Study on Social Integration Identification and Characteristics of Migrants from “Yangtze River to Huaihe River” Project: A Time-Driven Perspective

Abstract

The “Yangtze River to Huaihe River” project is another large-scale trans-basin water diversion project in China, following the South-to-North Water Division Project. The reservoir project is in the period of China’s rapid development of modernization, informatization and marketization. During this period, reservoir-induced migrants have great differences in resource endowment and development opportunities, so they show different characteristics in the integration process. In order to clarify these characteristics and formulate corresponding policies based on these characteristics to help them integrate into the society of the resettlement area as soon as possible, this paper has completed two aspects of work: First, a method for identifying social integration is proposed. Second, the characteristics of social integration of different migrant groups in different time dimensions are analyzed. This study has two important findings: First, the extent of information sources and understanding of policies has a greater impact on the social integration of migrants. The wider the source of information and the more understanding of policies, the more it helps migrants to integrate into society. Second, the social network relationship established by the female migrant group in the maintenance activities will greatly promote their social integration.

1. Introduction

Since the reform and opening up, China’s high-speed economic development has led to large-scale construction projects, such as dams, highways, and inter-basin water diversion projects, which have resettled about 50 million people [1]. These migrants are all internal migrants. Reservoir-induced migrants, as the largest group of project-induced migrants in China, have always attracted the attention of scholars [2,3,4,5,6].

As reservoir projects are mostly built in vast rural areas, and the reservoir-induced migrants are mainly farmers, they generally have a low level of education and skills, which makes their income levels generally low [7]. With the proposal of the precision poverty alleviation policy in 2013, China has carried out quantitative management of the income of the poor [8], and this management method has effectively improved the economic income of the migrants [9]. Although the economic income has been greatly improved, it is difficult for some migrants to integrate into the new social environment, which has led these migrants to collectively return to the area surrounding the reservoir [10]. Therefore, the issue of social integration of reservoir-induced migrants cannot be ignored.

At present, there is no uniform identification method for the social integration of migrants, that is, there is no systematic method to measure whether the migrants are integrated into the resettlement community. Scholars generally believe that economic integration is the basis and premise of social integration [11], so some scholars use economic indicators to identify the social integration of migrants. Huang et al. used income level as a standard to analyze the social integration. When the economic income of migrants is close to the indigenous people, it is considered that the migrants are integrated into the society of the resettlement area [12]. By constructing comprehensive economic indicators, Wang et al. used cluster analysis to identify the social integration of migrants [13]. However, in fact economic integration cannot be equated with social integration [14]. Some scholars consider that social integration is a psychological state that can be determined by the intuitive feelings of migrants [15], so they use questionnaires based on psychological awareness to quantify social integration [16]. However, in the actual process, the respondents’ answers may not reflect their actual situation given the possibility of biases in self-reported data. In addition, some scholars think that policy factors [17,18] and cultural factors [19,20] will also have a great impact on the social integration of reservoir-induced migrants. At the same time, environmental factors [21,22] and geopolitics [23,24] contribute to the sustainable development of migrants. Therefore, in identifying the social integration of migrants, multiple dimensions need to be considered.

The “Yangtze River to Huaihe River” project is a large-scale inter-basin water transfer project focusing on urban and rural water supply and development of Jianghuai shipping, combined with irrigation water supply and improvement of the Chaohu Lake and Huaihe River water ecological environment. The water supply scope of the “Yangtze River to Huaihe River” project covers 12 cities in the Anhui Province and 2 cities in the Henan Province, with a total of 55 districts and counties, covering an area of about 70,600 square kilometers. On the morning of December 29, 2016, the construction of the Yangtze River and Huaihe River project was officially started, and all reservoir-induced migrants were resettled. The project involved 79,465 migrants [25]. Unlike previous reservoir-induced migrants, the “Yangtze River to Huaihe River” project is under a special era background. During this period, China achieved rapid development of modernization, informatization and marketization, which resulted in greater heterogeneity in resource endowment among different migrant groups, and also caused uneven development opportunities [26]. Therefore, different migrant groups have great differences in the process of social integration [27,28]. In order to help the migrants of different individuals to integrate into the life of the resettlement area, it is necessary to analyze their integration characteristics, so as to be able to propose more effective plans to promote the social integration.

The development of time-geography theory provides a new idea for the study of social integration [29]. Some scholars began to analyze the relationship between people and regions using time as a dimension in the process [30,31]. Cumming proposed the concept of social exclusion in the time dimension, namely: limiting residents’ activities by limiting residents’ time budgets [32]. Based on Cumming, Hao et al. proposed the concept of spatial-temporal exclusion, that is, due to some special reasons, the self-dominant time of some social groups is deprived, and the activity spaces is closed and narrow [33]. Wang et al. systematically studied the spatial-temporal exclusion characteristics of urbanized migrants from 2011 to 2015 and their influencing factors [34]. Social integration is a time dependent process, and it changes over time. In order to clearly reflect this process, it is necessary to introduce the time dimension into the study of social integration.

To sum up, there is no unified social integration identification method for reservoir-induced migrants. Therefore, this paper proposes a method to quantitatively identify the social integration. Based on the grounded theory, the method uses structural equation model (SEM) and cluster analysis to identify the social integration status of the migrants from “Yangtze River to Huaihe River” project, and compares the recognition results with the results of the psychological integration questionnaire to prove its effectiveness. In addition, this paper uses the cox proportional hazards model to analyze the social integration characteristics of different migrant groups in different time dimensions, in order to help them integrate into the resettlement community faster in policy formulation.

2. Materials and Methods

2.1. Study Area

This paper studies the social integration of migrants from “Yangtze River to Huaihe River” project in the Feixi county, Hefei city, Anhui province, China. In order to obtain valid data and information, the research team conducted an investigation involving a questionnaire on the migrants resettled in Feixi County from January 5 to January 24, 2019. The investigation involved six resettlement areas, as shown in the Figure 1. These migrants lived in the countryside before resettlement. The six resettlement areas were selected for two main reasons: (1) they have a large number of migrants, accounting for 69.81% of the total migrants in Feixi County. Among them, there were 669 migrants from ChangZhen village, 1,607 from BinGuang village, 594 from WuHe village, 758 from ZiPeng community, 1,017 from LiangTing community and 1,269 from LiuHe community. (2) Migrants in the 6 resettlement areas are relatively concentrated, being both common and well represented in those areas.

2.2. Data Collection

In this sampling, only those who are capable of working and aged between 16 and 60 years old are sampled, with students at school and the elderly over 60 years old are not considered. This is due to social integration of migrants is based on the family, and only when the main labor force in the family is integrated into the society can the elderly and children be effectively integrated into the society [35]. A total of 449 samples were taken this time. Among them, ZiPeng community, LiangTing community and LiuHe community are urban resettlement areas, with 51, 68 and 85 people selected respectively for data collection; ChangZhen village, BinGuang village and WuHe village are rural resettlement areas, with 57, 137 and 51 people selected respectively.

In order to facilitate the calculation of cox proportional hazards model, the sample’s age, gender, location, political identity, and Educational background were converted into nominal variables for statistics. In the research on life needs, activities are divided into three types: work activities, maintenance activities and leisure activities [36]. Maintenance activities refer to activities that meet the physiological needs of individuals and families, including personal affairs, daily shopping, housework activities and etc. Leisure activities are social, recreational and other freely-involved activities that are motivated by cultural and psychological needs. Work activities are activities that provide an economic basis for the above two types of activities, and refer to work, business and other activities. Therefore, this paper conducts research from three time-based factors, namely working time, leisure time and maintenance time. The three time dimensions are measured by the daily average time of one month. Descriptive statistical results are shown in

Table 1. Variable Definition and Descriptive Statistics.

Variable	Type	Values	Mean Value	Standard Deviation
Covariate
Age	Nominal	1 = 16–30 years old 2 = 31–45 years old 3 = 46–60 years old	1.964	0.760
Gender	Nominal	1 = Male 2 = Female	1.438	0.496
Location	Nominal	1 = Urban 2 = Rural	1.546	0.498
Political identity	Nominal	1 = Communist Party 0 = Nonparty	0.200	0.401
Educational background	Nominal	1 = Primary school or below 2 = Middle school 3 = High school or above	2.160	0.682
Independent variable
Maintenance time	Scale	Actual (minutes)	120.247	31.517
Working time	Scale	Actual (minutes)	349.962	105.527
Leisure time	Scale	Actual (minutes)	263.613	85.524

Source: date collected from the field investigation.

.

2.3. Methods

2.3.1. Grounded Theory

Grounded theory (GT) is a systematic methodology in the social sciences involving the construction of theories through methodical gathering and analysis of data [37,38,39]. A study using grounded theory is likely to begin with a question, or even just with the collection of qualitative data. As researchers review the data collected, repeated ideas, concepts or elements become apparent, and are tagged with codes, which have been extracted from the data. As more data is collected, and re-reviewed, codes can be grouped into concepts, and then into categories. These categories may become the basis for new theory. Grounded theory approach involves the following basic steps, see Figure 2.

2.3.2. Structure Equation Modeling

The Structural Equation model [40] can be divided into two parts, the Measurement Equation and the Structural Equation. The Measurement Equation mainly represents the relationship between the observed variables and latent variables, and its expressions are shown in (1) and (2).

x = A_x ξ + δ

(1)

y = A_y η + ε

(2)

where, x is the exogenous observation variable vector; y is a vector composed of endogenous observation variables; A_x is the factor load of exogenous observation variables acting on exogenous latent variables; A_y is the factor load of endogenous observation variables acting on endogenous latent variables; δ is the error vector of the exogenous variable x; ε is the error vector of the endogenous variable y.

The structural equation mainly represents the relationship between latent variables, and its expression is usually written as:

η = Bη + Γξ + ζ

(3)

where η is the endogenous latent variable vector, ξ is the exogenous latent variable vector; B is the coefficient matrix of the endogenous latent variable η; Γ is the coefficient matrix of the exogenous latent variable $\mathsf{\xi }$; $\mathsf{\zeta }$ is the residual term of the structural equation, which indicating the part of the model that cannot be explained.

2.3.3. Cox Proportional Hazards Model

Let ${\mathrm{X}}_{\mathrm{i}}=\left({\mathrm{X}}_{\mathrm{i}1},{\mathrm{X}}_{\mathrm{i}2},\cdots ,{\mathrm{X}}_{\mathrm{ip}}\right)$ be the realized values of the covariates for subject i. The hazard function for the Cox proportional hazards model [41,42] has the form:

$$\mathsf{\lambda }\left(\mathrm{t},{\mathrm{X}}_{\mathrm{i}}\right)={\mathsf{\lambda }}_0\left(\mathrm{t}\right)\exp \left({\mathsf{\beta }}_1{\mathrm{X}}_{\mathrm{i}1}+{\mathsf{\beta }}_2{\mathrm{X}}_{\mathrm{i}2}+\cdots +{\mathsf{\beta }}_{\mathrm{p}}{\mathrm{X}}_{\mathrm{ip}}\right)={\mathsf{\lambda }}_0\left(\mathrm{t}\right)\exp \left({\mathrm{X}}_{\mathrm{i}}·\mathsf{\beta }\right)$$

(4)

This expression gives the hazard function at time t for subject i with covariate vector (explanatory variables) ${\mathrm{X}}_{\mathrm{i}}$. The likelihood of the event to be observed occurring with subject i at time ${\mathrm{Y}}_{\mathrm{i}}$ can be written as:

$${\mathrm{L}}_{\mathrm{i}}\left(\mathsf{\beta }\right)=\frac{\mathsf{\lambda }\left({\mathrm{Y}}_{\mathrm{i}}{ |\mathrm{X}}_{\mathrm{i}}\right)}{{\sum }_{\mathrm{j}:{\mathrm{Y}}_{\mathrm{j}}\ge {\mathrm{Y}}_{\mathrm{i}}}\mathsf{\lambda }\left({\mathrm{Y}}_{\mathrm{i}}{ |\mathrm{X}}_{\mathrm{j}}\right)}=\frac{{\mathsf{\lambda }}_0\left({\mathrm{Y}}_{\mathrm{i}}\right){\mathsf{\theta }}_{\mathrm{i}}}{{\sum }_{\mathrm{j}:{\mathrm{Y}}_{\mathrm{j}}\ge {\mathrm{Y}}_{\mathrm{i}}}{\mathsf{\lambda }}_0\left({\mathrm{Y}}_{\mathrm{i}}\right){\mathsf{\theta }}_{\mathrm{j}}}=\frac{{\mathsf{\theta }}_{\mathrm{i}}}{{\sum }_{\mathrm{j}:{\mathrm{Y}}_{\mathrm{j}}\ge {\mathrm{Y}}_{\mathrm{i}}}{\mathsf{\theta }}_{\mathrm{j}}}$$

(5)

where ${\mathsf{\theta }}_{\mathrm{j}}= \exp \left({\mathrm{X}}_{\mathrm{j}}|\mathsf{\beta }\right)$ and the summation is over the set of subjects j where the event has not occurred before time ${\mathrm{Y}}_{\mathrm{i}}$ (including subject i itself). Obviously $0<{\mathrm{L}}_{\mathrm{i}}\left(\mathsf{\beta }\right)\le 1$.

2.3.4. Research Procedures

This study is divided into the following two steps, the flow chart shown in Figure 3.

Social Integration Identification

First, 50 migrants with different occupations and income levels were selected from 449 migrant samples for interviews. The interviews mainly focused on the following four aspects: problems existing in the life after relocation, security mechanisms, compensation methods and government interaction. The data obtained from the interviews were analyzed, classified and summarized using grounded theory to obtain a theoretical framework for social integration analysis. Secondly, based on the analysis framework, relevant questionnaires were designed to construct a structural equation model for the social integration of migrants from “Yangtze River to Huaihe River” project. By normalizing the model’s standardized path coefficients, a comprehensive score calculation formula of social integration was obtained and comprehensive score of 449 migrant was calculated by the formula. Finally, cluster analysis was used to divide the sample’s comprehensive score into two groups. Among them, the high-scoring group was integrated, and the low-scoring group was non-integrated. This result is the experimental group. In order to verify the reliability of integrated and non-integrated migrants identified by the above methods, this paper designed a psychology-based control group questionnaire. The question is: “Do you think you have integrated into the society of the resettlement area?” Questionnaire options are divided into three categories: integrated, not integrated, and uncertain. By transforming the clustering method, when the results of the experimental group and the control group are closest, the results of the experimental group are considered reliable.

Analysis of Social Integration Characteristics in Different Time Dimensions

This article uses the Cox proportional hazards model to analyze the social integration characteristics of different migrant groups in different time dimensions. Among them, age, gender, location, political identity and educational background were used as covariates, and working time, leisure time and maintenance time were used as independent variables.

3. Results

3.1. Social Integration Analysis Framework and Theory Model

In the open coding, 16 concepts are obtained by analyzing the original data of the interview. It can be classified into 5 categories in the axial coding by refining the common features and internal relations of concepts. Based on the research results of other scholars, the connotation of 16 concepts is summarized, as shown in

Table 2. Multi-dimensional analytical framework of social integration.

Categories	Concepts	Connotation of Concepts
Political factors	Political participation	Limited political participation can lead migrants to resist the government and society [43].
	Employment discrimination	The loss of labor rights can cause dissatisfaction of migrants [44].
	Social obligation	Migrants with limited understanding of social obligation may suffer from exclusion and inferiority.
	Compensation benefits	The compensation for land expropriation and relocation of migrants affects their perception of fairness and trust in the government.
Economic factors	Household income	The limited income affects the living standards of migrants, which leads to anxiety and imbalance. [14]
	Revenue sources	Unstable income causes migrants to panic about future life [45].
	Consumption change	Changes in consumption concepts, patterns and structures lead to economic pressures on migrants.
	Household savings	The fear of emergencies arises from the fact that migrants have less household savings [14].
Cultural factors	Spiritual and cultural life	The spiritual and cultural needs affect the life satisfaction of migrants [46].
	Customs change	Migrants’ inheritance of customs influences their psychological identity of life in the resettlement place [47].
	Local behavior standard	Migrants’ maladjustment to local behavior will affect their identity [16].
Life factors	Social interaction	The homogenization of social interaction of migrants affects their psychological satisfaction needs and sense of belonging [48].
	Living environment	The living environment of resettlement area affects the living comfort of migrants [49].
	Activity participation	The limited participation in activities leads to the sense of closure and disappointment [48].
Security factors	Social security	The limited level of social security has caused migrants to feel greater pressure on the sustainable development of livelihoods.
	Skill training	The methods and effects of skills training affect the enthusiasm of migrants to integrate into society and accept identity.

Source: the framework is derived from interviews through grounded theory in the axial coding.

. Then, in the selective coding, the theoretical model of social integration of reservoir-induced migrants is constructed. This model describes the internal mechanism of the factors affecting social integration, as shown in Figure 4.

As can be seen from Table 2, consistent with the relevant research on international migration, political factors include political participation [43] and employment discrimination [44]. The political factors in this paper also include social obligations and compensation benefits. As reservoir-induced resettlement is mandatory, migrants are passive in accepting social obligations of resettlement sites. Therefore, local governments should actively popularize and guide migrants to fulfill the relevant obligations. At the same time, the government has requisitioned a large amount of agricultural land and houses, which requires reasonable compensation. Among the economic factors, household income [14], revenue sources [45] and household savings [14] are commonly used economic indicators. The economic factors in this paper also include consumption change, which can indirectly affect the economic level and therefore cannot be ignored. Cultural factors and living factors are consistent with the analysis of other scholars. This article classifies social security and skill training as security factors, but not as economic factors. On the one hand, it is because the security factors needs to accumulate for a certain period of time before it can affect the economy. For example, migrants must be employed after the training is over; with pension insurance available only after retirement. On the other hand, the correlation between economic factors and security factors is low, which is calculated by the following model.

It can be seen from Figure 4 that the factors affecting the social integration of reservoir-induced migrants can be divided into five categories: political factors, economic factors, cultural factors, life factors and security factors: Political factors are the important rights guarantee for the social integration, involving the basic civil rights and obligations of migrants. The deprivation of migrants’ basic rights will lead to their dissatisfaction and resistance to the government [50]. Migrants’ limited awareness of social obligations and inability to fulfill social obligations will lead to social exclusion [51]. Economic factors are the material basis for social integration [11]. Low income levels or consumption levels that cannot keep pace with the consumption level in the resettlement area will lead to mental imbalance, also leading to limited household savings which will make them feel economically insecure. Cultural factors are the internal driving force of social integration, with cultural maladjustment affecting migrants’ self-satisfaction and identity [15]. Life factors on the other hand are the external driving force of social integration. The living environment affects comfort of migrants in their living spaces, and a low level of living comfort will lead to the second resettlement of migrants [52]. Security factors are the institutional foundation for the social integration, which involves the basic insurance and skill training. Insufficient basic insurance will make migrants lack enthusiasm, and lack of survival skills can cause migrants to lose confidence in their future lives [53]. Political factors have a direct impact on life factors, security factors and economic factors. Security factors have a direct impact on economic factors. Economic factors have a direct impact on cultural factors and life factors. Cultural factors and life factors influence each other.

3.2. Comprehensive Score of Social Integration

Based on the analytical framework of social integration obtained from the grounded theory approach, this paper adopts Likert 5-point scale in questionnaire design. Among them, per capita household deposit, per capita household income, Engel coefficient, pension insurance and medical insurance are continuous variables, which are transformed into five grades by ward cluster method. This paper designed 17 questions according to above 16 factors [Appendix A]. Migrants scored each question, and then the scoring results were input into the second-order structural equation model built by Amos 17. The observed variables with too small a load were removed. The calculation results are shown in Figure 5.

The overall fit of the model was satisfactory but not excellent, χ^2/df = 1.865, p < 0.00 (RMSEA = 0.082, CFI = 0.956, GFI = 0.972, AGFI = 0.906, TLI = 0.943). The p value of Satorra-Bentler scale chi-square goodness test was less than 0.00, indicating that there was a discrepancy between the model and data. However, the argument has been made that testing a perfect fit is not appropriate for social sciences because of the complexity of the causal relationships; rather, it would be better to assess the fit of the model to the data [54]. Steiger developed the root mean square error of approximation and proposed that values below 0.05 indicate an excellent fit and those values less than 0.10 indicate a reasonable fit [55]. Thus, this model’s value of 0.082 indicates that the model fits reasonably well. Similarly, the goodness of fit index is pleasing at 0.972. As it is greater than 0.9, the model in its simple form fits the data in a satisfactory manner.

It can be seen from Figure 4 that political factors, cultural factors and living factors have a great influence on the social integration of the migrants in “Yangtze River to Huaihe River” project, and the most important factors are political factors (0.87), followed by cultural factors (0.83) and then living factors (0.81). Economic factors (0.71) and security factors (0.69) have a relatively small impact on the social integration of migrants. By normalizing the standardized path coefficient of the model, the comprehensive score of social integration can be calculated as follows:

PF = 0.336*PP + 0.320*ED + 0.344*SO

(6)

EF = 0.336*HI + 0.344*EC + 0.320*HS

(7)

CF = 0.326*SC + 0.343*CC + 0.331*LB

(8)

LF = 0.356*SI + 0.318*LE + 0.326*AP

(9)

SF = 0.345*EI + 0.341*MI + 0.315*ST

(10)

Comprehensive Score = 0.223*PF + 0.182*EF + 0.212*CF + 0.207*LF + 0.176*SF

(11)

3.3. Social Integration Identification

After calculation, it is found that the results of experimental group obtained by centroid cluster analysis are the closest to those of the control group. Therefore, the social integration identified by this clustering method is considered to be the most reliable. The results of experimental group showed that there were 362 people in the integrated part and 87 in the non-integrated part. Control group results are shown in Figure 6.

As can be seen from Figure 6, 83.80% of migrants in the integrated part of the experimental group think they have integrated, 9.86% think they have not integrated, and 6.34% are not sure whether they have integrated or not. Among the non-integrated part in the experimental group, 8.05% of the migrants thought they had integrated, 79.31% thought they had not integrated, and 12.64% of the migrants were not sure whether they had integrated or not. The conclusions of the experimental group and the control group are basically consistent.

3.4. The Characteristics of Social Integration in Different Time Dimensions

When using the Cox proportional hazards model for regression, the social integration status of migrants obtained from centroid cluster analysis is taken as the dependent variable; age, gender, location, educational background and political identity are used as covariates; and three time dimensions, namely working time, leisure time and maintenance time, are taken as independent variables respectively. The regression results are shown in

Table 3. Cox Proportional Hazards Model Calculation Results.

	Time Dimension
	Model 1 (Maintaining Time) Exp(B)	Model 2 (Working Time) Exp(B)	Model 3 (Leisure Time) Exp(B)
31–45 years old	1.168 *	1.231 ***	0.809 **
(Ref. category = 16–30 years old)	(2.106)	(3.753)	(2.619)
46–60 years old	0.514 **	0.406 ***	0.435 ***
(Ref. category = 16–30 years old)	(2.785)	(3.613)	(4.372)
Gender	1.996 **	0.443 **	0.517
(Ref. category = male)	(2.958)	(2.398)	(1.159)
Location	0.935	2.323 *	0.287 **
(Ref. category = Urban)	(1.404)	(1.958)	(2.546)
Political identity	1.094 ***	1.729 **	1.546
(Ref. category = Communist Party)	(3.757)	(2.742)	(1.243)
Middle school	0.967	1.301 **	0.740
(Ref. category = Primary school or below)	(0.926)	(2.659)	(1.215)
High school or above	1.488	1.512 **	0.764
(Ref. category = Primary school or below)	(1.653)	(3.020)	(1.715)
Chi-square	59.932	204.07	88.206
df	7	7	7
Sig	0.000	0.000	0.000

Note: T values are in parentheses; *, **, *** correspond to 10%, 5%, and 1% significance levels, respectively.

.

It can be seen from Table 3 that the chi-square values of the three models are 59.932, 204.070, and 88.206, respectively, and their corresponding significance levels are less than 0.00, indicating that all three models are statistically significant. The social integration rate in maintenance activities is significantly affected by age, gender and political identity. The social integration rate in work activities is significantly affected by age, gender, location, political identity and educational background. And lastly, the social integration rate in leisure activities was significantly affected by only the two factors of age and location. The specific analysis is as follows:

3.4.1. Age

Age has a significant impact on the social integration rate in the three time dimensions, as shown in Figure 7 below

From the perspective of maintenance time, the social integration rate of migrants aged 16–30 and 31–45 is basically the same with the change of maintenance time, and migrants aged 31-45 integrate slightly faster than those aged 16–30. Migrants aged 16–30 and 31–45 integrate faster than those aged 46–60. As maintenance time increases, all migrants aged 16 to 60 can integrate into society.

In terms of working time, migrants aged 31–45 are the fastest to integrate into society; followed by those aged 16–30; migrants aged 46–60 were the slowest to integrate. With working time increases, all migrants aged 16–45 can integrate into society, however, 13.74% of the migrants aged 46–60 cannot integrate into society.

From the leisure time point of view, migrants aged 16–30 are the fastest to integrate into society; those aged 31–45 followed; migrants aged 46–60 were the slowest to integrate. Along with leisure time increases, all migrants aged 16–45 can integrate into society, nevertheless, 8.62% of the migrants aged 46–60 cannot integrate into society.

3.4.2. Gender

Gender has a significant impact on the social integration rate in maintenance activities and work activities, as shown in Figure 8 below.

From the perspective of maintenance time, female migrants integrate into society faster than male migrants. As maintenance time increases, all female migrants can integrate into society, however, 1.07% of male migrants cannot integrate into society.

In terms of working time, male migrants integrate into society faster than female migrants. With working time increases, all male migrants can integrate into society, nevertheless, 1.84% of female migrants cannot integrate into society.

3.4.3. Location

Location has a significant impact on the social integration rate in work activities and leisure activities, as shown in Figure 9 below.

In terms of working time, rural migrants integrate into society faster than urban migrants. With working time increases, all rural migrants can integrate into society, nevertheless, 10.26% of urban migrants cannot integrate into society.

From the leisure time point of view, urban migrants integrate into society faster than rural migrants. Along with leisure time increases, both rural migrants and urban migrants can integrate into society.

3.4.4. Political Identity

Political identity has a significant impact on the social integration rate in maintenance activities and work activities, as shown in Figure 10 below.

From the perspective of maintenance time, the social integration rate of migrants who join the CPC (Communist Party of China) and those who do not join the party is basically the same as maintenance time changes. The CPC migrants integrate slightly faster than those who are nonparty migrants. As maintenance time increases, 0.80% of the CPC migrants and 1.02% of the nonparty migrants never integrate into society.

In terms of working time, the CPC migrants integrate into society faster than nonparty migrants. With working time increases, both CPC migrants and nonparty migrants can integrate into society.

3.4.5. Educational Background

Educational background has a significant impact on the social integration rate in work activities, as shown in Figure 11 below.

Migrants with a high school education or above are the fastest to integrate into society; followed by those with middle school education; Migrants with primary school education or below are the slowest to integrate. With working time increases, all migrants with a high school education or above and middle school education can integrate into society, however, 4.76% of the migrants with primary school education cannot integrate into society.

4. Discussion

4.1. Influence of Age on Integration Speed in Different Time Dimensions

The integration rate of migrants aged 46 to 60 in maintenance activities was significantly slower than that of migrants aged 16 to 45. This is most likely due to migrants aged 16–45 have better environmental adaptability, living ability and communication skills than those aged 46–60. On the contrary, migrants aged 46–60 are hardwired and their lifestyles and habits are thought to be more difficult to change [56].

The integration speed of migrants in work activities shows an inverted u-shaped relationship curve with age [26]. Migrants aged 31–45 are the fastest to integrate, with those migrants in this stage in their lives having the best state of both physical strength and energy whilst having more work experience. Migrants aged 16–30 rank the second. Although migrants of this age have certain advantages in learning and accepting new things, they are still in the accumulation period of social capital and economic capital, lack certain experience, and have high work pressure and strong job mobility [57]. Migrants aged 46 to 60 are the slowest to integrate. As migrants of this age grow older, their physical strength and energy gradually decline, which leads to the continuous decline of their competitive advantages in the labor market, making it difficult for them to integrate [58].

The integration speed of migrants in leisure activities is negatively correlated with age. Migrants aged 16–30 are the fastest to integrate. In fact, this is basically in line with the characteristics of this age. Migrants in this age are in the golden period of learning and imitation. In leisure and entertainment, they will find some local like-minded friends according to their hobbies, take the initiative to learn and accept their leisure and entertainment mode [59]. Migrants aged 31–45 rank the second. Migrants of this age group have formed a fixed entertainment and leisure mode in their life course. The traditional leisure and entertainment mode deeply embedded in their hearts has delayed the process of their social integration [60]. Migrants aged 46–60 are the slowest to integrate. Their social, entertainment and activity participation skills are seriously lagging behind, which makes it difficult for them to integrate into the leisure and entertainment life in resettlement areas, and are thus marginalized in social interaction [61].

4.2. Influence of Gender on Integration Speed in Maintenance Time and Working Time

Female migrants integrate into society faster than male migrants in maintenance activities. This may be related to the gender characteristics of women and the people who are in contact with the maintenance activities [62]. Through interviews, we found that small groups of women gradually formed in the resettlement areas. They would buy food together, shop together, and pick up children together. Members of these groups formed their own social network in the process of mutual interaction, communication and mutual assistance, which gradually extended from blood relationship to geographical relationship. In addition, based on the social bond between children and their classmates and teachers, female migrants can contact more people from different industries, thus storing certain social resources to help them integrate into the society faster. However, men are more likely to complete personal tasks as quickly as possible in maintenance activities [63].

Male migrants integrate into society faster than female migrants in work activities. First of all, the division of gender roles has become a common phenomenon in China’s grass-roots society. Most women are engaged in so-called “female jobs” in the service industry, such as housekeeping services, food processing, clothing and shoes making, cleaning and restaurant waiters, which puts them in a weaker position in the labor market competition. Second, for jobs that are suitable for both men and women, employers tend to be more inclined to men to reduce the additional labor costs of perinatal labor security. Finally, as women assume more of the family role of bearing and rising the next generation, they often choose jobs that are close to home and easy to take care of the family. Some women choose periodic employment for the sake of young children, which further puts them in a weak position in the labor market competition [64].

4.3. Influence of Location on Integration Speed in Working Time and Leisure Time

Rural migrants integrate into society faster than urban migrants in work activities. Rural migrants still have land and maintain the original mode of production mainly based on primary industries. Most of them can work directly without labor skill training. However, all the land of the urban migrants was expropriated. Their production methods are forced to shift from the primary industry to the secondary and tertiary industries, and they are mainly concentrated in the secondary labor market. Although the income of this labor market is higher than that of agriculture, the labor market is labor-intensive, with long working hours, fierce competition and high mobility [65]. Therefore, urban migrants integrate slower than rural ones.

Urban migrants however, integrate into society faster than rural migrants in leisure activities. This seems to be contrary to our common perception, because rural migrants still live in the countryside, and their entertainment and leisure modes have hardly changed [66], so rural migrants should be integrated into society more easily in their leisure activities. After the interview, we found that the leisure activities of rural migrants are mainly chatting with friends, watching TV and playing Mahjong. Although this does provide social contact and relaxation, the scope of social contact is relatively small and closed off, without introducing many new forms of information acquisition. However, urban migrants have abundant leisure activities, especially young people. They choose leisure activities such as jogging, reading, shopping and watching movies, which greatly increase their information content and subconsciously change their traditional thinking habits and living habits to more quickly integrate into their new location. Secondly, middle-aged and elderly people in urban communities will organize square dancing in the community, which enables them to have greater information exchange, emotional support and mutual assistance in life. The most important thing is that urban migrants may choose to communicate with their stakeholders in leisure activities. On the one hand, they can relax, and on the other hand, they can build their own social network, understand the industry information and relevant policies, so as to facilitate their life and work in the future. It can also be seen from Figure 4 that policy factors have the greatest impact on social integration. Therefore, urban migrants integrate into society faster than rural migrants in these areas.

4.4. Influence of Political Identity on Integration Speed in Maintenance Time and Working Time

The social integration rate of migrants who joined the party and those who did not join the party changed basically the same with the maintenance time, though with CPC migrants integrating slightly faster than those who are nonparty migrants. Although CPC members, as a kind of social resources and political resources, will likely receive some preferential treatment in their daily personal affairs, their social network resources that have been built up in their previous location would require time to rebuild to have the same effect as before. Therefore, the social integration rate of the CPC migrants and nonparty migrants do not differ by much [4]. Despite this, the potential benefits of CPC members will more or less make them integrate faster than nonparty migrants [26].

Migrants who join the party integrate into society faster in work activities than those who do not. First, the CPC migrants are more likely to find government jobs and are more likely to be protected by the current social system. In addition, China has strict requirements on the selection of CPC members, and the comprehensive ability of CPC members is generally strong, so the CPC migrants are more likely to be favored in the workplace. Finally, CPC migrants may have been village cadres before, with certain leadership and organizational skills [67], so they have a wider range of employment opportunities.

4.5. Influence of Educational Background on Integration Speed in Working Time

The degree of education determines the competitiveness of migrants in the job market. Migrants with a higher level of education have certain advantages in learning and accepting new things. They can engage in related work more readily after short-term employment training [68]. In addition, migrants with higher education level have stronger subjective initiative, and they can actively cooperate with the relevant policies of the government, so that they have more opportunities to improve themselves and adapt to the new working environment, thus accelerating the pace of social integration [69].

4.6. Policy Implication

Based on the conclusions drawn above, this paper proposes policy implications for the social integration of the migrants from the “Yangtze River to Huaihe River” project to help them integrate into the new life of the resettlement area. The policy implications are explained from three dimensions: work activity, maintenance activity and leisure activity.

4.6.1. Work Activities

Migrants aged 31-45 are the fastest to integrate into society; followed by those aged 16-30; migrants aged 46-60 were the slowest to integrate. the social integration rate of migrants of different ages varies significantly with the change of working time, and this is related to their physical condition and energy. In order to address this, it would be best to tailor market-oriented skills training provided by local government to the characteristics of different age brackets. As an example, for some young migrants, technical physical labor skills such as driving, hydraulic work, carpentry, cooking, electrician, plumber or masonry may be more suitable. Migrants over the age of 45 on the other hand, can learn the skills of the service industry such as housekeeping and sanitation work.

Female migrants are slower to integrate into society than men, and their basic legal rights are often not guaranteed. Therefore, in order to address this and promote even integration rates, the local government should further formulate specific and operational protection policies, such as: implementing the wage market positioning and wage growth standard of female migrants under the policy guidance; strictly preventing and combating the abuse of the probation period, and excessive extension of working hours and wage deduction.

Furthermore, the more educated migrants are, the faster they integrate into society at work. Therefore, different levels and types of education and employment training for different migrants would help in differentiating the needs to suit each type of migrant. First, the government should increase educational input in resettlement areas to improve the overall educational level of migrants. Second, for those migrants who are young and have a high school education level or above, medium and long-term adult education can be suited because they are expected to work longer and fulfill the conditions for further study. Third, for those older and less educated migrants, short-term training with strong applicability, practical features and low technical difficulty would be more suited to this demographic of migrants.

4.6.2. Maintenance Activities

Migrants aged 46–60 are the slowest to integrate into society in maintenance activities, the main reason being narrow interpersonal communication and single personal affairs in maintenance activities. At the same time, thought fixation hinders the change of their lifestyle and habits [56]. Therefore, various forms of community lectures on various targeted topics to address these obstacles to integration may prove effective. Topics such as health maintenance, public life, local culture and customs may prove to help migrants improve their quality of life, enhance their awareness of public life, and learn new codes of conduct and cultural customs. Through lectures, opportunities for public communication can also be created to transform the social integration of migrants from a negative assimilation process to a positive one. Secondly, it is suggested that local governments organize regular return visits to help solve the difficulties in the life of migrants, balance the cultural shock caused by forced relocation, and meet the cultural and psychological needs of migrants.

4.6.3. Leisure Activities

Rural migrants are slower to integrate into society than urban migrants in leisure activities, mainly due to the relatively closed information. In order to compensate, sources of information in a leisure context can be supplied to areas of rural migrant resettlement. This could take the form of radio stations, libraries and information exchange centers in order to learn more about the obligations, conduct codes and customs of resettlement areas.

5. Conclusions

This paper proposes a method to identify the state of social integration, and uses the Cox proportional hazards model to analyze the social integration characteristics of migrants from the “Yangtze River to Huaihe River” project in different time dimensions. Among them, age, gender, location, political identity and educational background are used as covariates, and working time, leisure time and maintenance time are used as independent variables. According to our research, we have the following conclusions:

(1)

Political factors, cultural factors and life factors have a greater impact on migrants’ social integration. Economic factors and security factors however, have relatively little impact on migrants’ social integration.

(2)

Migrants over the age of 45 have a serious lag in integrating into society in all aspects of activities.

(3)

Female migrants integrate into society faster than male migrants in maintenance activities and slower in work activities.

(4)

Rural migrants integrate into society faster than urban migrants in work activities and slower in leisure activities.

(5)

CPC migrants integrate into society faster than nonparty members in work activities and maintenance activities.

(6)

The higher the level of education, the faster migrants can integrate into society at work.

Most of the conclusions in the article are consistent with the existing research. It is worth noting that the source of information and the degree of understanding of the policy have a greater impact on the social integration of migrants from the “Yangtze River to Huaihe River” project. The wider the source of information and the more understanding of the policy, the more it helps migrants to integrate into society. This is also an important feature of migrants in the new era. In the past, project-induced migrants were not very concerned about policies and information, and always relied on government support and passively integrated into the new life of the resettlement area. With the arrival of the information age in China, people begin to realize that grasping government policies and understanding industry information are the key to wealth and success, and those migrants who are passively integrated always linger on the edge of poverty. Therefore, the grasp of policies and information is crucial for the social integration of migrants in the new era.

Another noteworthy phenomenon is that the social networking relationship established by the female group in the maintenance activities will greatly promote the social integration of female migrants. Women generally invest more in maintenance activities than men. The phenomenon of “clustering” in maintenance activities provides more communication opportunities for female migrants to become more united and more willing to help each other in unfamiliar environments. This has greatly promoted the social integration of female migrants.