel that the non-economic functions of agriculture are accompanied by economic functions that can offset the perceived costs.
5. Conclusions
The results indicate that self-identity is a vital factor that affects Chinese farmers’ land-responsibility behaviour intention in rural tourism areas. What is more, “not well respected, rather perceived as a low-rank profession” and “the low social status” are primary factors discourage youths from getting involved with farming [
58]. Therefore, the government should understand how farmer perspective the value of farming and consider its role in growing the rural economy and rural development.
Second, the conclusions of this study highlight the path dependence of farmers on the economic functions of agriculture. The income of farmers is generally low in China and obtaining economic benefits through agriculture remains the most important motivation for farmers to take land-responsibility behaviour. Therefore, only when farmers get the economic benefits of the tourism industry caused by the non-economic functions of agriculture, will farmers’ land-responsibility decisions be effectively stimulated.
Third, farmers’ agricultural non-economic function perception positively affects their agricultural economic function perception. That is, farmers can perceive the transformation of agricultural non-economic functions into economic functions, thereby increasing their understanding and support for sustainable land policies. For the sustainable development of villages, our findings are consistent with those of Ahnström et al. [
56] and Reimer et al. [
57], among others. We believe that investing in the non-economic functions of agriculture will be beneficial to rural communities and their sustainable development. The multifunctional development of agriculture is of great significance to the protection of the rural landscape, ecology, the cultural environment, and the protection of biodiversity and cultural heritage; these factors form the basis for agriculture to generate economic benefits in tourism and other industries. The tourism industry is an important means by which to transform agricultural non-economic functions into economic functions. When farmers in these areas understand that agricultural non-economic functions can achieve economic functions through tourism and other industries, they could be more likely to adopt land behaviours that are more conducive to sustainable development. For agricultural heritage sites and remote rural tourism sites with abundant tourism resources, agriculture and farming culture are important tourist attractions, and rural tourism can be developed with the help of agricultural non-economic functions. Therefore, rural tourism can not only be used as a means by which to alleviate poverty through the development of rural areas but could also enable farmers to experience the transformation of agricultural non-economic functions into agricultural economic functions, thereby stimulating local farmers to adopt more sustainable land decisions.