Introduction

Over the past decade, China’s central government has attached great importance to encouraging college students to return home and start businesses, in particular by supporting the expansion of the entrepreneurial space through IDRI and releasing the space for college students to return home and start businesses. As a practical response to the policy, through the policies of integrated development of rural industries (IDRI), Chinese governments and social organizations at all levels are promoting development models such as the integration of rural planting, breeding and processing industries, the integration of production, supply, and marketing, and the integration of agriculture and cultural tourism, so as to achieve the IDRI and promote the growth of the rural economy, and at the same time create more and better external environments for college students to return to their hometowns to start their own businesses. In addition, the Chinese government also encourages college students to return to their hometowns to development of entrepreneurial activities by providing direct support in the form of start-up funds. In 2023, China’s various government funds at all levels will directly support employment and entrepreneurship by more than 300 billion yuan (Data sourced from Economic Daily).

However, it is a basic fact that with the accelerated development of IDRI, the proportion of college students participating in entrepreneurial activities has increased more than fivefold compared to the past, but the success rate remains low due to the generally poor entrepreneurial performance. The entrepreneurial success rate of Chinese college students is 3% on average, while the entrepreneurial success rate of college students in the United States can reach 20–30% on average. It has been proved that IDRI can indeed promote the increase of entrepreneurial activities of college students returning to their hometowns, but the problem now is that the success rate of Chinese college students returning to their hometowns has not improved. It is not only a failure to achieve the desired objectives of the policy, but also a waste of government support funds. In order to fully resolve this issue, it is necessary to first clarify the mechanism of IDRI’s influence on the entrepreneurial performance of returning students.

We first reviewed the existing literature, up to now, there are two mainstream views in the academic community to address this issue. One view is that students’ own entrepreneurial ability determines whether the entrepreneurial activity can be successful or not (Paramba et al., 2023). Paramba et al. believe that college students’ own entrepreneurial ability, including intellectual capital and entrepreneurial strategic ability, are the key factors affecting the success rate of entrepreneurship. Some scholars believe that college students lack practical experience and social capital, which will lead to the possibility of failure of their entrepreneurial activities, therefore, strengthening the social activity ability of college students is conducive to expanding their social capital and increasing the success rate of entrepreneurship (Fokkema et al., 2017; Lu et al., 2023). Another view is that the external entrepreneurial environment has a greater impact on college students’ entrepreneurial activities. Some scholars believe that too strict legal system is not conducive to the development of entrepreneurial activities and the success of entrepreneurial activities (Williams, 2011); some scholars believe that the current entrepreneurial ecosystem, including entrepreneurial policies are not optimized enough will lead to the low entrepreneurial success rate of college students (Aguilar, 2021).

As mentioned above, although scholars have made a lot of important achievements in the field of research on college students’ entrepreneurial activities, so far, no scholars have used empirical methods to analyse the “black box” of the mechanism of IDRI’s influence on college students’ entrepreneurial performance, and there is no better solution to solve the above problems. Therefore, it is necessary to open the “black box” of IDRI and college students’ entrepreneurial performance, effectively clarify the internal mechanism, clarify the problem blocking points, and on this basis, extract the optimal plan and policy choice for the benign interaction between industrial integration and college students’ high-quality entrepreneurial cooperation. Compared with the existing research results, the contribution of this study is that this group spent 3 years (2022–2024) to conduct a large-scale questionnaire survey on returning entrepreneurial college students in 11 provinces in China, and obtained relevant and valid data. Based on these data, this paper adopts an empirical approach to clarify the influence mechanism of IDRI on the entrepreneurial performance of returning college students, and accordingly results forward suggestions to effectively improve the entrepreneurial performance. This study not only further enriches the existing theories of entrepreneurship, but also provides scholars with references on research methodology, and at the same time provides reference bases to help government departments adjust relevant industrial development policies and entrepreneurship policies.

Literature review

Research on the theory of integrated development of rural industries

Naraomi Imamura, a professor of the University of Tokyo, has already carried out theoretical research on the integration of agricultural industry. In the 90 s of the last century, he put forward the theory of the “sixth secondary industry”, the main content of which is to effectively stimulate farmers to carry out diversified activities and extend the agricultural industrial chain through the expansion of agricultural functions and the mutual extension of primary, secondary and tertiary industries (Naraomi and Keijuro, 1996), increase the added value of agricultural products, and then increase farmers’ income. This theory is also known as the “sixth secondary industry” because the multiplication and sum of the numbers 1, 2, and 3 in the primary, secondary and tertiary industries is 6. Subsequently, more and more scholars from different disciplines began to study the theoretical framework and influencing factors of industrial integration. Some scholars put forward the theory of “value chain integration”, which emphasizes that the value chain between different industries forms a new value chain through integration, so as to promote the development of industrial integration (Grossman and Helpman, 2002; Morosini, 2002); while some scholars found that policy support, technological innovation, and market demand are important factors to promote industrial integration, according to the study of specific industrial integration cases.

A large number of existing studies show that the mutual integration and development of rural industries is conducive to promoting the development of rural economy, society, and culture. The integration within the agricultural industry also has a great impact on the depth of development of rural areas and the stability of rural society. Most scholars believe that the IDRI has effectively extended the agricultural industrial chain, increased the added value of agricultural products, and promoted the increase of farmers’ income (Xiaoxiao, 2022). In recent years, more research results have shown that industrial integration leading the rural economy as an integral system can affect the sustainable development of multiple rural industries, and the integration and development of agriculture and other industries is more conducive to the success of entrepreneurs. For example, some scholars have proved through empirical analysis that the integrated development of agriculture and tourism not only significantly promotes the development of agricultural industries, but also significantly promotes the development of local tourism. Other scholars have verified through literature review and in-depth interviews that the integration of agriculture and cultural industries can not only enhance people’s sense of cultural identity, but also promote cultural heritage.

Theoretical study of CSREE performance

Previous research has found that IDRI is not only an effective model for promoting rural economic and social development, but also an effective way to improve the competitiveness of the agricultural industry and the income level of rural residents. Then, there is almost no literature to study these issues, such as whether IDRI will have an impact on the performance of CSREE, what kind of impact, how the influence mechanism and so on. The representative research results of enterprise performance evaluation in the existing literature are reflected in three aspects.

Research on the indicators and methods of enterprise performance evaluation. The theory and method of enterprise performance evaluation originated from business activities (Arena et al., 2015). The performance measurement system (PMS) was originally established from the perspective of enterprise profit (Speckbacher et al., 2003). PMS, a set of indicators to quantify the efficiency and effectiveness of actions (Neely et al., 2002), mainly includes four main stages (Bagnoli and Megali, 2011): Phase I (1920–1950): PMS focuses on limited decision areas, with particular attention to cost and efficiency (Bititci et al., 2011); Phase II (1950–1960): PMS expands its scope to departments and departmental budgets, but still pays special attention to economic and financial performance (Otley, 2003; Bititci et al., 2011); Phase III (1960–1980): PMS integrates new performance dimensions: quality, time, flexibility and customer satisfaction (Slack, 1983); Phase IV: PMS goes beyond the boundaries of the firm and takes more account of stakeholder impacts (Marchand and Raymond, 2008; Bititci et al., 2011).

Definition and measurement model of CSREE performance. CSREE performance refers to the economic, social, and environmental impacts and values generated by these entrepreneurial activities (Taouab and Issor, 2019). Researchers based on different research perspectives, their definition and understanding of CSREE performance are slightly different, there are two mainstream views: single-dimensional and multi-dimensional. The single-dimensional definition mainly focuses on the economic benefits of the enterprise, such as profit, sales, market share, etc. (Taouab and Issor, 2019), while the multi-dimensional definition considers both the social and environmental benefits of the enterprise, such as job creation, social responsibility, resource conservation, etc. These two definitions not only reflect different value orientations and evaluation criteria, but also influence the selection and application of subsequent measurement models. There are also various CSREE performance measurement models, such as the balanced scorecard, the performance prism (PP), and the performance pyramid (Taouab and Issor, 2019). These models have their own advantages and disadvantages, and need to be rationally selected and applied according to the specific research purpose and data characteristics.

The influencing factors and mechanism of CSREE performance. Most scholars believe that CSREE performance is influenced by many factors, such as personal characteristics, entrepreneurial motivation, entrepreneurial resources, and so on. These factors can be classified and analyzed from the individual level, the organizational level, and the environmental level, and can also be distinguished and compared from the perspective of internal and external factors; these factors affect enterprise performance through different paths and ways, such as technological innovation capability, cooperative development capability, resource acquisition capability and so on. These paths and ways can be distinguished and measured in terms of direct and indirect effects, and also identified and adjusted in terms of positive and negative effects. Different researchers use different methods to study these factors and mechanisms, such as empirical analysis, case study, structural equation model (Taouab and Issor, 2019) and so on. These methods have their own characteristics and limitations, and need to be appropriately selected and applied according to the specific research problem and data source.

Based on the above research results, the index system of IDRI and CSREE performance is constructed, and then the structural equation model is used to study the influence degree and mechanism of China’s rural industrial integration on CSREE performance, which is a kind of innovation in theoretical research, and the research methods and results can also provide reference for other scholars’ related research.

Methods

Structural equation model

SEM is a multivariate model that integrates multiple statistical methods. Its idea originates from the Genetic Path Modeling theory put forward by Wright S in the 1920s and 1930s. The path analysis method is based on the linear regression model, and reasonably explains the relationship between variables by establishing the path structure of “cause-result” which is consistent with the observed data. At that time, this method was mainly used to model the indirect relationship between dominant variables in the field of biology (Wright (1918); Wright (1920); Wright (1934)). The path analysis model solves the problem that the general regression model cannot deal with more than one independent variable and intermediate variables, but this method has some defects in use: path analysis assumes that there is no measurement error in variables; it can only deal with the causality of observable explicit variables, but not potential variables. Later, Karl Gustav Jöreskog, a Swedish statistician, improved on the basis of genetic path modeling and put forward the SEM theory (Jöreskog (1969); Jöreskog (1970); Jöreskog (1973)). SEM is an improvement of regression analysis and path analysis. Its advantages are: (1) introducing potential variables. There are no potential variables in regression analysis and path analysis, but SEM can include several potential variables and their measured variables in the same model to study the structural relationship between them. (2) SEM is also similar to the path analysis model solved by simultaneous equations, but SEM relaxes the restrictions of the model. At the same time, it allows independent variables and dependent variables to have measurement errors, and its adaptability is broader; (3) SEM overcomes the shortcomings of too many basic hypotheses in path analysis, inability to include potential variables, and inability to deal with reciprocal causality. Therefore, this paper adopts the SEM to analyse the structural relationship between IDRI and CSREE performance.

The basic formula of SEM is:

$$\eta ={\rm B}\eta +\varGamma \xi +\zeta$$
(1)

Where, \(\eta\), \(\xi\) and \(\zeta\) represent the endogenous latent variable, exogenous latent variable and residual term respectively, and the effect coefficient matrix of the sum of vectors respectively.

$$y={\varLambda }_{y}\eta +\varepsilon$$
(2)
$$x={\varLambda }_{x}\xi +\delta$$
(3)

Equations (2) and (3) are used for the measurement of the model, in which the endogenous latent variable η is measured y; exogenous latent variable vector ξ is measured by x; \({\varLambda }_{y}\) and \({\varLambda }_{x}\) represent the regression coefficient (factor load matrix) between latent variables and measured variables; \(\varepsilon\) and \(\delta\) are residual terms.

Variable design and description

Eight variables, including extension of the agricultural industry chain (EAIC), multi-functional expansion of agriculture (MFEA), internal integration of the agricultural industry (IIAI), technological innovation capability (TIC), resource acquisition capability (RAC), cooperative development capability (CDC), entrepreneurial survival performance, entrepreneurial growth performance, are selected as latent variables. Latent variables cannot be observed directly. According to the characteristics of latent variables, a total of 26 available measurement variables are designed. Based on the combing and integration of the existing research results, the scale of this study is designed. After combing through existing research, and several rounds of expert guidance, three parts of the scale are formed through several rounds of discussion on this basis: the first part is independent variable, including EAIC, MFEA, and IIAI, with nine items in total, to objectively evaluate the degree of IDRI; The second part is the intermediate variables, including RAC, TIC, and CDC three dimensions, a total of nine items; The third part is the dependent variable, including two aspects and a total of eight items, to evaluate the survival performance and growth performance of start-up enterprises. All scales were scored using 5-point Likert scale, there are 5 different levels, and the higher the value is, the higher the degree is.

Setting the independent variables

Independent variables are mainly used to measure the level of IDRI. The existing literature shows that the degree of IDRI can be measured by three indicators: EAIC, MFEA, and IIAI (Xiaochun and Dan, 2020; Zhao and Lv, 2023; Cui et al. (2021)). The EAIC index is the measurement of the vertical dimension of industrial integration (Yuanyuan and Yan, 2021), which is also called “the length of industrial integration” in this paper. The MFEA index is the measurement of the horizontal dimension of industrial integration (Marsden and Sonnino, 2008; Gulickx et al., 2013), also referred to as “the width of industrial integration” in this paper. The IIAI index is the measurement of the depth of industrial integration, which is also called “the depth of industrial integration” in this paper. In terms of measuring the degree of industrial integration development, the commonly used methods are patent coefficient method, Herfindahl index method and input-output method. Among them, the first two methods need to approximate the degree of industrial integration through the corresponding data to measure the technological integration among industries. The input-output method mainly measures the integration formed by industrial penetration, but cannot measure the complex industrial integration formed by industrial intersection and restructuring.

Determining the mediating variables

According to the existing research, IDRI not only has a direct impact on the performance of CSREE, but also has an indirect impact on the performance of CSREE through some intermediate variables. At present, most scholars argue that it is reasonable to use technological innovation capability, resource acquisition capability and cooperative development capability as intermediary variables of the impact of IDRI on enterprise performance (Teece, 2014; Mcafee, 2009; Kotey and Meredith, 1997). The evaluation of technological innovation capability of enterprises is generally reflected in the degree of technological innovation, the advanced degree of technological indicators, and the role of technological innovation in promoting scientific and technological progress and improving market competitiveness (Songqiang, 2017). Resource acquisition capability is a complex multidimensional concept, which is generally measured from the perspective of entrepreneurial cognition, that is, how much attention enterprise leaders pay to enterprise resource acquisition capability. Cooperative development capability can be divided into internal cooperation capability and external cooperation capability. Internal cooperation is a cooperative behavior between departments within an enterprise; external cooperation capability mainly refers to the ability to achieve partnership between enterprises.

Setting dependent variables

Dependent variables are mainly used to measure the performance of CSREE. According to the existing literature, enterprise performance can be measured by two indicators: survival performance and growth performance (Venkatraman, 1989; Lee and Wong (2015)). Enterprise survival is the basic dimension of enterprise performance, survival performance index measures the uncertainty of enterprise success, sustainable management is an important direction of enterprises, and growth performance indicators are the measure of the stability of enterprise development (Dewangan and Godse, 2014). In this paper, performance is measured from two aspects: survival performance and growth performance of student start-up enterprises. The main measurement indicators include enterprise market share, survival time, net profit, cash flow satisfaction, sales growth rate in the past 3 years, enterprise scientific and technological innovation capability, green organic agricultural product production quantity compared with the main competitors.

The designed variable system is shown in Table 1.

Table 1 Variable meaning and measure description.
Full size table

Theoretical basis and hypothesis

Direct effect hypothesis of rural industry integration development on college students’ entrepreneurial performance

Based on the above analyses, the following hypotheses are proposed, as shown in Table 2. The process of the IDRI is the process of CSREE to continuously improve their performance, and the IDRI is an important driving force for the further development of CSREE. In the entire agricultural industrial chain, the IDRI is a booster to improve agricultural productivity, and an important support for CSREE to achieve profit increase, agricultural modernization, and rural revitalization (Istiqomah and Adawiyah, 2018). Therefore, the IDRI will play a positive role in promoting the performance of CSREE.

Table 2 Hypotheses of direct effect.
Full size table

Mediation effect hypothesis of rural industry integration development on college students’ entrepreneurial performance

Technological innovation capability, which integrates organizational capability, adaptability, innovation capability, and information acquisition capability, is a more comprehensive and systematic capability (Enos and Park, 1988). The development of the industrial integration enables the agricultural industry to obtain production technology and form new technologies more smoothly within and between industries, thus improving the technological innovation capability of enterprises. Ding (2021) discussed the relationship between IDRI and scientific and technological innovation through the vector autoregressive model (VAR) and concluded that the development of the industrial integration will promote the investment in scientific and technological innovation of enterprises, and then promote the development of science and technology of enterprises (Wang et al., 2024). Barney believes that technological innovation capability is conducive to maintaining the competitive advantage of enterprises, thus improving profitability (Barney, 2001). In recent years, academics have used a variety of methods to conduct a lot of research on the relationship between technological innovation capability and enterprise performance, and most scholars believe that the relationship between them is positive.

Entrepreneurial resources are the collection of resources for the development of enterprises. Entrepreneurial resource acquisition refers to the behavior of acquiring resources through certain means in the process of enterprise establishment and development. According to the definition of three-industry integration, the process of three-industry integration includes the process of optimizing the allocation of resources and other elements (Wei et al., 2023). Entrepreneurial resources are one of the necessary elements of entrepreneurial activities that determine the success of enterprises in the market. Cao and Nie (2021) concluded that corporate resources play a role in promoting corporate performance through a questionnaire survey. Many studies have shown that entrepreneurial resources are important factors affecting entrepreneurial performance (Jiang, 2015). The more resources college student entrepreneurs get, the greater the probability of entrepreneurial success.

Cooperation capability is the capability of an enterprise to integrate internal and external resources. Coordination activities across functions and across organizational boundaries are extremely difficult because of goal conflicts and fierce competition for scarce resources, and cooperation is an effective way to break through the bottleneck of its own resources (Arranz et al., 2017). In the process of the development of the integration of the three industries, it is helpful to break down the barriers between various industries, expand the scope of cooperation among enterprises, increase the duration of cooperation among enterprises, and make agricultural production, processing, sales, and other service industries closely linked and coordinated development (Wang and Pan, 2015). Social network theorists believe that the initial establishment of smaller enterprises due to the scarcity of internal resources, good cooperation is a tool for start-ups to create competitive advantage (Liu et al., 2020).

Based on the above view, the following hypotheses are put forward, as shown in Table 3.

Table 3 Mediation effect hypothesis of rural industry integration development on college students’ entrepreneurial performance.
Full size table

Data sources

Sample selection and data sources

The data used in this study are all from interviews with Chinese CSREE business owners or key managers between June 2022 and April 2024. The 11 provinces involved in the survey are: Jilin and Hebei provinces in northern China; Shandong, Henan, Anhui, Jiangsu, and Zhejiang provinces in central and eastern China; Sichuan and Guizhou provinces in western China; and Guangdong and Fujian provinces in southern China. The provinces surveyed cover the main grain and vegetable production areas in China, as well as the main areas of agricultural production and college students returning home to start businesses in China, so the survey samples are highly representative. In the actual sampling process, according to the demonstration enterprise list published on the government website and the CSREE list provided by government departments, stratified sampling is carried out according to the overall distribution characteristics of enterprise size and industry types, so as to reduce the sample selectivity deviation as much as possible. In addition, the research group conducted a pre-survey before the formal survey, and adjusted the questionnaire items according to the results to further improve the quality of the survey. 14 typical CSREEs from Anhui, Hebei, and Sichuan provinces were selected for the pre-survey, and 124 valid questionnaires were actually collected. On this basis, the experience was summarized and the questionnaire was optimized and revised. The formal research relies on the support of the National Bureau of Rural Revitalization and other relevant departments, and the research group investigates the basic situation, internal development and performance of China’s CSREE tri-industry integration through field research, online questionnaires, online interviews, and voice calls. Since all questionnaires in this study were self-reported by business owners, in order to avoid the influence of common methodology bias, a three-stage survey method was adopted in this study, and business owners participating in the survey were asked to leave the last four digits of their mobile phone numbers in order to establish correlation between the three survey results. The questionnaire was distributed and collected for three times, each with an interval of 5 weeks. In the first stage, questionnaires on the integration of control variables and IDRI were collected, a total of 950 questionnaires were sent out and 883 were recovered. In the second stage, questionnaires on TIC, RAC, and CDC were collected. Questionnaires were distributed to business owners who participated in the first survey, and 740 questionnaires were collected. In the third stage, questionnaires on survival performance and growth performance were collected. Questionnaires were distributed to business owners who had participated in the first two surveys at the same time, and 689 questionnaires were collected. The questionnaires of business owners who went bankrupt or changed their jobs and those that were not completed were manually reviewed and eliminated, and the questionnaires that failed to pass the logical rationality test were eliminated by SPSS software. Meanwhile, for those business owners who did not complete the questionnaires, the parties concerned were contacted for supplementary investigation as far as possible. After screening out invalid questionnaires and eliminating abnormal values, the questionnaire samples were filled in regularly, 611 valid data samples were obtained, with an effective rate of 88.68%.

Description of sample characteristics

Among the 611 CSREE surveyed, 211 enterprises are mainly engaged in planting development, 159 enterprises are mainly engaged in aquaculture development, 120 enterprises are engaged in agricultural products processing, 90 enterprises are engaged in trade and service industry. In terms of the distribution of business scale, most of the sample enterprises are “small” (the capital scale is less than 1 million RMB), accounting for 50.90%; the proportion of medium-sized enterprises (capital scale is 1 million-5 million RMB) is 30.61%; the proportion of large enterprises (capital scale is more than 5 million RMB) is 18.49%. In the sample, start-up enterprises account for 12.44%, growth enterprises account for 42.55%, mature enterprises account for 30.44%, and enterprises in the transition stage account for 14.57%.

Reliability test

The reliability of 611 samples was tested by SPSS software, and the internal consistency index Cronbach’s Alpha value of the questionnaire was obtained.

The results showed that the overall Cronbach’s alpha coefficient value of the scale was greater than 0.7, and the Cronbach’s alpha coefficient value corresponding to the eight dimensions was greater than 0.7, indicating that the reliability of the questionnaire was good and the internal consistency was good.

The corrected item total correlation (CITC) is an index that measures the correlation between a single item and the whole. If the correlation coefficient is less than 0.3, the item does not meet the condition. According to the reliability test and analysis, the correlation coefficients of CITC in the questionnaire are all more than 0.3, which indicates that each item is related to the whole and the discrimination is strong.

Validity test

It is necessary to determine that the correlation between the original variables is strong before using factor analysis, so KMO and Bartlett values are selected to test the sample data. The value range of KMO test results is 0–1. The closer the value is to 1, the stronger the correlation. It is generally believed that factor analysis can be done when the KMO value is greater than 0.6. The purpose of the Bartlett test is to detect whether the data come from a population that follows a multivariate normal distribution (Yang and Xiangyi (2022)). In this study, the KMO value is 0.805 (Table 7) > 0.8, the Bartlett test approximates the chi-square value of 6704.184, the degrees of freedom is 325, and the P value is 0.000 < 0.01, which passes the significance test. Based on the above results, the data of the scale are suitable for factor analysis.

Factor analysis was used to measure the construct validity of the scale. The common factor is extracted according to the principle that the eigenvalue is greater than 1, and the variance interpretation rate values of the eight factors were 10.424%, 10.303%, 8.801%, 8.746%, 8.703%, 8.677%, 8.610%, and 8.349%, respectively, indicating that the eight common factors had good interpretation of the original data. In addition, the rotational component matrix (Yang and Xiangyi (2022)) is obtained by using principal component analysis and Kaiser standardized maximum variance method according to the existing literature. According to the meaning of the item in the scale and the rotation component matrix, the load value is more than 2016, which indicates that it can be analyzed as an important item. At the same time, the result of the rotation component matrix is consistent with the scale and measurement dimension designed by the questionnaire, so the validity of this questionnaire is high.

Correlation analysis and differential validity

Correlation analysis and discrimination validity is also one of the important criteria to test the rationality of the design between various dimensions of the model (Podsakoff et al., 2012), and Table 4 is obtained after analyzing the latent variables of the model. Table 8 shows that there are significant correlations between seven items: survival performance, growth performance, EAIC, MFEA, IIAI, RAC, TIC, and CDC (P < 0.01); The absolute values of the correlation coefficients of each latent variable are less than the square root of the corresponding mean variance mean. This indicates that there is a certain correlation among the latent variables and a certain degree of differentiation among them, which indicates that the discriminant validity of each dimension of the scale data is ideal and there is no multicollinearity among the variables, which can be carried out for subsequent empirical testing and analysis.

Table 4 Pearson correlation.
Full size table

Empirical analysis

Model fit test

In order to test whether the theoretical model construction is applicable to the actual survey results, the model suitability is first tested, and the results are shown in Table 5. The results show that the root mean square (RMSEA) value of approximate error is 0.047, less than 0.08, the chi-square value (CMIN) is 650.717, the degree of freedom (df) is 275, and the normalized chi-square value (CMIN/DF) is 2.366, less than 3. On the whole, EAIC, MFEA, IIAI, RAC, TIC, CDC, and CSREE survival performance and growth performance overall model adaptability is relatively good.

Table 5 Model adaptability.
Full size table

Direct effect test

Path analysis was performed using AMOS software, structural equation models were used to test the above hypotheses (Ramadani et al., 2022), and the size and significance of the path coefficients between variables were analyzed by Maximum likelihood. Firstly, Bootstrap, that is, the uniform sampling method with put back, is employed to calculate and process the collected 611 sample data, and the model path is fitted, and the path coefficient plot is obtained as shown in Fig. 1.

Fig. 1: Test results for structural equations based on the bootstrap method.
figure 1

The path coefficients shown in the figure are standardized path coefficients with two decimal places; A dashed line indicates that the path test failed.

Full size image

On this basis, the models of EAIC, MFEA, IIAI, RAC, TIC, CDC, and CSREE survival performance and growth performance are clarified, and the path coefficients are sorted from largest to smallest as shown in Table 6.

Table 6 Path test results.
Full size table

Table 6 shows that EAIC, MFEA, IIAI should play a significant positive role in promoting the main effect of CSREE survival performance and growth performance. At the same time, the development of the industrial integration plays a positive role in the resource acquisition capability, technological innovation, and cooperative development of CSREE, indicating that hypotheses H3a, H3b, H3c, H4a, H4B, H4c, H5a, H5b, and H5c are supported. From the effect of the integrated development of China’s rural industries, IDRI can not only directly promote the performance of enterprises, but also improve the ability of technological innovation within enterprises, attract professionals, retain professionals, and then improve the opportunity and level of technological innovation, promote industrial cooperation, integrate multiple resources, improve the ability of cooperative development, and constantly give birth to integrated new technologies. From the results of the influence of intermediate variables on enterprise performance, technological innovation capability, and cooperative development capability have a significant positive influence on enterprise performance, and the hypotheses of H4d, H4e, Hb1, Hb2, Hb3, Hb4, Hb5, Hb6, H5d, H5e, Hc1, Hc2, Hc3, Hc4, Hc5 and Hc6 are acceptable. Table 6 also shows that the path coefficients of path 17 and path 18 are insignificant, indicating that the ability to acquire enterprise resources has no significant impact on CSREE performance, and the hypotheses Ha1, Ha2, Ha3, Ha4, Ha5, and Ha6 are rejected. From the perspective of improving enterprise performance, technological innovation capability and cooperative development capability play an important role, but the direct effect of resource acquisition capability is not significant.

Intermediary effect test

Three indicators of RAC, TIC and CDC are selected as intermediate variables, and the effect value and confidence interval of each intermediate path are obtained by bootstrap operation to test whether they play an intermediate effect between the development of the industrial integration and CSREE performance. The test results are shown in Table 7, which shows that the intermediary effect of technological innovation capability and cooperative development capability is significant at the 5% level, but the intermediary effect of resource acquisition capability is not significant. According to the data in Table 7, Tables 8 and 9 can be obtained after calculation. Table 8 shows that IIAI has the greatest impact on CSREE performance, with a path coefficient of 0.306; followed by MFEA, the influence of the path coefficient is 0.261; the influence of the path coefficient on CSREE performance is 0.258, with a path coefficient of 0.261. Table 9 shows that the intermediary effect of the integrated development of China’s rural industries on enterprise performance through technological innovation capability is 0.100188, and that on enterprise performance through cooperative development capability is 0.110838. Therefore, the intermediary effect value of CDC is higher than that of TIC, which means that the development of the industrial integration plays a key role in improving the survival performance and growth performance of enterprises by improving the vertical integration cooperation, horizontal cooperation, and diversified cooperation of CSREE. In fact, professional technical personnel is the key consideration in the design of variable technological innovation capability and cooperative development capability. The contribution of specialized management and technical personnel to the improvement of CSREE performance is much greater than the acquisition cost of market policy information, social capital and physical capital.

Table 7 Intermediary Effect test.
Full size table
Table 8 Path coefficients of IDRI influence on CSREE performance.
Full size table
Table 9 Path coefficients of mediation variables on CSREE performance.
Full size table

The intermediary effect of the development of the industrial integration on the performance of college students’ returning entrepreneurial enterprises through resource acquisition capability is 0.049714, but it fails to pass the significance test, so the hypothesis of Ha1, Ha2, Ha3, Ha4, Ha5, and Ha6 is not established, which is consistent with the result that resource acquisition capability does not have a significant impact on enterprise survival performance and growth performance in path analysis.

Conclusion and discussion

Conclusion

The structural equation model (SEM) is used to study the impact of IDRI on CSREE performance. Through empirical analysis, the following conclusions are drawn: (1) IDRI does effectively improve the performance of CSREE, which is in line with the “sixth industry” theory. (2) The three dimensions of IDRI, namely, EAIC, MFEA, and IIAI, have a positive direct impact on the survival performance and growth performance of CSREE, and the effects are significant. Among them, IIAI has the greatest direct and comprehensive impact on CSREE performance, followed by MFEA and EAIC. (3) In the process of China’s IDRI, the intermediary effect of technological innovation capability and cooperative development capability is significant. Among them, cooperative development capability has the greatest intermediary effect on CSREE performance, followed by technological innovation capability, while the intermediary effect of resource acquisition capability is not significant. (4) The direct influence of the three indicators of IDRI on CSREE performance is greater than the indirect influence. According to Tables 6 and 7, the direct influence path coefficients of the three indicators of IDRI on CSREE performance are significantly higher than the indirect influence path coefficients of two intermediary variables, indicating that IDRI can directly promote the improvement of CSREE performance.

Based on the above conclusions, the following policy recommendations are given. (1) IDRI should be promoted in an all-round way to provide a better external environment for college students to return home to start a business. The government should make overall planning, identify the local leading industries, synchronously promote the development of EAIC, MFEA and IIAI, build industrial competitiveness, and improve the success rate of college students returning home to start a business. (2) the development of IIAI should be promoted. From the above empirical results, the performance improvement effect of IIAI enterprises is the most significant, so guiding college students to choose the industrial development model of the combination of planting and breeding when they return home is not only conducive to promote the internal integration and development of the primary industry, but also help to improve the performance of enterprises. (3) TIC and CDC of college students’ entrepreneurial enterprises should be greatly improved. Under the background of the integrated development of agricultural industry, improving the technological innovation capability and cooperative development capability of agricultural enterprises is helpful to effectively improve the performance of agricultural enterprises. When college students return to their hometown to start a business, they should fully do a good job in market research, give priority to entrepreneurial projects with high technical level, and introduce scientific and technological industries with a high degree of innovation, such as informationization, intelligence, and so on, which can not only effectively enhance the willingness of college students to return home for entrepreneurship and employment, but also help to promote the overall transformation of traditional agriculture to modern agriculture. At the same time, it is suggested that government departments should give full play to government service functions, take the initiative to provide more and better cooperation opportunities for CSREE, strengthen multi-enterprise, multi-industry, and multi-department cooperation and collaboration, give full play to the mediating effect of technological innovation and cooperation development ability on the performance of CSREE, and take more measures to improve the performance level of CSREE in an all-round way.

Discussion

Based on the above empirical evidence of China, IDRI can significantly improve the performance of CSREE, in order to further test this conclusion, the research team obtained the relevant micro-survey data of the United States, Canada and Japan from 2018 to 2022, and used the SEM model for calculation. From the empirical results, promoting the IDRI can improve the entrepreneurial performance of returning college students, which is fully consistent with the research conclusion of this paper and also consistent with the “sixth industry” theory (Naraomi and Keijuro, 1996) proposed by Japanese scholar Naraomi Imamura. This study also found that returning entrepreneurial students’ own technological innovation ability and cooperation development ability have a significant positive impact on the performance of entrepreneurial firms, which is consistent with the findings of Hajdari and Gruenhagen (Hajdari et al., 2023; Gruenhagen et al., 2020). At the same time, there are some differences between the results of this study and those of Hajdari and Gruenhagen. This study found that although the resource acquisition ability of college entrepreneurs has a positive impact on the performance of entrepreneurial ventures, the impact is relatively weak and less significant. However, the results of Hajdari and Gruenhagen’s study showed that the resource acquisition ability of college entrepreneurs has a significant and positive impact on the performance of entrepreneurial ventures (Hajdari et al., 2023; Gruenhagen et al., 2020). After further examination of the data and comparative analyses, it was found that the differences in the policy environment could lead to differences in the results. The Chinese government’s agricultural industry integration development policy, college student entrepreneurship policy, and entrepreneurial financial support provide a lot of resource support for the entrepreneurial activities of returning college students, which to some extent weakens the ability to obtain the resources needed for entrepreneurial activities, and also leads to returning college students need more cooperative development ability and technological innovation ability to achieve better entrepreneurial performance. These findings indicate that the Chinese government’s policies on the integrated development of the agricultural industry and on encouraging college students to return to their hometowns to start their own businesses have played a positive role in stimulating the entrepreneurial activities of college students returning to their hometowns, but the existing policies have not been able to significantly improve the entrepreneurial performance of returning college students. According to the findings of this study, the Chinese government should adjust the policy direction as soon as possible and introduce more policies that are conducive to enhancing the cooperative development ability and technological innovation ability of returning college students, so as to more effectively improve the entrepreneurial performance of college students. For example, more policies should be introduced to promote cooperation among returning college students and cooperation between entrepreneurial college students and related enterprises to achieve joint entrepreneurship and co-creation of entrepreneurship, so as to enhance the cooperative development ability of returning college students. Establishing an entrepreneurial mentoring system, recruiting experienced entrepreneurs and related technical experts as entrepreneurial mentors to provide technical guidance and market guidance to returning college students, so as to improve the technological innovation ability of returning college students. Considering the differences in resource endowment, social system, political system and humanities in different countries and regions, it is suggested that the proposed research framework and methods can be used to learn from the proposed research framework and methods when studying the development of local industrial integration, and more effective variable indicators can be selected for empirical research according to local characteristics, which will further improve the accuracy of research results.

Limitations and future research

This paper focuses on the mechanism and extent of the impact of IDRI on the performance of entrepreneurial ventures of returning college students in China, and proposes relevant policy recommendations based on the findings of the study. In reality, improving the performance of college students’ entrepreneurship requires not only the support and participation of the government, but also the participation of higher education institutions, social organizations, and returning university students. However, limited by the length of the article and availability of data, this paper does not explore the impact of higher education institutions, social organizations, and students’ own characteristics on entrepreneurial performance. Therefore, after completing this paper, our team is now working on a new research project, Reform of Entrepreneurial Ability Cultivation Mode of College Students in Higher Education Under the New Situation, which focuses on further improving and optimizing the teaching methods, teaching contents and talent cultivation programs in higher education institutions under the new situation and new environment, in order to improve the entrepreneurial capabilities and performance of college students. At present, the project has entered the field research stage, and the team looks forward to cooperation and support from colleagues and friends.