Influence of perceived value of rural labour in China on the labourers’ willingness to return to their hometown: the moderating effect of social support

ABSTRACT

China is vigorously implementing rural revitalization strategy, emphasizing the significance of human resources in the sustainable development of rural areas and agriculture. As China’s economy is entering a new normal stage, the growth rate of rural labourers’ income in cities is slowing down, and the gap between urban and rural net incomes is gradually closing. The labourers who have migrated are gradually showing the willingness to return to their hometowns for development. This study constructs a theoretical framework for the studying of rural labourers’ willingness to return to their hometowns. It finds that the perceived value of rural labour positively influences their willingness to return. Further perceived extrinsic–intrinsic benefits, perceived extrinsic costs, and their perceived values are all significantly and positively correlated. Moreover, formal support has a positive moderating effect between the perceived value of rural labour and willingness to return. The perceived value of rural labour increases with higher perceived benefits and lower perceived costs; the stronger the perceived benefits and perceived value of labour, the greater the willingness to return.

KEYWORDS:

• Rural labourers
• perceived value
• willingness to return
• social support

1. Introduction

In recent years, the focus on agriculture in China has shifted to the comprehensive promotion of rural revitalization. Consequently, the central government has prioritized solving issues related to agriculture, rural areas, and farmers. The key to rural revitalization lies in people. The 2018 Central Document No. 1 issued by the General Office of the CPC Central Committee and the State Council has proposed that implementing a rural revitalization strategy must break the talent bottleneck, create more local talent, and gather talent from across the country. The Opinions on Accelerating the Revitalization of Rural Talents in 2021 pointed out that local talents must be vigorously cultivated, all kinds of talents must be attracted to promote rural revitalization, and the villagers’ endogenous motivation must be stimulated and cultivated. From a practical perspective, the people most likely to be attracted to rural areas are mainly those born and raised locally, namely rural migrant. Rural migrants commonly consist of unskilled migrants and rural talents. Rural talents encompass those with high educational attainment or rich working experiences. The rural talents mentioned here include young college graduates who have received higher education, professionals with rich work experience (such as teachers, doctors, government officials, entrepreneurs, professional managers, etc.), experienced workers who have not received higher education, as well as retired soldiers, and so on. After all, the number of rural talents is limited, and rural revitalization requires a large number of ordinary workers without high education and skills. Rural immigrants include not only rural labour between the ages of 16 and 60, but also the elderly and children. Rural labourers at this age group contribute a significant amount of income to their families and have significant power in deciding whether to return to their hometowns. Therefore, both the survey object and research object of this article specifically refer to rural labour. The total number of labour workers in China was 171.72 million in 2021. Among them, 100.42 million cross-provincial labour workers went up by 1.4%, accounting for 58.5% of the total number of workers outside the country. Notably, as the total number of migrant workers increases, the labour export of migrant workers shows an increasing trend, but the mobility radius is further reduced, and the tendency of employment in the vicinity or returning to the hometown increases (Chen & Wang, 2019).

Although the export of rural labour has decreased, and the willingness to migrate has changed, the loss of rural talent is still relatively serious. Many labourers leave the countryside for school or work, and the deterioration of the countryside’s demographic structure leads to the lack of endogenous power (Yang et al., 2016). With the intensification of urbanization, ‘urban diseases’ (or Problems of big cities) are becoming increasingly serious, such as traffic congestion, housing shortage, insufficient water supply, energy shortage, environmental pollution, and so on. Therefore, cities’ living costs are soaring, and the quality of life and happiness of rural labourers living in cities are decreasing. At the same time, with the gradual implementation of a rural revitalization strategy, rural conditions have steadily improved. Many rural labourers have again compared the psychological expectations of urban and rural life, believing that the attractiveness of cities has gradually declined; thus, some have chosen to return (Tang & Feng, 2015). Therefore, while China’s rural labour force continues to migrate out of the country, it also continues to migrate back in a process influenced by a combination of factors. Some papers have studied the labour force return problem in terms of influx and outflow factors, as well as social factors, but less research has been done at the level of individual psychology and personal perception (Bilecen, 2022; Mohamed & Abdul-Talib, 2020). Individuals’ behaviour is planned, and decisions are made after careful consideration and comparison. Individual perception factors can have a direct impact on their willingness to act (Baker et al., 2002; Kuo et al., 2009). The perceived value of rural labour return refers to the comparison between the perceived benefits and costs of labour when making return decisions. When making this decision, rural labour not only considers the corresponding benefits and costs. They will also examine the formal support provided by rural government departments (support policies, guidance, training, etc.) and the informal support (emotional support, material support, action support, and information support, etc.) provided through informal channels such as relatives, friends, neighbours, and community members (Li et al., 2020). According to the literature we have collected, there are few articles exploring the impact of perceived value on rural labour return willingness in the Chinese context, and examining the moderating effects of formal and informal support. As a result, itis necessary to explore the influence of labourers’ perceptions on labourers’ return from the perspective of individual perceived cost–benefit trade–off comparison, and examine the moderating effect of social support factors. This study will expand the application field of perceived value theory, deepen the research on the issue of rural labour return, and provide a theoretical basis for attracting labour return to rural areas.

2. Literature review and research hypothesis

2.1. Theory of labour mobility

Labour migration and mobility have always been much debated issues in economics research. In the 1950s, Lewis, an American development economist, proposed the first Dual Sector model (Lewis, 1954). This was refined by Todaro, an American development economist, who proposed a two-sector model of urban and rural population migration. The Todaro model states that if one job is added in the industrial sector at the minimum wage, more than one surplus rural labour may move to the city. Further, due to the movement of the surplus rural labour force, the income of the traditional agricultural sector will increase, which, in turn, will likely trigger a reverse movement of labour back to the countryside (Harris & Todaro, 1970). In the late 1950s, Bogue systematically elaborated on the push–pull theory, arguing that labour mobility is influenced by both influx and outflow, with favourable living conditions in the influx forming the pull force and unfavourable living conditions in the outflow forming the push force and that the interaction of the push and pull forces determine the direction of surplus labour mobility (Junge et al., 2015). When the rural push is greater than the pull, the labour force is more inclined to move to the city; when the urban push is greater than the pull, the labour force is more inclined to return to the rural area. In the 1980s, the new economic theory of migration, represented by Stark, Taylor, and Bloom, proposed that individuals and families are interdependent, individual urban and rural mobility decisions are made jointly by family members, and families are the main decision-makers of labour mobility (Stark & Bloom, 1985). From the perspective of the new migration economy theory, the surplus labour force in rural areas goes to work in the city for a period, and when the personal and family expectations are satisfied, the labour force will return home (Jean-Pierre, 2013). With the development of the social network theory, domestic and foreign scholars began to study the relationship between social networks and labour mobility, which emphasized the influence of social networks based on kinship, blood, and geography. Portes (1998) argued that social networks connect the inflow and outflow of labour by allowing friends and relatives to share information and the surplus labour force to choose whether to move to the city. This network system helps the labour force enter the influx and adapt quickly to the environment, in line with our countryside’s human and social context.

2.2. Return migration theory

‘Returning immigrant’, also known as ‘re-migration’ or ‘circular migration’, is usually defined as a type of immigration where immigrants return to their home country or region from the place where they previously aspired to settle as residents (Bilecen, 2022). The literature on migrant return can be divided into two areas: international migrant return and domestic migrant return. Due to the significant influence of politics, law, religion, culture, values, etc. between different countries, international migration differs greatly from domestic migration. Compared with the domestic immigrants that this article focuses on, there are significant differences. Therefore, we will not provide a detailed review of the literature on international migration here. Interested readers can read the review article Bilecen (2022), Mohamed and Abdul-Talib (2020).

The literature on domestic migrant return has different research perspectives (Li et al., 2020; Othniel & Gopal, 2008). Most of the literature focused on motivational factors, such as economic, psychological and emotional factors. Piotrowski and Tong (2010) examined economic and non-economic determinants of return migration young migrants from rural Thailand, and found that the effects of non-economic family-related factors were as strong in magnitude as economic effects in determining return. Hirvonen and Lilleor (2015) used 19-year tracking data from rural Tanzania to estimate the patterns and determinants of return migration, and find that return is largely associated with unsuccessful migration. Rerat (2014) addressed the migration behaviours of young university graduates from a rural region in Switzerland, and compared graduates’ current place of residence with characteristics related to their socio-familial, migration and professional trajectories. Wang et al. (2023) empirically tested the impact of rural e-commerce development on the return intentions of migrants, and found that rural e-commerce development promotes the return intentions of migrants. Tang et al. (2020) studied the factors that are most valued by returning labour when evaluating return decisions, and the order is external environment, individuals and families, and social network. Zhou and Tang (2022) examined the relationship between local-hukou residents’ attitudes towards migrants and the intention for return migration among rural-to-urban migrants while accounting for socio-economic heterogeneity in migrants.

Several papers have explored the impact of immigrant return on their career choices, investments, etc. Wu et al. (2023) evaluated how migration experience influences rural families’ choices for productive investments and the underlying mechanism of village poverty alleviation. Zhang et al. (2021) investigated the effect of outgoing working experience on return migrated labour's non-agricultural occupational change. Similarly, Wang and Yang (2013) studied the impact of return migration on the occupational choice in rural China.

However, there are few articles exploring the impact of perceived value on rural labour return intention, especially in Chinese context, and examining the moderating effects of formal and informal support.

2.3. Perceived value theory

Perceived value theory was formally proposed by Zeithaml in 1988. The economic view is that perceived value is closely related to price, and the psychological view is that perceived value is a matter of individual cognition and emotion (De Medeiros et al., 2016). Zeithaml considered perceived value as the consumer’s overall assessment of the utility of a product based on the perception of what was received and given; therefore, the perceived value represents a trade-off between perceived gain and give (Zeithaml, 1988). Wood and Scheer suggested that consumers’ perceived quality of products and services is derived from the product’s functional characteristics and all tangible and intangible aspects that consumers care about, including psychological risk, opportunity, and emotional costs (Wood & Scheer, 1996). Hansen et al. argued that perceived value is measured by the many benefits gained and the perceived benefits lost; therefore, perceived value equals the benefits minus the cost (Hansen et al., 2008). The hierarchical and trade-off models best typify the perceived value. As the research becomes more in-depth, the division of perceived value becomes specific, with the hierarchical and the trade-off models being the most representative. The hierarchical model assumes that perceived value is determined by the outcome of an individual’s processing of perceived information, which operates on three cognitive levels: (i) the individual’s initial perception of the product’s specific attributes and effectiveness; (ii) the individual’s overall evaluation of the product; (iii) the final perceived value formed through the comparative analysis between the initial perception and the overall evaluation (Woodruff, 1997). The trade–off model considers perceived value as an individual’s subjective perception formed by weighing perceived benefits and losses (Parasuraman & Grewal, 2000). Perceived benefits are the sum of the benefits individuals feel the product will deliver, including monetary and non–monetary gains. Perceived losses are the sum of the potential costs individuals must pay to obtain the product, including tangible and intangible costs (Huber et al., 1997; Lee & Jun, 2007; Sheth et al., 1991). The greater the individual’s perceived net benefit, the higher the level of individually perceived value and the more pronounced the effect on behavioural intentions. Based on these, perceived value is a subjective comparison of individual benefits and costs, including tangible and intangible gains and losses. In the context of the traditional rural cultural background, the perceived benefits and costs dimensions are divided from the intrinsic and extrinsic costs–benefits perceived by individuals.

Thus, individual behavioural decisions compare their expected cost–benefit trade-offs, essentially influenced by cognitive-level factors. Cognitive differences in individuals determine their perceived value, resulting in differences in willingness to take behavioural decisions. As the application area of perceived value expands, research has begun to focus on the impact of perceived value on labour force return. Liao et al. studied farmers’ regret of moving away from the countryside in terms of demand level, internal factors of perceived value and social influence, and external factors of cost, which significantly influenced farmers’ regret of leaving the countryside (Liao et al., 2018). Ghazali et al. studied individuals’ willingness to purchase and move based on the push–pull theory in terms of the cost-push of the outflow location and the attraction pull of the perceived value of the inflow location. The stronger the push of the outflow location and the attraction of the inflow location, the stronger the willingness to purchase a home and the individual’s willingness to move (Ghazali et al., 2020). Dalen and Henkens explored the main determinants of labour migration intentions in the Netherlands by delineating three dimensions–economic, social, and psychological factors–from an economic and sociological perspective. They found that perceived labour market opportunities, shared household expectations, and employee personalities might stimulate international labour mobility (Dalen & Henkens, 2012). Wang and Fan find that rural labourers’ return decisions reflect their perceived institutional and social disadvantages in the city, driving them to maintain strong ties to the countryside and that as labourers’ perceived costs and risks in the city increase, their perceived value in the city decreases, forcing them to return (Wang & Fan, 2006). Wang found that perceived benefits outside positively affect willingness to go out, and perceived costs outside negatively affect willingness to go out. If migrant workers perceive lower benefits and higher costs outside, they will be more willing to stay in their hometown (Wang, 2017). However, there are more job opportunities, education, medical, and other social security resources in cities than in rural areas, and these conditions are strongly attractive (Simões et al., 2021). Considering that returning to the countryside costs a lot of time and energy, while employment opportunities and employment channels are fewer, returning to the countryside means giving up existing contacts and economic sources, which will break the existing comfortable life and increase the risk, so the perceived cost may hinder the return of labour. When rural labourers return to their hometowns, if the perceived cost paid is greater than the benefits they receive, the lower the perceived value of labourers, the less likely they will return. Conversely, if the perceived benefits are higher than the perceived costs, the higher the perceived value of the labour force and the more likely it is to return. The perceived value of rural labour has an impact on the willingness to return, so the hypothesis is proposed:

H1: The perceived benefits of rural labour have a significant positive effect on perceived value;

H1a: The perceived extrinsic benefits of rural labour have a significant positive effect on perceived value;

H1b: The perceived intrinsic benefits of rural labour have a significant positive effect on perceived value;

H2: The perceived cost of rural labour has a significant negative effect on perceived value;

H2a: The perceived extrinsic cost of rural labour has a significant negative effect on perceived value;

H2b: The perceived intrinsic cost of rural labour has a significant negative effect on perceived value;

H3: The perceived benefits of rural labour have a significant positive effect on their willingness to return;

H4: The perceived cost of rural labour has a significant negative effect on its willingness to return;

H5: The perceived value of rural labourers’ return has a significant positive influence on their willingness to return.

2.4. Social support theory

Research on social support was first concentrated in the fields of psychology and medicine. The term was formally introduced in the 1970s. Cobb believed that social support is an explicit message based on which individuals can feel care, respect, and belonging to a group (Sidney, 1976). Thoits pointed out that social support is access to financial, instrumental, and informational help through relatives, friends, organizations, and so on (Thoits, 1985). Further, Chen and Choi argued that social support provides tangible resources, information, emotional needs, and companionship (Chen & Choi, 2011). In terms of support subjects, some scholars have divided social support into two dimensions, formal and informal support (Lipman & Longino, 1982), with the former mainly referring to government departments and the latter to non–governmental organizations and personal relationships. Martina and Melissa argued that local governments play the role of the lead agency and facilitator, active service provider, and passive delivery supporter, and government support has an impact on the renewal of local communities and the long–term settlement of migrants and refugees (Boese & Phillips, 2017). Scholars have found that formal support includes policy support, financial support, fiscal, and technical support, whereas informal support does not involve direct government intervention but is provided by important people with whom the recipient has a close relationship (Li & Kwaku, 2001; Lipman & Longino, 1982). Combining our national situation and rural development, the study defines social support as the resources and services rural labourers receive when deciding to return to their hometowns. It divides social support into two dimensions: formal support and informal support.

Social support is an important influence that individuals obtain through various social networks. It can help individuals reduce stress and improve their ability to adapt to society (Turner, 1992). Formal and informal support are two intertwined norms of support in social development. Formal support provides a solid guarantee for promoting rural development and resolving livelihood issues. Li et al. used activity theory analysis to study the role of governments in the rural e-commerce ecosystem and its impact on poverty alleviation. The study showed that the government plays a leading role in promoting rural e–commerce development, and the poor can also benefit from it (Li et al., 2019). As the gap between rural and urban areas becomes smaller, the workforce chooses to return to their hometowns to work, live, and start businesses, and the government provides them with support in every aspect. However, policy support in cities may also discourage labourers from returning to their hometowns. Yue et al. studied the willingness of China’s urban and rural migrant population from an intergenerational perspective and found that the educational environment, quality of education and educational policy support in cities were extremely attractive factors for young people from rural areas, which, to some extent, discouraged labourers from returning to their hometowns (Yue et al., 2010). Informal support can provide individuals with solid livelihood security and emotional attachment. Lu (2012) found that the emotional and family life of labourers working away from their hometown for a long time is affected, increasing the emotional burden of family members. They may decide to return to their hometown to alleviate the emotional stress and gain family support. Ryan et al. argued that the connection of individuals’ social networks could influence the decision to migrate, relatives and partners provide emotional support when individuals are lonely, and acquaintances’ relationships can provide convenience in work and life. These forms of support reduce individuals’ anxiety during migration (Ryan et al., 2008). Rural labourers who have been away from their hometowns for a long time are deeply concerned about their parents and children. Once suitable employment and business opportunities in their hometowns exist, they return without hesitation. Therefore, the hypothesis is proposed:

H6a: Formal support plays a positive moderating role between rural labourers’ perceived value and their willingness to return;

H6b: Informal support plays a positive moderating role between rural labour's perceived value and its willingness to return.

In the context of the above analysis, this study constructs a relevant model based on the trade-off model of perceived value and combines the research hypothesis to construct a relational model of the influence of the perceived value of rural labour on its decision of willingness to return, as shown in Figure 1.

Figure 1. Model of the influence of perceived value of rural labour on its decision of willingness to return.

3. Empirical analysis

3.1. Descriptive statistics

The study drew on the scale of Sweeney and Soutar (Sweeney & Soutar, 2001) and other scholars, combined with the current situation of talent revitalization for rural revitalization in China, and designed a questionnaire in five parts. The data were obtained through questionnaires, and preliminary descriptive statistics was conducted based on personal characteristics, family situation, work experience, and economic characteristics noted in the questionnaire. The selected labour export destinations for this survey are several villages under the jurisdiction of Hengshui City, a city in Hebei Province, China. This survey focuses on people born in these villages and work in city, aged 16–60 years old, and have at least six months of experience working outside. This survey distributed the questionnaire in a snowball format. The questionnaire investigated personal characteristics, family conditions, migrant work experience, economic characteristics and other aspects. We choose to distribute the questionnaire from the end of January to early February 2022, which coincides with the Chinese New Year, when a large number of migrant workers return home for the holiday. A total of 450 questionnaires were distributed in the survey, and 322 valid questionnaires were collected, with an effective rate of 71.56%. The basic information of the survey sample is given in Table 1.

Table 1. Distribution of individual characteristics of the rural labour force.

Statistics	Categories	Sample size	Effective percentage
Gender	Male	194	60.2
	Female	128	39.8
	Total	322	100.0
Age	Under 16 years old	0	0
	16–30 years old	138	42.9
	31–40 years old	101	31.4
	41–50 years old	75	23.3
	51–60 years old	8	2.5
	61 years old and above	0	0
	Total	322	100.0
Education	Junior high school and below	10	3.1
	High school and junior college	39	12.1
	College and undergraduate	160	49.7
	Postgraduate and above	113	35.1
	Total	322	100.0
Marital Status	Married	194	60.2
	Unmarried	128	39.8
	Total	322	100.0
Family members	Under 3	61	18.9
	3–6	252	78.3
	More than 6	9	2.8
	Total	322	100.0
Number of outworkings	Under 3	190	59.0
	3–5	98	30.4
	More than 5	34	10.6
	Total	322	100.0
Occupation	Students	11	3.4
	Farmers	39	12.1
	Workers	90	28.0
	Enterprise employees	58	18.0
	Freelancers	31	9.6
	Individuals	32	9.9
	Teachers	9	2.8
	Civil servant	52	16.1
	Others	11	3.4
	Total	322	100.0
Annual income	Under 3w	96	29.8
	3w-6w	66	20.5
	6w-9w	58	18.0
	9w-12w	47	14.6
	More than 12w	55	17.1
	Total	322	100.0

3.2. CFA model test

Based on the model test results in Table 2, CMIN/DF = 2.141, which is in the excellent range of 1-3, and RMSEA = 0.060, which is well within the range of 0.08. In addition, IFI, CFI and TLI tests all reached a good level of 0.8 or more. Therefore, the combined study analysis results indicate that the scale CFA model has a good fit.

Table 2. CFA Model Test.

	Standard	Result
CMIN/DF	1–3 excellent, 3-5 good	2.141
RMSEA	<0.05 excellent, <0.08 good	0.060
IFI	>0.9 excellent, >0.8 good	0.891
TLI	>0.9 excellent, >0.8 good	0.872
CFI	>0.9 excellent, >0.8 good	0.889

3.3. Reliability and validity tests

Reliability analysis tests the reliability, stability, and consistency of questionnaires in a study. Reliability analysis is performed through Cronbach’s α. Scholars usually consider the reliability of the questionnaire to be higher when Cronbach’s α is greater than 0.8. Table 3 shows the results of the reliability test. Except for the Cronbach’s α of 0.676 for the informal support variables, the Cronbach’s α of all other questionnaire dimensions and the overall reliability test results are greater than 0.7. Hence, the reliability of the questionnaire is good.

Table 3. Reliability test.

Component Dimension	Item number	Cronbach’s α within dimensions	Item number
PEB	PEB1	0.840	5
	PEB2
	PEB3
	PEB4
	PEB5
PIB	PIB1	0.718	5
	PIB2
	PIB3
	PIB4
	PIB5
Total PB		0.856	10
PEC	PEC1	0.733	4
	PEC2
	PEC3
	PEC4
PIC	PIC1	0.829	3
	PIC2
	PIC3
Total PC		0.866	7
PV	PV1	0.745	3
	PV2
	PV3
FS	FS1	0.727	3
	FS2
	FS3
IS	IS1	0.676	4
	IS2
	IS3
	IS4
Total SS		0.740	7
Will	Will1	0.735	3
	Will2
	Will3

Validity analysis measures the reasonableness of the questionnaire and verifies the validity of the questionnaire. The questionnaire dimensions explored in the study are designed by referring to the literature of existing academic scholars and combining them with the current situation of rural revitalization and talent revitalization in China. Hence, the content validity is good. First, SPSS 22.0 data analysis software is used to validate the structural validity, and the structural validity of the questionnaire scales is judged by the KMO and Bartlett’s tests. The structural validity of the questionnaire scale is good when the KMO is greater than 0.8 and P < 0.05. As seen in Table 4, the KMO is 0.917, which is greater than 0.8, and the Bartlett test is 4181.775 with a df of 435, and p < 0.001. Hence, the construct validity is good.

Table 4. KMO test and Bartlett’s test results.

KMO		0.917
Bartlett’s test	χ^²	4181.775
	Df	435
	P	0.000***

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

Then, a validation factor analysis is conducted using AMOS 22.0 to verify the convergent and discriminant validity of the questionnaire, where convergent validity requires standardized factor loadings, composite reliability (CR), and average variance extracted (AVE) are greater than 0.7 and 0.5, respectively. AVE between 0.36-0.5 is an acceptable threshold for convergent validity (Fornell & David, 1981; Huang et al., 2013; Yin, 2018). In summary, as seen in Table 5, each dimension has acceptable convergent validity and well-combined reliability Table 5.

Table 5. Convergent validity test.

			Estimate	AVE	CR
PEB1	<—	PEB	0.774	0.5213	0.8427
PEB2	<—	PEB,	0.683
PEB3	<—	PEB	0.553
PEB4	<—	PEB	0.760
PEB5	<—	PEB	0.811
PIB1	<—	PIB	0.561	0.3528	0.7277
PIB2	<—	PIB	0.685
PIB3	<—	PIB	0.580
PIB4	<—	PIB	0.663
PIB5	<—	PIB	0.452
PEC1	<—	PEC	0.616	0.4064	0.7286
PEC2	<—	PEC	0.778
PEC3	<—	PEC	0.541
PEC4	<—	PEC	0.590
PIC1	<—	PIC	0.809	0.6263	0.8335
PIC2	<—	PIC	0.842
PIC3	<—	PIC	0.718
FS1	<—	FS	0.643	0.4875	0.7372
FS2	<—	FS	0.615
FS3	<—	FS	0.819
IS1	<—	IS	0.498	0.3506	0.6809
IS2	<—	IS	0.641
IS3	<—	IS	0.555
IS4	<—	IS	0.660
PV1	<—	PV	0.713	0.5088	0.7541
PV2	<—	PV	0.607
PV3	<—	PV	0.806
Will1	<—	Will	0.654	0.4787	0.7334
Will2	<—	Will	0.703
Will3	<—	Will	0.717

Drawing on existing studies in the literature, discriminant validity tests were conducted by multi-factor analysis (Shijian et al., 2020; Yang et al., 2015; Zhang et al., 2019), and in the discriminant validity tests, all other multi-factor models passed the significance test at a significance level of 0.001, but indicators such as CMIN/DF and RMSEA became worse compared to the original model, therefore, indicating that the original model had good discriminant validity see Table 6.

Table 6. Discriminant validity test.

	factors	CMIN	DF	P	CMIN/DF	RMSEA	IFI	TLI	CFI	Comparison	△CMIN	△DF
1	8	807.187	377	.000***	2.141	0.060	0.891	0.872	0.889
2	7	914.827	384	.000***	2.382	0.066	0.865	0.846	0.864	2vs1	107.64	7
3	6	1061.327	390	.000***	2.721	0.073	0.830	0.808	0.828	3vs1	254.14	13
4	5	1128.246	395	.000***	2.856	0.076	0.814	0.793	0.812	4vs1	321.059	18
5	4	1198.069	399	.000***	3.003	0.079	0.797	0.776	0.795	5vs1	390.882	22
6	3	1216.434	402	.000***	3.026	0.079	0.793	0.774	0.791	6vs1	409.247	25
7	2	1502.999	404	.000***	3.720	0.092	0.720	0.696	0.718	7vs1	695.812	27
8	1	1669.773	405	.000***	4.123	0.099	0.678	0.651	0.675	8vs1	862.586	28

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

3.4. Hypothesis test

Perceived value is a cognitive evaluation obtained by individuals after comparing perceived benefits and perceived costs, so the perceived value variable of rural labour was recalculated and coded as PV. Based on the logical behavioural framework of perceived value, a linear regression model was constructed to examine the mechanism of the role of each dimension of perceived benefits and perceived costs on perceived value and rural labour's willingness to return.

To investigate the effects of each dimension of perceived benefits and perceived costs of rural labour on perceived value, multiple linear regression analysis was conducted on perceived extrinsic benefits, perceived intrinsic benefits, perceived extrinsic costs, and perceived intrinsic costs of rural labour, and the results of the analysis are shown in Tables 7 and 8. The closer R² is to 1, the better the model fit is. According to the parametric test, the adjusted R² = 0.916 and DW = 1.855, which are approximately equal to 2, indicate that the residual series are not autocorrelated, and the F-statistic is 873.670 with p = 0.000 < 0.001, indicating that it is significant at the test level of α = 0.001 and the fitted multiple linear regression equation is statistically significant.

Table 7. Model parameter test.

R	R^2	Adjusted R^2	Std. Error	Durbin-Watson	F	P
0.958	0.917	0.916	0.24653	1.855	873.670	0.000

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

Table 8. Regression model.

	Coeff.		Std. Coeff	T	P	Collinear Statistics
	B	Std. Error	Beta			Franchise	VIF
	0.517	0.154		3.352	.001
PEB	0.393	0.026	0.336	15.114	.000	0.532	1.879
PIB	0.484	0.031	0.323	15.467	.000	0.603	1.658
PEC	−0.308	0.027	−0.276	−11.208	.000	0.433	2.308
PIC	−0.220	0.023	−0.246	−9.495	.000	0.390	2.565

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

According to the criteria of the covariance index, the model is well constructed when VIF < 5 and there is no multicollinearity. The VIF values in this analysis are all less than 5, which indicates that the multiple linear regression model is well constructed and the equation is meaningful. Through regression analysis, the following analysis results were obtained in Table 8. The regression coefficient of PEB was 0.393, p = 0.000 < 0.001, significant at the 0.001 level, indicating that PEB has a significant positive effect on PV, and the higher the perceived extrinsic benefits of rural labour, the higher its perceived value, and hypothesis H1a was verified. The regression coefficient of PIB was 0.484, p = 0.000 < 0.001, indicating that PIB has a significant positive effect on PV. The higher the intrinsic benefits of rural labour, the higher its perceived value and hypothesis H1b is verified. The regression coefficients of PEC and PIC are both negative, with regression coefficients of −0.308 and −0.220, respectively.

It means hypothesis H2a and H2b are both verified. Therefore, the higher the perceived extrinsic or intrinsic cost of rural labour, the lower its perceived value, and the more negative individual perceptions.

In summary, the perceived extrinsic benefits and perceived intrinsic benefits of rural labour have a significant positive effect on their perceived value, with regression coefficients of 0.393 and 0.484, while perceived extrinsic costs and perceived intrinsic costs have a significant negative effect on perceived value, with regression coefficients of −0.308 and −0.220, respectively. The model equation of the multiple linear regression is: Return of rural labour PV = 0.517** + 0.393*** PEB + 0.484*** PIB – 0.308*** PEC – 0.220*** PIC.

Meanwhile, to investigate the role of the overall perceived benefits and overall perceived costs of rural labour on their perceived values, PB and PC were put into a linear regression model with PB and PC as independent variables, and the analysis results are shown in Tables 9 and 10.

Table 9. Model parameter test.

R	R^2	Adjusted R^2	Std. Error	Durbin-Watson	F	P
0.957	0.916	0.915	0.24717	1.827	1736.486	0.000

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

Table 10. Regression model.

	Coeff.		Std. Coeff	T		P	Collinear Statistics
	B	Std. Error	Beta				Franchise	VIF
	0.577	0.151		3.811		.000
PB	0.864	0.029	0.590	29.599		.000	0.664	1.505
PC	−0.524	0.022	−0.486	−24.382		.000	0.664	1.505

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

According to the regression model analysis in Table 10, the VIF values of PB and PC are both less than 5, indicating that the regression model is well constructed. The regression coefficient of PB is 0.864, p = 0.000 < 0.001, which is significant at the 0.001 level. The perceived benefits of rural labour are significantly and positively correlated with their perceived values. The higher the benefits perceived by individuals, the higher their perceived value and the more positive perceptions they bring, so hypothesis H1 is verified. The regression coefficient of PC is −0.524, p = 0.000 < 0.001, which proves that the cost factor is significant at the 0.001 level. The higher the perceived cost of rural labourers, the more their perceived value will decrease. Individuals will generate more negative perceptions, and hypothesis H2 will be verified.

Through the logical framework of perceived value behaviour, the role of each dimension of perceived benefit and perceived cost of rural labour on perceived value was examined, followed by the analysis of the influence of each dimension of perceived value of rural labour on willingness to return, and stepwise regression was conducted to obtain the research results in Tables 11–13 below. Model 1 was obtained by PB and PC step by step, and then model 2 was obtained by PB and PC with PV step by step. Model parameter test results are shown in Table 11, the adjusted R2 of models 1 and 2 are 0.313 and 0.331, respectively, and the model fits well. DW = 2.082, F is 147.009, p = 0.000 < 0.001, indicating that it is significant at the test level of 0.001 and the equation is statistically significant. Based on the above analysis results, we can infer the following viewpoint. If rural governments hope that returning home can bring higher perceived value to rural labour, they must take measures from two aspects. One is to find ways to enable them to obtain more benefits, both external and internal. The second is to make every effort to reduce their return costs, both external and internal.

Table 11. Model parameter test.

Model	R	R^2	Adjusted R^2	Std. Error	F	P	Durbin-Watson
Model 1	0.561	0.315	0.313	0.61181	147.009	.000
Model 2	0.579	0.335	0.331	0.60374	80.289	.000	2.082

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

Table 12. Stepwise regression of variables for each dimension of perceived value.

Model		Coeff.		Std. Coeff	T	P	Collinear Statistics
		B	Std. Error	Beta			Franchise	Std. Error
Model 1		1.050	0.236		4.458	.000
	PB	0.714	0.059	0.561	12.125	.000	1.000	1.000
Model 2		1.624	0.297		5.465	0.000
	PB	0.394	0.118	0.310	3.329	0.001	0.241	4.151
	PV	0.251	0.081	0.288	3.101	0.002	0.241	4.151

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

Table 13. Excluded variables.

Model		Beta	T	P	Franchise	VIF
Model 1	PC	−0.076	−1.336	0.183	0.664	1.505
Model 2	PC	0.184	1.953	0.052	0.232	4.311

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

The results of the stepwise regression test are shown in Tables 12 and 13. Model 1 adds PB and PC, and the coefficient of PB is 0.714, p = 0.000 < 0.001, which indicates that the perceived benefits of rural labour have a significant positive influence on the willingness to return when the perceived benefits of labour in the countryside increase and perceive that the countryside can provide more opportunities and resources, their willingness to return home will also increase. The perceived cost of rural labour is p = 0.183 > 0.05, so the perceived cost of rural labour has no significant effect on the willingness to return, so the stepwise regression excludes the perceived cost of rural labour. The coefficient of PB is 0.394, p = 0.001 < 0.01, and the coefficient of PV is 0.251, p = 0.002 < 0.01, which is significant at 0.01 level, indicating that the perceived value of rural labour has a significant positive influence on its willingness to return. H3 and H5 are verified.

The higher the perceived value of the rural labour force, the higher the degree of the labour force's willingness to return to their hometown, probably because they see the benefits such as material and opportunities that their hometown can offer them and pay less cost; the perceived cost of rural labour force p = 0.052 > 0.05 has no significant effect on the willingness to return, so the perceived cost factor is excluded from the hierarchical regression. H4 is not verified. With regard to this result, it may be explained as follows. Hometown attachment or Native Land Complex is particularly prominent in Eastern culture (Wang et al., 2018), and Confucian culture teaches that people should stay in their hometown. As a result, China's migrant rural labours may be particularly sensitive to the benefit forces that draw them back and less so to the costs of returning home. This is consistent with their cultural values (Li et al., 2020). However, it does not undermine the importance of perceived cost factor of rural labour. It should be noted that compared to reducing the return cost of rural labour, rural governments should find ways to create greater benefits for them, such as employment opportunities, school conditions, medical conditions, transportation conditions, and so on.

To investigate the role of social support in the influence of perceived value of rural labour on their willingness to return, hierarchical regression was used. Firstly, the independent variables (PV) and moderating variables (FS, IS) were decentralized to obtain perceived value ZPV, formal support ZFS, and informal support ZIS. Next, the perceived value ZPV, moderating variables (ZFS, ZIS), and the interaction between perceived value and social support (ZPVZFS, ZPVZIS) were sequentially incorporated into the model to explore the moderating role of social support in the impact of perceived value of rural labour on return intention. The model parameters were tested as Table 14. Model 3 incorporates ZPV to obtain an adjusted R2 of 0.310 with a significant p-value of 0.000 < 0.001, model 4 incorporates the independent variables ZPV, ZFS, and ZIS into the model and obtains an adjusted R2 of 0.374 with a significant p-value less than 0.001, and model 5 incorporates ZPV, ZFS, ZIS and the interaction terms ZPVZFS, ZPVZIS are included in the model to obtain the final hierarchical regression model with an adjusted R2 of 0.378 and a significance p-value of 0.000 < 0.001. From model 3 to model 5, the adjusted R2 increases from 0.310–0.378, the model fits better, and the degree of explanation for the willingness to return to rural labour increases, indicating that two dimensions of social support play a role. The DW value is 2.090, within the standard range of approximately 2, the residual series are not autocorrelated, all are significant, and the hierarchical regression equation is statistically significant.

Table 14. Model parameter test.

Model	R	R^2	Adjusted R^2	Std. Error	F	P	Durbin-Watson
Model 3	0.558	0.312	0.310	0.61318	144.928	.000
Model 4	0.616	0.380	0.374	0.58394	64.886	.000
Model 5	0.623	0.388	0.378	0.58183	40.078	.000	2.090

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

The results of the hierarchical regression model are shown in Table 15. Model 3 puts ZPV into the model, and the regression coefficient is 0.485, p = 0.000 < 0.001. The perceived value of rural labour has a significant positive effect on its willingness to return to the countryside decision, indicating that as the perceived value of rural labour increases, the willingness of this part of rural labour to return to the countryside will also become stronger. Model 4 puts ZPV, ZFS and ZIS into the model, where p = 0.193 for ZFS on rural labour’s willingness to return and p = 0.000 < 0.001 for ZIS on rural labour’s willingness to return, indicating that formal support provided by rural area government has no significant effect on rural labour’s return, while emotional and financial support brought by family, relatives and community, etc. have a significant positive impact. After adding the two variables, ZPV is still significant at the 0.001 level, indicating that rural labourers’ subjective value still plays an influential role. The higher the perceived value of rural labourers, the higher the cognitive motivation, and the stronger the influence on their willingness to return to their hometowns, and the more emotional support and substantial help NGOs and individuals can provide, the stronger the willingness of rural labourers to return to their hometowns. Model 5 was analysed by adding the interaction terms ZPVZFS and ZPVZIS to model 4, and the results showed that ZPVZIS did not pass the significance test and hypothesis H6b was not tested. The p of ZPVZFS on the perceived value of rural labour is 0.041 < 0.05, which is significant at the 0.05 level, and the regression coefficient is 0.114, indicating that formal support plays a significant positive moderating role in the influence of the perceived value of rural labour on its willingness to return to the countryside. The stronger the perceived formal support, the more significant the moderating effect is. Hypothesis H6a is tested. The above results indicate that although formal support provided by rural governments does not have a significant direct impact on the willingness of rural labour to return, it can still have an indirect effect, and positively strengthen the impact of perceived value on the willingness to return. Unlike formal support, informal support such as emotional and economic support brought by family, friends, neighbours, and communities has a direct and significant positive impact on the willingness of rural labour to return. However, informal support did not have an indirect effect, and the impact of perceived value of rural labour on return intention did not change significantly due to the level of informal support.

Table 15. Regression model.

Model		Coeff.		Std. Coeff	T		P	Collinear Statistics
		B	Std. Error	Beta				允差	B
Model 3		3.876	0.034		113.423		.000
	ZPV	0.485	0.040	0.558	12.039		.000	1.000	1.000
Model 4		3.876	0.033		119.102		.000
	ZPV	0.316	0.048	0.363	6.567		.000	0.637	1.569
	ZFS	0.076	0.058	0.064	1.305		.193	0.819	1.221
	ZIS	0.387	0.072	0.296	5.342		.000	0.635	1.575
Model 5		3.866	0.036		106.430		.000
	ZPV	0.312	0.049	0.359	6.309		.000	0.597	1.676
	ZFS	0.140	0.066	0.118	2.134		.034	0.637	1.569
	ZIS	0.379	0.073	0.290	5.174		.000	0.617	1.621
	ZPVZFS	0.114	0.055	0.119	2.051		.041	0.571	1.751
	ZPVZIS	−0.047	0.067	−0.039	−0.697		.486	0.615	1.625

Note: ***p < 0.001, **p < 0.01, * p < 0.05.

In order to study the impact of individual characteristic variables such as gender, age and family on the return intention of rural labour force, the return intention of rural labour force was re-coded, and intentions 0–3 were coded as 0, representing ‘no’, and intentions 3–5 were coded as 1, representing ‘yes’. Chi-square analysis was conducted on the individual characteristic variables and return intention, and the specific analysis was as follows.

The basic characteristic variables such as gender, age and education background of the survey subjects were analyzed with Chi-square analysis respectively, and the cross-table results were obtained, as shown in Table 16. In terms of gender, among the 322 respondents, 175 men had the intention of returning, accounting for 90.21% of males, and 19 had no intention of returning, accounting for 9.79% of males. There were 98 women with the intention to return, accounting for 76.56% of females, and 30 women without the intention to return, accounting for 23.44% of females. The Chi-square value between the gender of the rural labour force surveyed and the intention to return is 11.127, and the p value is 0.001, which is significant at the level of 0.01. Therefore, the intention to return is significantly correlated with gender. The traditional idea is that women are responsible for taking care of the elderly and children, and their willingness to return will be higher than that of men. However, the survey shows that women's willingness to return is lower than that of men. This may be because with the changes of The Times, after receiving higher education, getting used to the pace of migrant work and the concept of cities, the traditional concept of women staying behind has been broken. Rural labourers have stable jobs and satisfactory salaries outside, and they will consider bringing their parents and children to live in cities together, so that they can still take good care of their families, and women's willingness to return will also be reduced.

Table 16. Cross-table analysis of return intention and individual characteristic variables.

			Willingness to return		Total
			Yes	No
Gendar	Men	count	175	19	194
		Percentage	90.21%	9.79%
	Women	count	98	30	128
		Percentage	76.56%	23.44%
	Total	count	273	49	322
		Percentage	84.78%	15.22%
	chi-square	11.127
	P	0.001*
Age	16–30	count	109	29	138
		Percentage	78.99%	21.01%
	31–40	count	92	9	101
		Percentage	91.09%	8.91%
	41–50	count	66	9	75
		Percentage	88.00%	12.00%
	51–60	count	6	2	8
		Percentage	75.00%	25.00%
	Total	count	273	49	322
		Percentage	84.78%	15.22%
	chi-square	7.903
	P	0.048*
Education	Junior high school	count	10	0	10
		Percentage	100.00%	0.00%
	High school	count	33	6	39
		Percentage	84.62%	15.38%
	undergraduate	count	138	22	160
		Percentage	86.25%	13.75%
	Postgraduate and above	count	92	21	113
		Percentage	81.42%	18.58%
	Total	count	273	49	322
		Percentage	84.78%	15.22%
	chi-square	3.055
	P	0.383

From the perspective of age, 109 rural workers aged 16–30 had the intention to return, accounting for 78.99%, and 29 had no intention to return, accounting for 21.01%; Among the 31–40 years old, 92 people had the intention to return, accounting for 91.09% of this age group, and the other 9 people had no intention to return, accounting for 8.91%; Among the rural labour force aged 41–50, 66 people (88.00%) had the intention to return, and 9 people (12.00%) had no intention to return. In the 51–60 age group, 6 people had the intention to return, accounting for 75.00%, and 2 people had no intention to return, accounting for 25.00%. The Chi-square value between the age of the respondents and the return intention was 7.903, and the p value was 0.048, which was significant at the level of 0.05. Therefore, the return intention of rural labour was significantly correlated with age. Cities are more attractive to young rural workers. Cities can not only meet the material and cultural life of young people, but also provide high career expectations. However, in the past two years, affected by the COVID-19 epidemic, urban living costs have increased, real income has decreased, and young workers’ willingness to return is also increasing.

From the academic background of the rural labour force, there are 10 respondents who have a junior high school education or below, and all of them are willing to return. Among the rural labour force with high school and technical secondary education, 33 people have the intention to return, accounting for 84.62%, and 6 people have no intention to return, accounting for 15.38%; Among those with college or bachelor's degree, 138 people have the intention to return, accounting for 86.25%, and 22 people have no intention to return, accounting for 13.75%. There are 92 people with graduate degree or above who have the intention to return, accounting for 81.42%, and 21 people who have no intention to return, accounting for 18.58%. The Chi-square value between rural labour's educational level and return intention is 3.055, and the p value is 0.383 > 0.05. Therefore, there is no significant correlation between rural labour's educational level and return intention.

4. Discussion and conclusion

4.1. Discussion

Promoting a rural revitalization strategy is the inevitable way to promote China's high-quality development and accelerate the modernization of agriculture and rural areas. To revitalize the countryside, the key lies in ‘local people’. We should provide reasonable guidance to the rural labour force and correctly understand the phenomenon of returning to the countryside. Only by organically combining the attraction of talents with rural revitalization, breaking the existing system, building a good rural environment and infrastructure, enhancing the attraction, and relying on the returning labour force for development can rural revitalization be promoted in a lasting way. Only with sufficient labour force talents can we promote rural development and attract more labour force to return.

Perceived benefits of rural labour have a significant positive impact on perceived value, while perceived costs have a significant negative impact. Perceived benefits have a positive impact on return intention, and perceived value also positively affects the decision of return intention of labour。These are consistent with previous studies(Wang, 2017). The perceived cost has no significant effect on the willingness to return, which is different from previous studies (Galipo, 2018; Ghazali et al., 2020; Wang, 2017). Due to China's traditional agricultural civilization, those born and raised in rural areas have a natural and strong local sentiment towards their hometown. Hometown has a special status, and people have a strong attachment and identity towards it. Traditional nostalgia concepts such as ‘homesickness’ and ‘falling leaves returning to their roots’ are also driving forces for the return of labour. This will make rural labours, especially those with a strong hometown complex, not completely rational in treating all factors equally when making return decisions, but more concerned with benefits rather than costs.

Formal support does not have a significant direct impact on the willingness of rural labour to return, but has a positive moderating effect on perceived value and return intention of rural labours. This finding is not consistent with existing papers. Luo & Chen, 2020 found that formal support (government policy perception) has a direct impact on willingness to return to hometown, but didn’t explore the moderating effect of formal support (government policy perception) in the relation between perceived value and return intention. The reason for the inconsistent research results may be that their interpretation of policy perception is whether the rural labours are aware of and satisfied with the relevant policies for rural revitalization. The formal support in this article not only includes policies, but also includes the public service currently provided by the government, such as education, medical care, transportation, and other conditions. In the past, rural governments failed to implement the policies initially proposed due to various reasons and did not achieve the initial promised results. This leads to a low level of trust in policies among rural residents. Therefore, even if rural labourers are aware of the various support policies provided by local governments, they may believe that the policies will take a long time to be effective, or may not necessarily achieve corresponding results. Therefore, formal support did not have a significant impact on the willingness to return.

Informal support has a significant impact on the willingness of rural labour to return, but has no significant moderating effect on the relation between perceived value and return intention of rural labours. It is counter-intuitive that informal support does not have a corresponding moderating effect. This may be because the majority of rural migrant workers are young adults, who have studied away from home, received higher education, and worked or settled directly in the city after finishing their studies, adapting to the pace of urban life and integrating into the development of the city. For them, they have not established a deep emotional connection with the countryside for a long time, and the relationship with the rural community is not close, and the rural life is strange and helpless. Even if they choose to return to their hometown for development, their relatives and friends at home may not be able to provide them with appropriate resource conditions and opportunities.

4.2. Policy recommendations

The results of the study indicated that the perceived value of rural labour had a significant effect, and formal support had a moderating effect on their decision to return to their hometown. Based on those, the paper proposes policy suggestions to enhance rural labour's perceived value and promote rural labour's return from the perspectives of material benefits, emotional belonging, and policy and service security system. The specific recommendations follow below.

4.2.1. Increase material benefit to enhance the return intention of rural labours

Perceived benefits will directly affect the judgment of rural labour on perceived value, thereby affecting their willingness to return. Enabling rural labour to obtain better material benefits is an important means of improving perceived benefits. When deciding whether to return, rural labour will first determine whether their hometown has good, at least not bad, job opportunities and income. Hence, rural governments should rely on local advantageous resources, attract investment, and vigorously develop characteristic industries. For example, Anji County, Zhejiang Province, as the birthplace of the theory that ‘green waters and green mountains are golden mountains and silver mountains’, relies on the local good ecological resources to actively attract rural labour to return and develop the rural tourism and homestay industries, achieving great success.

Secondly, rural labour will also examine whether their hometown has good public service conditions and whether appropriate economic subsidies can be provided. Certain economic subsidies should be provided to attract labourers to return to their hometowns. The rural government should establish an information registration platform for returning labourers and offer subsidies to those who return to their hometowns for re-employment to help tide them over the difficulties. Although many rural governments have limited financial resources, there are still many policies and funds available from channels such as central, provincial, and municipal governments. In recent years, China has taken advantage of the unique nature of its governance model and adopted a strategy of ‘the Rich First Pushing Those Being Rich Later’. It has widely implemented multiple counterpart assistance plans, such as the ‘East West Cooperation Assistance Plan’, the ‘Key Cities in Revolutionary Old Areas Cooperation Work Plan,’ the ‘Counterpart assistance plan for Xinjiang’, the ‘counterpart assistance plan for Tibet’, ‘Implementation Plan for Paired Cooperation between Shanghai, Jiangsu, and Zhejiang Cities to Assist Cities in Northern Anhui’ and so on. In addition, China is comprehensively promoting rural revitalization, formulating various policies, and providing a large amount of funds to support projects related to rural revitalization. Multiple national departments have issued subsidy application projects, focusing on planting, breeding, processing, and circulation in the agricultural and rural areas. Rural governments should actively utilize national support policies and apply for support projects to improve local medical, educational, transportational, ecological and other conditions. Moreove, rural governments should assist rural labours in applying for support funds in order to provide more perceived benefits for rural labours and attract them to return to their hometowns.

4.2.2. Enhancing emotional belonging and hometown identity to improve the return intention of rural labours

From thousands of years until recent decades, China has always been a rural society. People have a natural attachment to the countryside. Traditional villages are an important carrier of cultural inheritance, carrying people's long nostalgic memories. Rural governments should increase cultural excavation and protection efforts, explore traditional village protection and development models, and organically combine protection, inheritance, and development to retain the charm of the countryside. It should actively build rural libraries, museums, cultural squares, fitness facilities, harden roads, lay rainwater and sewage pipelines, build public toilets and amusement parks, promote garbage classification and management, and vigorously improve the hardware and software conditions of rural life. Through the above measures, the rural environment is made more pleasant, the rural scenery is more charming, and life is comfortable, thereby enhancing the sense of hometown identity of returning labours and enhancing the perceived benefits at the spiritual level. Only in this way can rural labour be better attracted back and retained.

Additionally, services that enhance emotional belonging cannot be ignored in attracting labourers to return home. Villages should establish an online exchange and mutual assistance platforms to connect the outgoing labour force with those left behind in villages; strengthen the connection between the outgoing labour force and their hometowns; and send signals to the outgoing labour force that they are valued and needed, thus, making them more willing to return to their hometowns for development owing to a sense of responsibility. In addition, collective activities are a way to enhance feelings and maintain good relations. During the time when a large number of migrant workers return home, such as the Spring Festival and Mid-Autumn Festival, activities can be held to maintain and cultivate feelings.

4.2.3. Strengthen social support and promote the return of rural labours

Formal support has played a moderating role in the relationship between rural migrants’ perceived value and return intention to hometown. Therefore, the rural government should have a certain failure relief mechanism for labourers who return to their hometowns for employment and entrepreneurship and establish a professional entrepreneurial guidance team jointly with investment enterprises to provide targeted guidance to solve problems or help entrepreneurial teams analyse the reasons for failure. In addition, the government should pay greater attention to the social security of the returning labour force, use the current information technology to speed up the improvement of social security and medical insurance transfer mechanisms in different places, establish a comprehensive medical insurance system, set up a minimum appeasement system for unemployment and pensions, and make up for the shortcomings of public services. In terms of educational resources, the interaction between rural and urban areas should be strengthened. Urban primary and secondary schools should assist rural schools and promote the development of rural education. Simultaneously, the rural government should cooperate with the provincial universities in conjunction with local education's actual needs and formulate policies and measures to promote college students to stay and find employment in the countryside.

Although informal support did not played a significant moderating role in the relationship between rural migrants’ perceived value and return intention to hometown, it has a direct and significant positive impact on the willingness of rural labour to return. China has always been an acquaintance society, especially in rural areas. Family, relatives, friends, neighbours, etc. are right next to them. After returning to their hometown, rural labourers no longer have to endure the pain of missing their hometown and family, thus obtaining emotional comfort, which is often difficult for them to obtain in cities. When they encounter difficulties in their hometown, they can receive help from relatives, friends, neighbours, etc. Therefore, rural governments should take measures to enhance informal support. Mobilize rural cadres, CPC party members, and other individuals with strong influence to regularly inform rural labourers working outside of their hometown of the development situation, various support policies, projects, and funds provided by the government. In traditional festivals such as the Spring Festival, Tomb Sweeping Day, Dragon Boat Festival, Mid-Autumn Festival, etc., the rural government should actively organize people who stay in their hometown to plan fellowship activities, invite returning labour force to participate, and provide opportunities for everyone to communicate with each other, increase trust, and enhance feelings. In addition, after the labour force returns to their hometown, it is also necessary to fully leverage the role of village committees and other organizations to help resolve various conflicts between family members and neighbours, maintain a harmonious atmosphere in the countryside, and create a rural environment that can attract, retain, and lead a happy life.

4.3. Future prospects

Based on the perceived value theory, this paper explores the influence of each dimension of the rural labour force's perceived value on their willingness to return and the moderating role of social support. However, the influencing factors perceived by the returning labour force are complex and variable, and there are several aspects that this study has not been able to cover. First, many subordinate counties and rural areas questionnaire distribution should be expanded to obtain more sample research surveys and increase the persuasiveness and generality of the research. Second, the study examined fewer external influencing factors, and future studies should consider the influence of realistic and objective factors and conduct comparative analysis. Finally, the study only investigated the factors behind rural labourers’ willingness to return to their hometowns. In the future, another survey should be conducted on the returning behaviour of labourers in this study area to analyse whether their behaviour is consistent with their willingness. If there are behaviours that are not consistent, it is necessary to further explore the influencing factors to make the research conclusion more comprehensive.

Author contributions

Lei Wang contributed to the conception and model construction of the study and paper revision and improvement; Zhaoxin Zhang contributed to the questionnaire survey, empirical analysis, and manuscript writing.

Institutional review board statement

The study was conducted by the Declaration of Helsinki and approved by the Ethics Committee of ***.

Acknowledgements

We are indebted to the anonymous reviewers and editors.

Data availability statement

The data presented in this study are available upon request from the corresponding author.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Additional information

Funding

This work was supported by Science and Technology Commission of Shanghai Municipality: [grant no 19YF1437000]; Shanghai Academy of Educational Sciences: [grant no C2021064].