According to social capital theory, the relationships and trust that individuals build in social networks can provide resources and information (Cui et al., 2022). For relocated farmers, good social integration can enhance their social networks (Tuominen and Haanpää, 2022), provide more employment opportunities and market information, thereby increasing household income. Social integration is also closely related to the development of human capital (Xu et al., 2022). By integrating into the new environment, relocated farmers can gain access to training and skill upgrading opportunities, thereby increasing their employability and labor productivity (Cao et al., 2015). This ability to improve can directly contribute to the growth of household income. On this basis, the role of social capital is not limited to the individual level, but also affects the lives of relocated farmers through intergenerational transmission and social structure (Zhao et al., 2021). Social capital cannot only affect the economic conditions of the current generation of farmers, but also affect the education, employment and social integration level of the next generation through the intergenerational transmission mechanism. Social networks and the accumulation of trust facilitate the transfer of resources across generations, including financial support, information flow and social support, which can have a profound impact on children’s education, employment choices and social status. In addition, the role of social capital involves changes in the social structure, especially in new communities after migration. The social capital accumulation of relocated farmers is not only the result of individual behavior, but also affected by community structure, public resource allocation and social policy (Zhang et al., 2022). Through mutual support and resource sharing among communities, relocated farmers can better integrate into the new environment, obtain more social support, and promote social integration. Therefore, the hypothesis is proposed:

The income of relocated households consists of operating income, wage income, property income, and transfer income. Social integration can affect operating income by enhancing social capital. The social capital theory holds that social networks and social relationships can provide resources, information, and support, all of which have a positive impact on business activities (Cao et al., 2015). For example, relocated farmers who integrate into the new community can utilize the local commercial network and resources to obtain market information, technical support, and cooperation opportunities, thereby increasing their operational income. Individual skills, knowledge and education levels have a significant impact on income, and after integrating into the new community, relocated farmers may have the opportunity to receive better vocational training and education, enhancing their employability and market competitiveness (Liu et al., 2023). Social integration helps relocated farmers accumulate various forms of capital, including financial and material capital. By participating in investments such as renting out houses and land to earn rent, participating in cooperative dividends, and earning interest income from funds, relocated farmers can increase their property income. Social integration enhances the social support network of relocated farmers, making it easier for them to receive social assistance and government subsidies, and increasing their transfer income (Cao et al., 2015). Based on this, the following research hypothesis is proposed:

Human capital theory emphasizes the importance of an individual’s knowledge, skills, and experience to economic output (Wu et al., 2022). Social integration helps farmers establish broader social networks, promote their access to more employment information and opportunities, and enhance their knowledge, skills, and labor productivity (Cui et al., 2022). These all belong to the category of human capital and directly affect economic income. Non-farm employment is usually accompanied by higher wages and more stable job opportunities (Huang et al., 2022). With the strengthening of social integration, relocated farmers can more easily enter the non-agricultural employment market, thereby increasing household income. This process indicates that non-agricultural employment plays an important mediating role between social integration and household income, and this mediating effect is closely related to the improvement of human capital.

The information asymmetry theory states that access to and use of information has a significant impact on an individual’s decision making (Mavlanova et al., 2012). The existence of information asymmetry often leads to inefficient decision-making and misallocation of resources, which can be significantly improved by social integration. Social integration enhances relocated farmers’ access to information, enabling them to keep abreast of market dynamics, employment opportunities, and relevant policy information (Vancea and Boso, 2015). This access to information can help relocated farmers make better economic decisions and thus improve household income.

Life satisfaction reflects an individual’s subjective assessment of his or her own life status, and an individual’s life satisfaction affects his or her overall well-being and job performance (Ahn et al., 2023). Good social integration enables relocated farmers to feel a sense of belonging and support in the new environment, which in turn increases their life satisfaction (Piao et al., 2022). Positive emotional state and mental health promote better work performance and productivity (Walasek et al., 2019) which in turn affects household income. Synthesizing the above analysis, the following hypothesis is proposed:

Based on the previous theoretical analysis, the theoretical analytical framework of social integration affecting relocated household income is constructed (Figure 1) to verify the mechanism of social integration affecting relocated farmers’ income and the mediating effects of non-agricultural employment, information acquisitions and life satisfaction.

The data used in this paper are from the field research conducted by the research group from January to March 2024 in Liangshan Prefecture, Sichuan Province, on the theme of “social integration of farmers relocated to alleviate poverty through land resettlement.” In the survey, firstly, a typical survey was used to select five counties with more resettlement from the 17 counties (cities) in Liangshan Prefecture, and finally, Ganluo, Butao, Yuexi, Meigu, and Zhaojue counties were selected as the sample source counties. Secondly, the method of combining stratified sampling and random sampling was adopted, taking full consideration of the differences in the level of economic development of different regions and the types of characteristics of different relocated farmers, and, as far as possible, a variety of types of relocated farmer samples were selected to ensure that the research results were closer to the actuality. One to three resettlement districts were randomly selected from each county, and 50–80 research samples were selected from each resettlement district, and household surveys were carried out. A total of 661 questionnaires were recovered, and after eliminating invalid questionnaires, 610 valid questionnaires were finally obtained, with the validity rate reaching 92%.

In order to investigate the influence of social integration on family income level, this paper adopts the method of gradually adding variables. Specifically, from model (1) to model (4) in Table 2, social integration variables, household head characteristics variables, family characteristics variables and resettlement characteristics variables are gradually introduced, respectively. The estimated results in Table 2 show that the social integration variables in models (1) to (4) are at least significant at the 5% level, and the sign is positive. This indicates that social integration significantly promotes the improvement of household income level of relocated farmers, and the influence of social integration on household income is stable after gradually strengthening the constraints. The estimation results of model (4) show that social integration significantly increases the household income of relocated farmers after controlling the characteristics of household head, family and resettlement. When social integration increases by 1 unit, the household income level of relocated farmers increases by 0.359 units, and hypothesis 1 is verified.

In terms of the impact of control variables on the household income of relocated farmers, gender is significant at the 5% statistical level, and the coefficient estimate is 0.119, which indicates that the income-enhancing effect brought by men contribute 11.9% more to household income than women. Skills training has a significant positive impact on household income, making it of great policy significance for the government to increase vocational skills training for relocated farmers and guide them to non-farm employment. Family size is significant at the 1% statistical level, and the coefficient estimate is 0.098. For every additional member of the family, the total family income increases by 9.8%. This suggests that the larger the household size, the more potential labor force there is, and the more total household income is generated. The proportion of labor force significantly increases the level of income, when the proportion of household labor force increased by 10%, the total household income is increased by 3.5%, the larger number of labor force in the household means that more members can engage in economic activities and work, increasing the total income of the household. The impact of livelihood capital on household income is significantly positive, with total household income boosted by 33.5% for every one unit increase in livelihood capital, probably because having more and higher quality livelihood capital means that households have stronger productive capacity, employability, and risk resistance, which not only directly increases income, but also, through optimizing resource allocation, improving productive efficiency, and expanding the scope of economic activities, can indirectly contribute to the growth of household income (Wang et al., 2023; Zhao and Lan, 2023). The location of resettlement is significant at 1% statistical level, and the coefficient is 0.205, which indicates that the location of resettlement is one of the important factors affecting the household income of relocated farmers.

When the benchmark regression model estimates the impact of social integration on the household income of relocated farmers, there may be endogeneity problems caused by factors such as mutual causation and missing variables. It is necessary to select appropriate instrumental variables to overcome the potential endogeneity problems. Drawing on Zhao et al. (2022) and Cui and Wang (2020), the “mean value of social integration of other relocated farmers in the same resettlement community” and the “duration of residence in the new community after relocation” were selected as instrumental variables for social integration. Since the same community is an acquaintance society and the “peer effect” prevails, the social integration of relocated farmers will be affected by the mean value of social integration of other relocated farmers in the same community, which is in line with the principle of correlation of instrumental variables. At the same time, the social integration of other relocated households in the surrounding area will not directly affect the farmer income of the relocated household, and there is no inevitable connection between the two, which is in line with the requirement of exogeneity of instrumental variables. The longer the time of living in the new community after relocation, the more time and opportunity the family members have to establish contact with the residents of the new community, build a new social network, increase social capital, and thus improve the level of social integration. The length of residence in the new community after relocation mainly affects the process of family adaptation to the new community and the establishment of social networks, rather than directly determining the income level of the family. Therefore, the “duration of residence in the new community after relocation” also meets the requirements of relevance and exogeneity of instrumental variables.

The 2SLS results are shown in Table 3 for model (1) and model (2). From the results of the non-identification test, the Kleibergen-Paap rk LM statistic is 69.501, corresponding to a p-value of 0.000, rejecting the original hypothesis of “non-identification.” From the test of weak instrumental variables, the Cragg-Donald Wald statistic is 39.952, which is greater than the critical value of 19.93 at the 10% level of bias, rejecting the hypothesis of “the existence of weak instrumental variables.” In the over-identification test, the Hansen J statistic corresponds to a p-value greater than 0.1, which means that the original hypothesis of “all instrumental variables are exogenous” is accepted, and the instrumental variables meet the requirement of exogeneity. From the results of the second stage regression, after overcoming the potential endogeneity problem, the regression coefficient of social integration on the household income of relocated farmers is 0.964, which is significant at the 1% statistical level. Although the estimated coefficient is larger than that of the baseline regression, the sign and significance of the coefficient have not changed substantially. This indicates that the conclusion that the improvement of social integration level will increase the household income of relocated farmers is significant and robust.

Table 5 reports the estimation results of social integration on the income structure of relocated households using the Tobit model. The regression results of model (1) show that social integration has a positive impact on operating income, indicating that social integration has a significant positive effect on the operating income of relocated households. This may be because social integration helps relocated farmers establish connections with local residents, obtain more market information, technical knowledge, and business experience, thereby improving the efficiency and effectiveness of business activities. The coefficient of social integration in model (2) is 1.070, and it has passed the significance test at the 1% level, indicating that social integration has a significant positive impact on the wage income of relocated farmers. This may be because by integrating into the new social environment, relocated family members of farmers can more easily access local employment information, including job advertisements, business needs, etc., which helps them find suitable jobs. According to the regression results of model (3), social integration has a significant impact on property income, and the coefficient is positive, which is consistent with research hypothesis 2. On the contrary, the results of model (4) show that the social integration of relocated farmers has no effect on their transfer income, which is partially biased from hypothesis 2, indicating that the current level of social integration of relocated farmers in the research subjects is not significant enough to significantly promote the increase of family transfer income.

According to the results in Table 6, the coefficient of the total effect of social integration on household income is 0.359, and the lower limit is 0.267 and the upper limit is 0.452, and the range excludes 0, which indicates that the total effect of social integration on the household income of relocated farmers is significant. The test coefficient of the mediating effect of non-farm employment between social integration and household income is 0.075, and the lower limit is 0.043 and the upper limit is 0.112, the range does not include 0, indicating a significant intermediary effect. The test coefficient of the mediating effect of information acquisition between social integration and household income is 0.071 and the lower limit is 0.039 and the upper limit is 0.108. The range does not include 0, so the mediating effect is significant. The test coefficient for the mediating effect of life satisfaction between social integration and household income is 0.101 and has a lower bound of 0.060 and an upper bound of 0.143, which excludes 0 and has a significant mediating effect.

Table 6 presents not only the results of the tests of the path of influence, but also the results of the total indirect and direct effects. The total indirect effect is the overall mediating effect of the three variables of non-farm employment, information acquisition and life satisfaction between social integration and household income, with a coefficient of 0.246, a lower bound of 0.182 and an upper bound of 0.314, with confidence intervals that do not include 0, indicating that these three mediating variables generally mediate between social integration and household income of relocated farm households. The direct effect of social integration on household income, i.e., the “net effect” of social integration itself on household income after controlling for the mediating effects of non-farm employment, information acquisition and life satisfaction. The results show that the coefficient of the direct effect is 0.113, with a lower confidence interval of 0.023 and an upper confidence interval of 0.203, which does not include zero, indicating that social integration itself has a significant contribution to household income. Therefore, social integration not only indirectly affects household income through mediating variables such as non-agricultural employment, information acquisition, and life satisfaction, but also directly has a significant positive impact on household income. Hypotheses 3, 4, and 5 have been validated.