The Influencing Mechanism of Three-Dimensional Capital on the Employment of Land-Expropriated Farmers —Based on Structural Equation Model

Land-expropriated farmers are a new group accompanied by rapid urbanization in China. The problem of employment not only is related to their own survival and development, but also directly affects the quality and effectiveness of economic and social development of China. This paper introduces the three-dimensional capital factors of human capital, social capital and psychological capital, comprehensively analyzes the mechanism of their “independent” and “synergistic” influences on the employment of land-expropriated farmers, and constructs a theoretical model of the influences. Based on the survey data of land-expropriated farmers in Daxing district of Beijing, this paper uses the structural equation model to test the theoretical model empir-ically, and obtains the influence paths and coefficients of three-dimensional capital on the employment of land-expropriated farmers. It is helpful to grasp the key to solve the problem of land-expropriated farmers at the deci-sion-making and implementation levels, and to defuse social risks caused by unemployment of land-expropriated farmers more effectively.


Introduction
With the rapid progress of urbanization in China, a large number of rural collective lands have been expropriated, and the scale of land-expropriated farmers (LEFs) has become increasingly large. According to data released by National key to solve the problems of LEFs. Compared with other types of social security, employment is an active security, which plays an important role in the sustainable development of LEFs in cities. It is not only related to their survival and development, but also directly affects the quality and effectiveness of economic and social development of China.
The income of LEFs mainly comes from working in the factory or engaging in the rural service industry. As the capacity of rural industry to absorb surplus agricultural labors has declined in recent years, rural service industry has become the main channel for surplus agricultural labors to transfer to non-agricultural industries in China. The service industry is mainly labor-intensive or labor-capital-intensive and many jobs have low skill requirements that can reduce training costs. Moreover, the rural service industry is close to the local reality, which can directly provide local employment opportunities and reduce the cost of finding jobs for farmers. With the development of economy and the improvement of income per capita, people's consumption demand for services will be increasingly large, creating conditions for the service industry to absorb a large number of labors. The development of rural service industry (Agricultural Producer Services, Rural Consumption Services and Rural Public Services) is an important path to increase employment opportunities and increase income of LEFs ( Figure 1). However, the rural service industry needs a large number of practical and technical personnel engaged in deep processing of agricultural products, products marketing and circulation services. Lack of technical and skills training, many LEFs are unable to engage in relevant work, and unemployment is the biggest problem facing LEFs.
The employment problem of LEFs has been highly valued by the government and widely concerned by the society. For example, in view of the slow progress of citizenization of rural migrant population and the quality of urbanization needing to be improved in Daxing district of Beijing, the Outline of the 13th Five-Year Plan for National Economic and Social Development of Daxing District and Beijing Economic and Technological Development Zone emphasized to give priority to employment, to promote the employment of key groups, to actively expand employment channels, to achieve the rural migrant population employment and green employment, and to promote the professionalization and Journal of Service Science and Management citizenization of farmers. In the current context, the research on the employment of LEFs is helpful to respond to the above concerns, and can provide references for relevant decision-makers.
The main contribution of this paper is to introduce the three-dimensional capital factors of human capital, social capital and psychological capital, comprehensively analyzes the mechanism of their "independent" and "synergistic" influences on the employment of LEFs, and constructs a theoretical model of the  [5]. The United Kingdom absorbed rural surplus labors by building a service system suitable for agricultural industrialization to develop rural tertiary industry [6], and Japan increased employment opportunities for LEFs by cultivating comprehensive functions of local small cities [7].
The research on LEFs and their employment is mainly carried out from the following three perspectives. The first is the research and evaluation on the resettlements of LEFs. Monetary resettlement is a fixed amount, but it cannot maintain sustainable livelihood [8]. Purchasing insurance for LEFs, and changing one-time land compensation to lifelong guarantee of "land for social security" have problems of unreasonable compensation standards, incomplete land acquisition compensation procedures, and low degree of democratic participation in decision-making of farmers [9]. Under the condition of market economy, the problem of government job placements is that when enterprises dismiss LEFs, the government has no right to interfere [10]. The second is to discuss the difficulties of LEFs employment from the perspectives of relevant system, employment environment and LEFs themselves. For example, Wang (2015) studied the imperfection of land expropriation system, social security system and training system, which restricted the employment of LEFs [11]. Zhou (2015) found that the weak abilities of intermediary organizations and the social exclusion become obstacles to the employment of LEFs [12]. Wang et al. (2017) found that LEFs face severe employment situation due to their low knowledge level and skills level [13]. The third is to analyze the employment paths and necessary as- Combining three-dimensional capital with employment, the main discipline basis is labor economics, sociology and psychology. The existed researches from the perspective of three-dimensional capital mainly include three aspects. Firstly, it is applied to the research of poverty alleviation and anti-poverty mechanism [16]. Secondly, it is applied to the study of urban integration behaviors of rural migrant workers [17]. Thirdly, it is applied to the research on employment assistances, and the objects of assistances include knowledge talents like college students [18]. Although it has been extended to LEFs, a vulnerable group in non-agricultural employment [19], the relevant literature is very limited and the research lacks completeness. It is not only a weak point but also a growth point to study the influences of three-dimensional capital on the employment of LEFs and then to carry out employment assistances.  [20], and are also important influencing factors for the transformation of farmers from low-level urban integration to high-level urban integration [21]. However, the human capital accumulation of LEFs is at a low level, and they are in a weak position in the employment competition due to the constraints of resources, policies, views and values, education and other factors.

The Independent Influences of Three-Dimensional Capital Factors on the Employment of LEFs
Therefore, the following hypothesis is established: H1: The human capital accumulation of LEFs plays a positive role in their employment.
2) The Influences of Social Capital. Social capital is essentially a specific social network formed by human interactions and its potential social resources. It high information credibility and easy to form "group advantages" [23]. The social capital stock of LEFs is generally low and the homogeneous social capital is the main way for them to obtain employment information. The homogenous social capital (strong-relationship) is characterized by easy communication, high stability and strong willingness, which can improve the employment possibility of LEFs. However, Ci (2011) found that it was the weak-relationship rather than the strong-relationship that was more conducive to individuals to obtain higher quality job information and find a better job [24]. The richer the extended heterogeneous social capital (weak-relationship), the more conducive it is to the em-Journal of Service Science and Management 3) The Influences of Psychological Capital. Psychological capital is a core element that transcends human capital and social capital. It is the psychological resource to promote the personal growth and performance improvements, and is the key to win in human resources [25].

The Synergistic Influences of Three-Dimensional Capital Factors on the Employment of LEFs
The core idea of synergy is that if each subsystem in a system can cooperate well,  2) The Synergistic Influences of Social Capital and Psychological Capital The LEFs generally lack heterogeneous social capital, namely, weak-relationship social capital. The weak-relationship promotes the information flow between different groups, disseminates information that people are unlikely to see, and plays a positive role in obtaining employment information of LEFs. Studies have shown that weak-relationship is most likely to provide friends with information that would otherwise be difficult for them to access. Ci (2011) found that a weak-relationship increased the likelihood of information sharing by nearly 10 times, while a strong-relationship increased the likelihood by only 6 times [24]. For LEFs, more strong-relationship and weak-relationship can strengthen their social capital and improve their self-confidence in employment. Positive psychological capital plays a positive role in promoting the accumulation of social capital, especially heterogeneous social capital. LEFs with higher self-efficacy will actively and consciously expand their interpersonal network and social space. Meanwhile, LEFs who are optimistic, confident and resilient have strong interpersonal attraction, which is more conducive to expanding social radius, accumulating more heterogeneous social capital and winning more employment opportunities. Therefore, the following hypotheses are established: H6: The social capital of LEFs plays a positive role in their psychological capital accumulation.
H7: The psychological capital of LEFs plays a positive role in their social capital accumulation.
3) The Synergistic Influences of Human Capital and Psychological Capital Firstly, the accumulation of human capital has a positive impact on the psychological capital accumulation of LEFs. The higher the human capital level of LEFs is, the stronger the employment opportunity perception and employment ability is, and the more likely they are to find a job. This will increase their self-confidence and firm their belief that they can achieve their dreams through hard work. Even in the face of setbacks and difficulties, they can still maintain a positive attitude towards employment with optimism and tenacity, and constantly improve their anti-frustration ability in adversity. Secondly, the psychological capital of LEFs has a correlated impact on their human capital accumulation. The more positive psychological components LEFs have, the stronger their self-efficacy is, and the more willing they are to adapt to new fields, accept new skills and new ideas, and show more flexibility and creativity in the process of career search. The higher the level of psychological capital of LEFs, the more able they are to stimulate their own motivation of achievement, thus consciously strengthening the improvement of knowledge, skills and experience. On the contrary, the lower the level of human capital is, the more likely for LEFs to be frustrated in the process of employment, and also prone to produce a negative psychological state. The more negative psychological components of LEFs, the more pessimistic they are about their career and future, the more closed and re-Journal of Service Science and Management jected they are to the outside world, and the more passive they are to participate in employment training. This negative psychological state is bound to restrain the accumulation of human capital and the display of its effectiveness. If not properly handled, it may lead to many negative misconducts and bring about a series of social problems. Therefore, the following hypotheses are established:   (1988) found that SEM could be applied to estimate the parameters of potential variable or complex independent variable and dependent variable in the prediction model, which could not only help analyze the measurement error, but also analyze the structural relationship between potential variables, which was an important method to deal with the problem of multiple variables (including potential variables) in social science research [28]. This study belongs to the category of social science research, including many variables, which is suitable to use SEM for measurement. Journal of Service Science and Management Theoretical analysis shows that the more abundant the human capital, social capital and psychological capital the LEFs have，the stronger the employment competitiveness and more employment opportunities they have. According to relevant studies on the employment of LEFs, combined with characteristics of LEFs and questionnaire survey, observed variables of human capital, social capital, psychological capital and employment are extracted as shown in Table 1.

Empirical Analysis
Furthermore, according to the hypotheses, a theoretical model of LEFs employment can be constructed (Figure 3).

2) Data Description
The data are from a questionnaire survey of 33 land-expropriated villages in   The 5-point Likert Scale was used to measure the items in this questionnaire.
The measuring scale was 5 points for very large impact, 4 points for relatively large impact, 3 points for general impact, 2 points for relatively small impact, and 1 point for very small impact.

3) Questionnaire Test
Reliability analysis can be used to test whether the measurement results reflected the real characteristics of the respondents' stability and consistency. The higher the reliability of the scale is, the smaller the standard error is. Reliability coefficient above 0.7 is acceptable, and above 0.8 means high internal consistency.
Firstly, SPSS18.0 software was used to test the independence of the questionnaire data, and the questions with low identification degree (Employment Expectation) were eliminated. The Cronbach coefficient was further used for reliability analysis, and the Cronbach's Alpha coefficient of the scale in the questionnaire was 0.916. This indicated that the questionnaire had good internal consistency and the evaluation results had high reliability. Secondly, the validity of the questionnaire was analyzed. Six experts were invited to evaluate the content and structure of the questionnaire, and the expert identification rate was 95.12%, indicating that the validity of the questionnaire was high. Table 2 reflects the degree of fitting between structural equation model and survey data. The chi-square value of the model is 871.339, the degree of freedom is 403, and the statistical significance P of chi-square test is less than 0. 001. Chi-square value and degree of freedom ratio are 2.162, less than 3, which can truly reflect the survey data. The other two absolute fitting indexes GFI is 0.921, Note: a) *P < 0.05, **P < 0.01, ***P < 0.001; In the structural equation model, the significance level reaches 0. 05, which means there is significant difference between the hypothesis model and the observed data. RMSEA is the root mean square error of approximation and value of 0.05 (or 0.08) or below indicate good model fitting. GFI is the absolute fitting index. Value of 0.90 and above indicate good model fitting. b) CFI, IFI and NFI are relative fitting indices. Value of 0.90 and above indicate good model fitting. c) AIC and BIC are information indices, and the smaller they are, the simpler and more efficient the model will be. After a series of steps as normality test, outliers test and parameter correction, the modified structural equation model ( Figure 4) and the effects of three-dimensional capital latent variables on employment (Table 3) are obtained.

4) Hypothesis Test a) Fitting Degree Analysis
In Figure 4, the arrows represent the direct influences between variables, and the values represent the path coefficients, i.e. the direct influence of one variable on another. The larger the value is, the greater the influence of one variable on the other is. The specific analysis results are as follows: firstly, the modified  In Table 3