The impact of digital identity on the labor supply of the young elderly: an empirical study from China

Introduction

In recent years, the new generation of Internet information technology has been rapidly updated and iterated. It has not only promoted digital transformation in the production field such as intelligent production methods, digitalization of industrial forms, changes in industrial organization platforms, and the creation of new work models and new jobs. It also promotes digital transformation in the areas of life such as smart healthcare, digital learning, smart transport, smart home, online socializing and online shopping. At the same time, the traditional lifestyles of the elderly have been impacted or subverted, resulting in the passive integration of many elderly into digital social life. As a result, labour supply at the intersection of population ageing and digital transformation has become the focus of research in various countries.

In China, due to a number of factors such as rapid aging of the population, the lag in the construction of the social security system and the urban-rural dichotomy, labour supply is not only an important way of social participation for the young elderly (the retired elderly between 55 and 70 years of age), but it also has a positive impact on the improvement of the economic situation, social circle, physical health and psychological well-being of the young elderly (Zhang X.X., and Nedospasova O., 2023) [1]. An analysis of the existing literature reveals that scholars' findings on the evolution and scope expansion of identity theory indicate that human identity changes significantly with changing times (technological innovations, economic scale, development patterns, etc.), which leads to the emergence of new identities; while other scholars have confirmed that Internet use has a significant impact on labour participation, community activities, volunteerism and family care. Based on these two theoretical perspectives, the authors propose a brand-new concept, digital identity, which provides a new way of thinking to unravel the mystery that constrains the labour supply of young seniors. Accordingly, the authors propose four hypotheses:

Hypothesis 1: Internet use has a positive effect on the labor supply of the young elderly.

Hypothesis 2: The higher the frequency of Internet use, the greater the possibility of labor supply for the young elderly.

Hypothesis 3: The social security system has an impact on the labor supply of the young elderly.

Hypothesis 4: There is significant urban-rural heterogeneity in the labor supply of the young elderly.

In order to test these four hypotheses, firstly, the mechanism of the identification of the digital identity of the elderly was dissected from the perspectives of triggering, catalysing, and shaping, and its mechanism of action on the labour supply of the elderly; then, a number of proxy variables were selected from the CGSS2017 database and the above hypotheses were verified by using the Zero-Inflated Poisson model in the Stata 17 software.

Literature review

Studies related to identity theory. Some philosophers and psychologists have argued that a person's identity is shaped by his (or her) interactions with others and the social environment (Busacchi V. and Martini G., 2021) [2]. Identity is an understanding of ‘who I am’ (Glover J., 1988; Ludwig A., 1997) and a reflection on ‘the sense of belonging to particular attributes’ (Alanen L., 1999) [3-5]. It is the ability to recognise ‘common features with others or groups’ (Hall S., 1996) [6], which develops gradually through internalisation of cultural and linguistic rules and interactions (Mead G.H., 1934) [7]. Individuals experience opposition, conflict and integration in this process, gradually recognising their freedom and independence (Hegel, 1807) [8]. In addition, identities are dynamic systems that change and evolve after formation (Freud S., 1949; Cordon L., 2018) [9-10]. Sociological psychologists divide identity into self-identity and social identity. Self-identity is concerned with the confirmation of individual identity, the theory of self-development suggests that the self develops gradually from physiological to social to psychological (Allport G., 1961) [11], and the theory of self-identity suggests that its formation is a central task of adolescence, a mental process of convincing oneself that I am one and not the other (Erikson E., 1998) [12]; and social identity explains the formation of the individual's identification with the group (Tajfel H. and Turner J.C., 1979). Social identity theory refers to identification based on individual qualities as personal identity and identification based on group identity as social identity (Turner J.C. et al., 1987; Hogg M.A., 2016) [13-15]. In addition, some scholars of communication and economics have argued that technological advances, political upheavals, and economic development have prompted people to adjust or integrate their identities in accordance with the macro-environment (Meyrowitz J., 2015; Breakwell G., 2015; Akerlof G.A., Kranton R.E., 2010) [16-18], which, in turn, has led to a reshaping of societal roles and behavioural norms ( Kranton R.E., 2016) [19]. Clearly, as society develops, the understanding of identity evolves.

A study of the impact of Internet use on the elderly. Since the 1980s, the rapid development and diffusion of Internet communication technologies have driven the global economic and social transition to digitalisation, changing the production and life scenarios of social groups (Castells M., 2006) [20]. The application of the Internet provides new opportunities for older people's identity. On the one hand, Internet use is beneficial for older people's physical and mental health, helping with self-health assessment (Tavares A.I., 2020) [21], increasing the frequency of physical activity and thus improving physical health (Nakagomi A. et al., 2022) [22]; enhancing older people's connectedness (Wang H. and Wellman B., 2010 ) [23], reducing loneliness (Patrícia S. et al., 2022) [24], and enhancing life satisfaction (Du P., 2020) and mental health (Shelia R. et al., 2014) [25-26]. In addition, Internet use improves e-health literacy among older adults, which helps prevent and manage chronic diseases (Li P. et al., 2023) [27]. On the other hand, the Internet has become an important tool for older adults to continue socialising (Chen X.L., 2015) [28]. Through the Internet, older adults enhance information acquisition and cognitive abilities (Kamin S. and Lang F., 2020) [29], broaden social networks (Barbara N. and Jaime F., 2015) [30], and accumulate social capital (Barbosa B. et al., 2018) [31]. Modern technology training helps to improve the psychosocial environment of older adults (Taylor M. and Bisson J., 2021) [32] and enhance human capital, self-efficacy, and motivation for social participation (Jin Y. and Zhao M., 2019) [33]. With increased frequency of Internet use and skills, older adults become more socially active (Grigoryeva I. and Kelasev V., 2016) [34], and re-employment and community activities become important ways for older adults to realise their self-worth (Li B. et al., 2023) [35].

However, the negative effects of digital transformation on the elderly cannot be ignored. On the one hand, the elderly are affected by objective factors such as physiological decline, relatively low level of education, reduced interest in accepting new things, insufficient mastery of digital skills, and fewer training avenues, which in turn creates psychological resistance or rejection of digital use (Berkowsky R., 2018) [36]. On the other hand, a certain degree of internal powerlessness exists among the elderly due to the faster updating of digital technology, the low degree of aging of digital devices and software, the existence of loopholes in digital security, and the difficulty in distinguishing the authenticity of digital content (Gulbrandsen K. & Sheehan, M., 2020) [37]. Both of these aspects significantly exacerbate the digital divide among the elderly (Friemel, 2016) [38].

In conclusion, multidisciplinary research has shown that identity shifts with changes in the objective world, emphasising the interaction between society and self. It is clear that the widespread use of digital technology will lead them to a new identity, i.e. digital identity identification.

Theoretical mechanisms and research hypotheses

So, what exactly is digital identity identification? This article defines digital identity identification as: personal display in digital society (such as showing your interests, opinions, etc. on platforms such as social media, blogs, forums, etc.), social interaction (such as communication, comments, likes, sharing, etc.), Internet behavior (such as search, shopping, reading, etc.), community ownership (such as adding online interest groups, communities, etc.) and other online activities formed by digital social identity (or character recognition) formed. It can reflect personal characteristics such as personal interests, values, social circles, etc., and can also reflect the brand image and value proposition of an organization or group. Moreover, with the continuous iteration of digital technology, people's digital identity will also change further.

(1) The spread of digital technology as a trigger for the identity crisis of the elderly. When individuals or groups are faced with new circumstances, challenges or pressures, they may begin to question their identity, feel confused, anxious or lost, and lead to significant changes in mood and behaviour. It is only by resolving psychological conflicts that self-identity can be achieved and the desired inner confidence can be gained (Erikson E., 1968) [39]. Currently, the rapid development of digital technology brings convenience and innovation to people's lives, but it also brings challenges to the elderly. Older people may feel marginalised due to technological complexity, lack of relevant knowledge and skills, and cultural differences. Learning and adapting to new technologies can be a serious challenge for the elderly, and this identity crisis may have a negative impact on their self-esteem and mental health (Matthews S., 2008) [40]. It can be seen that the popularity and widespread use of digital technology has triggered an identity crisis among the elderly.

(2) Acceptance of digital technology as a catalyst for the digital identity identification of the elderly. The Technology Acceptance Model (TAM) suggests that an individual's acceptance and use of new technology is primarily influenced by perceived utility and perceived ease of use (Davis, F., 1989) [41]. In the current rapid transformation of the world to a digital society, the living environment and social connections of the elderly are also covered by digital technologies (Shalagina G. & Shalagin S., 2022) [42]. For the elderly, on the one hand, acceptance and use of digital technology is only possible if they perceive it to be of practical value to their daily lives. On the other hand, the ease of learning and operating digital technologies is a key consideration for acceptance and use by the elderly. In addition, factors such as personal attitudes, social relationships and experiences may also affect their acceptance and use of digital technologies. In this scenario, acceptance of digital technology plays a catalytic role in the digital identity identification of the elderly.

(3) Identifying with digital technology as a shaping agent of digital identity identification for the elderly. Information Technology Identity (ITI) theory suggests that individuals view information technology as a self-concept over a long period of interaction with it (Carter M., 2012) [43]. As Information Technology Identity is not only a product of an individual's interaction history, but also an important force that influences his/her thinking and behaviour. Typically, IT identity affects the digital identity of the elderly through three dimensions: relevance, emotional engagement, and dependence. First, relevance refers to the elderly's integration of information technology into their self-concept and their identification with its importance in their lives. Second, affective engagement refers to seniors' emotional attachment to information technology, which motivates them to use digital technology more actively. Finally, dependence implies that seniors rely more on digital technology to fulfil a wide range of social activities. It is clear that information technology identity plays a significant role in shaping the digital technology identity of the elderly through relevance, emotional engagement and dependence (Emelyanova O., 2023) [44] and in order to create a solid foundation for the digital identity identification of the elderly.

Source: drawn by author

Figure 1 - The mechanical flowchart of the elderly digital identity identification and social participation

(4) The labor supply of the elderly is an important way of digital identity identification. Numerous studies have shown that economic pressures are the main reason why elderly re-enter the labour market. This has led some elderly to work to cope with the decline in quality of life caused by reduced income and lower pension replacement rates. Currently, with the wide application and constant updating of digital technology and network communication, people's needs for information acquisition, dissemination and utilisation in the real society and digital virtual world have gradually merged, initially forming a digital society (Wang F. & Guo L., 2022) [45]. On the one hand, the spread of digital technology has provided an important foundation for the labour supply of elderly, allowing them to use digital technology to find work or conduct business from home via the Internet. On the other hand, the development of the digital economy has created many new occupations and more flexible working patterns, enabling elderly to better balance work and family life. It is evident that labour supply under economic pressure has become an intrinsic driver of elderly's digital identity identification. Overall, the labour supply of elderly in the digital age not only reflects their adaptation and acceptance of new things, but also intuitively reflects their digital identity identification.

Data sources, variable selection and model construction

Analyzing the impact of digital identity on the labour supply of low-income older people requires the collection of survey data on their acceptance, proficiency and evaluation of digital technology, as well as data on labour hours, income, and personal, family and social characteristics (George B., 2019) [47]. Therefore, the authors chose the ‘Chinese General Social Survey 2017)’ data set, which has more than 12,500 valid samples, contains more than 780 variables, and includes a large amount of data on Internet usage. Since the survey respondents in this paper are limited to 55 to 70 years old, after removing the missing values, a total of 3464 valid samples were obtained.

(1) Dependent variables. Because the labor supply is the decision of the working hours that the family or individual is willing to provide under certain wage rates ( Hicks J.R., 1963 ) [48]. Therefore, we select the CGSS2017 data set about ' how many hours do you usually work per week? ' to measure the labor supply situation of the young elderly, named Labor Supply (LS). According to the survey data, the average weekly working hours of employees in Chinese enterprises in 2016 was 46.1 hours, but the working hours of residents in rural areas were generally more than 65 hours [49] . Therefore, the average working time of the young elderly is about 56 hours per week.

(2) Independent variables. As it can be seen from the previous section, digital identity identification is the result of an individual's acceptance and identification with digital technology, so the questionnaire items related to digital identity identification in the CGSS2017 data set were selected. It mainly includes Online social interaction, Online Self-Presentation, Online rights protection, Online Leisure and Entertainment, Online Information Access and Online Trading ( see Table 1 ). The answers to these six questions are all composed of ordered data of 1, 2, 3, 4 and 5, representing 'never', 'rarely', 'sometimes', 'often' and 'always ‘respectively.

Table 1 Independent variable related content

(3) Control variables. In terms of personal characteristics, gender, age, physical health status(PHS), educational level(EL), and Marital status(MS) are the factors that labour supply of the elderly is based on (French E., & Jones J. ,2017 ) [50]. Subjective well-being is an important manifestation of the mental health status of the elderly, named Mental Health status(MHS); reading in the leisure time of the elderly is an effective way to improve their cognitive ability (Heckman J. et.al., 2006) [51], named Cognitive Ability(CA). For family characteristics, the number of children of the elderly was used as a proxy variable for the family size of the elderly, named Children.The income data of the elderly's family throughout the year was used to measure the family economic status (HES). In terms of social characteristics, the subjective perception of socioeconomic status was used to measure the judgement of social class position and identity (Zheng L. & Xu M., 2022) [52] , named socio-economic stratum (SES). The frequency of the elderly's participation in offline social interaction activities is used to measure their social capital status, named Social Capital (SC). Social security such as Medicare and Pension have important impacts on the retirement decisions of the elderly (Ceni R., 2017; French E. & Jones J., 2011) [53-54], named Medical insurance(MI) and Pension, respectively.In addition, the long-term place of residence is used to measure the difference in the labour supply of the elderly in urban and rural areas, named Living Place (LP).

From the perspective of the selected data type, since the dependent variable Labor Supply ( LS ) belongs to the count variable, the main value is [ 0, 1, 2, ... n ], and the proportion of 0 is more than 74.4 %. Therefore, the Zero-Inflated Poisson (ZIP) model is used for regression analysis. Since the goal of the ZIP model is to estimate the zero-inflated probability and the mean parameter of the count part, the distribution of the count data and the probability of occurrence of the zero value are predicted. By adding the independent variable Xi ( i = 1, 2, ..., n ) and the control variable Z_j ( j = 1,2, ..., m ), the probability function can be expressed as :

In equation (1), Y is the count data of an event, P(Y=0) is the probability that the event does not occur, and P(Y=y) is the probability that the event occurs y times. α is the intercept of the zero- inflated portion, β is the coefficient of the independent variable X, γ is the coefficient of the control variable Z, λ is the parameter of the Poisson distribution, e is the base of the natural logarithm, y is the count value, and y! indicates the factorial of y.

Therefore, the function after adding dependent and control variables can be expressed as:

In Equation (2), β is the coefficient of the independent variable X_i and γ_j is the coefficient of a number of control variables such as Gender_j , Age_j , PHS_j , EL_j , CA_j , MS_j and MHS_j .

Empirical analyses

Table 2 shows that the median weekly labour supply of the low elderly is about 8.5 hours. They frequently use life-related digital technologies such as OSI, OSP, OLE, and OLA, but less often use complex ORP and OT.In terms of individual characteristics, the median age of the young elderly is 62.6 years old, they are in good health, and most of them are married but with lower education and cognitive ability. In terms of family characteristics, the median number of children was about 2, and the median annual income was about $60,698. In terms of social characteristics, they have a median perceived socioeconomic class of 2.14, a median social capital of about 2.7, a preference for offline socialising, a median of 0.93 and 0.81 for having health insurance and pension respectively, and a high proportion of urban permanent residents.

Table 2 Descriptive statistics of variables

In Table 3, Model (2) shows that Online Self-Presentation (OSP) has a significant positive effect on the labour supply of the young elderly (p-value = 0.006).Online Self-Presentation refers to the presentation of one's abilities, experiences and achievements through platforms such as the Internet and social media. For the lower-aged elderly, online self-presentation can increase their visibility and influence, and share their experience and wisdom with others. This can not only increase the self- confidence and self-esteem of the young elderly, but also stimulate their creativity and motivation to participate in social activities. Model(4) shows Online Leisure and Entertainment (OLE) positively impacts younger elderly labour supply (p-value=0.038). OLE provides relaxation, reduces pressure, and enhances mental health, increasing labour motivation. It also offers social interaction, reducing loneliness and expanding social support. Cognitive activities in OLE maintain cognitive function, benefiting labour participation. Model(6) shows Online Trading (OT) positively affects younger elderly labour supply (p-value=0.000). Digital technology offers new job opportunities for older adults to use their skills online, making work hours flexible and improving quality of life. It enhances social participation and mental health while providing opportunities to learn new digital skills.

The positive effects of Model (2), Model (4) and Model (6) on the labour supply of the young elderly suggest that the higher the frequency of the young elderly's use of digital technology the higher the probability of their labour supply. At a deeper level, the process of accepting and identifying with digital technologies such as Online Self-Presentation, Online Leisure and Entertainment, and Online Trading is a concrete manifestation of the digital identity identification of the young elderly. Moreover, as the digital identity of the young elderly increases, the likelihood of their labour supply increases significantly. These regression results also show that the authors' hypotheses 1 and 2 are valid.

Table 3 - Descriptive statistics of variables

Source: Authors' calculations based on CGSS2017 database

In terms of individual characteristics, older men are more likely than women to have access to labour opportunities because of their digital identity. Good physical and mental health contributes to young elderly's labour supply; married people are more likely to have access to labour supply than unmarried people, while age and education are the main barriers. Although the number of children has a negative effect on labour supply, it is not significant; household income has a positive effect. Lack of social resources significantly limits labour supply, and having public health insurance and a pension reduces the likelihood of labour supply, with health insurance having a particularly strong effect. young elderly living in urban areas are more likely to participate in the labour force than in rural areas, as there are more jobs in urban areas, which supports hypotheses 3 and 4.

In Table 4, the Incident Rate Ratio (IRR) of the digital identity identification of the young elderly for their labor supply shows that for every unit increase in OSP, the possibility of their labor supply increases by about 4.5 %. For every unit increase in OLE, the possibility of labor supply increases by about 2.9 %. For every unit increase in OT, the possibility of labor supply increases by about 6.7 %.

Table 4 - Incident rate ratio of digital identity identification of young elderly on their labour supply

Digital Identity Identification of the young and old is a combined result of multiple digital technologies. The regression results of the composite variable Digital Identity Identification (DII) better show its effect on labour supply. The regression results of Model (7) in Table 2 show that DII has a significant positive effect on the labour supply of the under-aged elderly (p-value = 0.001). for every 1 unit increase in DII, the likelihood of labour supply increases by about 0.91% (see Table 4). This proves the robustness of the model and the validity of the hypothesis.

Conclusions

Empirical results show that Online Self-Presentation, Online Leisure and Entertainment, and Online Trading significantly promote young elderly labor supply. Digital technology acceptance enhances their digital identity, impacting social labor participation. Although OSI, Online Information Access, and Online Rights Protection are crucial for digital identity and social participation, their regression results lacked statistical significance, possibly due to sample size and data processing issues. Future studies should design better questionnaires and ensure adequate sample size and quality for reliable data.

At present, the development of the digital society is irreversible, and the young elderly are bound to accept a variety of digital technologies and gradually agree with the positive effects of digital technology, thus realizing their digital identity identification. Moreover, the process of the young elderly's digital identity identification will also trigger significant changes in their social participation behaviors and approaches. Therefore, in order to avoid the young elderly becoming marginalized in the digital society, they should be helped to quickly adapt to and accept their digital identity identification. Specifically, the government should clear the obstacles for the young elderly to cross the digital divide through policy support, improving infrastructure, and building a digital literacy education system; while enterprises should pay attention to the age-adapted modification of digital hardware and software to adapt to the usage needs of the young elderly. The younger generation in the family should proactively convey digital thinking, digital skills, and cybersecurity awareness to the young elderly, inspire confidence in the use of digital technology, and drive the young elderly to better adapt to digital life.