The Conditional Effects of Internet Use on Well-being: Heterogeneous Trust Mechanisms and the Boundary of Subjective Social Status

The Conditional Effects of Internet Use on Well-being: Heterogeneous Trust Mechanisms and the Boundary of Subjective Social Status | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article The Conditional Effects of Internet Use on Well-being: Heterogeneous Trust Mechanisms and the Boundary of Subjective Social Status Qing Wang, Xuebo Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9225549/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 11 You are reading this latest preprint version Abstract The increasing integration of the internet into public life has ignited an intense academic debate regarding whether digital technology fundamentally fosters or erodes social capital and subjective well-being. To address this complexity, our study rigorously utilized the Lagged Fixed Effects (LFE) model and multi-period panel data from the China Family Panel Studies (CFPS) to examine the long-term, cross-period causal structure of internet use and well-being. The research confirms a highly conditional and heterogeneous dual-track mechanism. While internet use generates psychological costs via family and neighbourhood trust (bonding capital), its ability to foster stranger trust (bridging capital) is strictly contingent upon subjective class identity. However, even for individuals with higher class recognition who successfully accumulate this bridging trust, it ultimately lacks the emotional depth to translate into long-term subjective well-being. Crucially, the study found a persistent barrier preventing this interpersonal trust from converting into institutional trust, despite political trust being the key driver of long-term well-being. By uncovering these asymmetric mechanisms and boundary conditions, this research significantly advances social capital theory and social comparison theory in the digital context. The findings provide crucial evidence for understanding the digital well-being paradox, offering targeted insights for public governance aimed at mitigating the unequal distribution of digital benefits across different social strata. trust heterogeneity internet usage interpersonal trust political trust lagged fixed effects Figures Figure 1 Figure 2 Figure 3 Introduction The pervasive integration of the internet into public life has sparked intense academic debate regarding its impact on subjective well-being (SWB). While the digital environment empowers connectivity, it concurrently risks alienating individuals, underscoring a persistent digital well-being paradox (Diener et al., 1999 ; Diener et al., 2003 ; Yap & Lim, 2024 ; McLean et al., 2025 ; Silchenko, 2025 ). This scholarly divergence suggests that the transition from digital connectivity to psychological well-being is not linear, but rather governed by complex, often contradictory, mediating mechanisms. To unravel this paradox, extant research frequently utilizes Social Capital Theory, positing that internet use fosters social ties and trust, subsequently enhancing SWB. However, a critical limitation is the tendency to treat trust as a monolithic construct. This unidimensional view overlooks the profound theoretical insight that trust is inherently heterogeneous. This paper addresses this underdeveloped issue, demonstrating that trust heterogeneity can actually flip the sign of internet use effects. Treating trust as a single mediator conceals both the benefits and the psychological costs of digital engagement, particularly within China’s unique differential trust structure (Fei, 1992 ; Li, 2012 ). Accordingly, this study explores two core questions: (1) How does internet usage influence SWB through the distinct mediating pathways of stranger (bridging), family-neighbour (bonding), and political (institutional) trust? (2) How does subjective class identity condition these mechanisms? This study is theory-driven, with social capital theory serving as the focal lens for explaining how internet use shapes trust-based resources. Differential trust pattern theory complements this lens by clarifying why trust is not a single construct in China and why trust directed at close ties, generalized others, and institutions operate through distinct mechanisms. Social comparison theory further explains why subjective class identity conditions whether internet-enabled exposure translates into beneficial trust. Theory selection strictly follows the IMPACT logic (Hollebeek et al., 2025 ): interestingness lies in explaining why the same digital behaviour generates both social gain and psychological cost; matching aligns each trust dimension with its theoretical mechanism; parsimony limits the framework to the necessary trust pathways; applicability links findings to concrete well-being concerns; conceptual rigor specifies clear boundary conditions; and testability anchors each claim to panel-based empirical tests. Utilizing four waves of longitudinal data from the China Family Panel Studies (2016–2022) and employing a lagged fixed-effects approach, this research estimates cross-period associations while mitigating unobserved time-invariant confounding. Moving beyond the limitations of cross-sectional scale data, we introduce subjective class identity as a critical boundary condition. Guided by the typologies of theory development (Lim, 2026 ), this study contributes through theoretical modification and extension. First, we extend social capital theory by uncovering the hidden costs of digital connectivity. We critically demonstrate that internet-enhanced bonding trust (family and neighbours) acts as a psychological burden in a digitally mediated environment, highlighting the cultural specificity of strong ties. Second, we expand boundary conditions by demonstrating a structural paradox in digital bridging trust. While higher subjective class identity enables individuals to successfully convert internet use into stranger trust, this digitally acquired capital remains emotionally superficial, failing to deliver SWB gains across all class strata. Theoretical interestingness arises in revealing the illusion of digital social capital, as the internet reinforces structural inequalities in network building, yet these networks ultimately prove hollow for deep psychological fulfilment. Third, we contribute to political trust literature by clarifying the boundaries of digital social capital. We reveal that digitally formed interpersonal trust struggles to cross structural barriers to enhance institutional (political) trust, clarifying which trust pathways are actually available for policy leverage. Literature Review Internet Usage and Subjective Well-Being As internet technology increasingly permeates public life, its impact on subjective well-being (SWB) is highly complex and multifaceted. The internet offers significant benefits by enhancing political transparency, daily efficiency, and interpersonal connections (Becchetti et al., 2021; Büchi et al., 2018; Yang et al., 2022; Yoon, 2014; Zhang & Li, 2023). It facilitates instantaneous, low-cost communication that transcends time and space, provides broad access to information, and fosters diverse community support, particularly for marginalized groups (DiMaggio et al., 2001; Cardozo, 2023). However, these benefits are countered by documented risks. Excessive internet use is associated with fragmented attention, social comparison, and cyberbullying, which can persistently erode mental health and trigger anxiety and loneliness (McCrae et al., 2017; Shensa et al., 2020). For instance, internet-addicted groups show pronounced levels of depression and suicidal ideation (Seabrook et al., 2016). While digital engagement serves as a valuable resource for connectivity, it simultaneously acts as a catalyst for social comparison and psychological burden (De Hesselle & Montag, 2024; Mathy & Cooper, 2003). Consequently, positing a strictly unidirectional effect on SWB is theoretically insufficient. Aligning with the digital well-being paradox, the direct effect of internet use on well-being may be deeply contested. Therefore, rather than assuming a uniform outcome, this study proposes the following competing hypotheses to capture the baseline relationship before introducing our trust-based mechanisms: Hypothesis 1a (The Enhancement Hypothesis): Internet usage is positively associated with residents' subjective well-being. Hypothesis 1b (The Burden Hypothesis): Internet usage is negatively associated with residents' subjective well-being. Internet Usage and Heterogeneous Interpersonal Trust: Bridging Gains and Bonding Costs The classic distinction between particularistic (bonding) trust and universalistic (bridging) trust is central to understanding Chinese society's differential trust pattern (Fei, 1992). Universalistic trust extends to generalized others and non-kin (strangers), forming bridging social capital, while particularistic trust is based on strong kinship or geographical ties (family and neighbors), forming bonding social capital (Coleman, 1988; Portes, 2000; Putnam, 2000; Hall et al., 2021). While traditional frameworks often assume the internet uniformly enhances interpersonal trust, differentiating these trust targets reveals a divergent mechanism. Regarding bridging trust, the prevailing view indicates a positive association. The internet functions as a low-cost, boundary-spanning tool that effectively reduces communication barriers, allowing individuals to connect with heterogeneous groups and accumulate stranger trust (Ratan et al., 2010). Because this expanded social network provides novel social support and fulfills psychological needs, it translates into higher subjective well-being (SWB) (Martínez et al., 2019; Li et al., 2025). Therefore, we hypothesize: H2: Internet usage positively influences stranger trust. H3: Internet usage exerts a positive indirect effect on subjective well-being through the mediation of stranger trust. Conversely, the impact of internet use on bonding trust (family and neighbors) may paradoxically manifest as a hidden psychological cost. In the digital age, ubiquitous social platforms (such as WeChat in China) facilitate perpetual connectivity, which can continuously reinforce and intensify interactions within primary social circles (Davidow, 2011). While traditional social capital theory assumes strong ties are unequivocally beneficial, digital "over-connection" fundamentally alters this dynamic. The constant online presence blurs personal boundaries, subjecting individuals to inescapable social surveillance, rigid reciprocal norms, and the heavy emotional labor required to maintain these strong ties (Van Bruyssel et al., 2024; Zhou, 2019). Consequently, rather than providing liberating social support, this digitally intensified bonding trust transforms into a psychological burden (Rotter, 1980; Horak et al., 2022). This inescapable network of localized obligations and constant social presence ultimately diminishes overall life satisfaction. Thus, we propose: H4: Internet usage positively influences family and neighbourhood trust. H5: Internet usage exerts a negative indirect effect on subjective well-being through the mediation of family and neighbourhood trust. Internet Usage, Political Trust, and Subjective Well-Being The widespread adoption of internet technology also exerts a complex influence on public political (institutional) trust. While the internet facilitates information flow, higher usage frequently exposes individuals to unverified critical discourse, governance flaws, or social inequalities, which can incrementally erode institutional confidence (You & Wang, 2020). Cross-national analyses consistently show that frequent internet use can widen the gap between heightened democratic expectations and perceived government performance, negatively impacting institutional trust (Howard, 2010; Norris, 2011). Political trust is a well-documented macro-level driver of SWB. High levels of institutional trust reduce perceived societal uncertainty by strengthening expectations of institutional safeguards and life stability (Cárcaba et al., 2022). It reflects public confidence in policy response mechanisms; when citizens believe their concerns are systematically addressed, their well-being significantly increases (Hamilton et al., 2016). Therefore, if internet usage compromises political trust, it subsequently removes a critical pillar of subjective well-being. Based on this mechanism, we propose: H6: Internet usage negatively influences political trust. H7: Internet usage exerts a negative indirect effect on subjective well-being through the mediation of political trust. Subjective Class Identity as a Boundary Condition Subjective class identity refers to an individual's self-perception and emotional identification with their position in the social hierarchy (Weber, 1968). This identity serves as a critical socio-psychological filter regulating this conversion (Hargittai, 2001). Individuals with a high subjective class identity generally possess greater psychological security, cognitive resources, and optimism (Uslaner, 2002). This enables them to use the internet strategically, approaching unfamiliar online environments with confidence to cultivate generalized trust and extract bridging resources (Miao, 2023). In contrast, those with lower subjective class identity often experience higher baseline vulnerability and resource anxieties (Zhang et al., 2025). For them, the digital environment may appear riskier, limiting their capacity or willingness to transform digital connectivity into genuine stranger trust (CAI & HOU, 2014; CHEN & YANG, 2025; Bagherianziarat & Hamplová, 2025) Therefore, subjective class identity acts as a boundary condition, determining whether online exposure successfully translates into beneficial forms of generalized trust. H8: Subjective class identity positively moderates the relationship between internet usage and stranger trust; specifically, the positive effect of internet use on stranger trust is stronger for individuals with higher subjective class identity. Therefore, we construct the following theoretical model (see Figure 1 ): Method The data for our study are sourced from the China Family Panel Studies (CFPS), a nationwide longitudinal fixed-sample tracking survey conducted by the Institute of Social Science Surveys (ISSS) at Peking University. This project systematically collects microdata at the individual, household, and community levels to monitor and analyse dynamic changes across multiple dimensions of Chinese society. Its findings primarily provide empirical evidence for academic discourse and public policy formulation. Specific survey topics encompass employment status, educational attainment, family structure and interactions, population mobility, and physical and mental health conditions. To maintain panel data continuity and variable consistency, exclusion method was applied to address missing values. The resulting valid sample comprises 14,328 observations from 3,582 residents tracked longitudinally across four waves of data: 2016, 2018, 2020, and 2022. Subjective well-being is the dependent variable in our study. The CFPS data poses the question, "To what extent are you satisfied with your life as a whole?" The scale ranges from 1 to 5, with 1 representing 'very dissatisfied' and 5 representing 'very satisfied'. Responses are scored on a scale of 1 to 5, with 1 representing the weakest response and 5 representing the strongest. Internet usage is the core independent variable in our research. In the context of surveying internet usage, the CFPS data typically poses the following questions: "Do you use mobile internet?" and "Do you use computer internet?" The values assigned to the responses were as follows: 1 for "yes" responses and 0 for "no" responses. Furthermore, the CFPS data inquired about the frequency of public consumption, daily activities, and shopping on internet platforms. The following scale was utilized: 1 = Almost daily, 2 = 3–4 times per week, 3 = 1–2 times per week, 4 = 2–3 times per month, 5 = Once per month, 6 = Once every few months, 7 = Never. The integration of these scores with the initial two questions resulted in the formation of a novel variable pertaining to internet usage (Blank & Groselj, 2014 ). To ensure that higher numerical values represent a higher frequency of digital engagement, the original frequency scales (1–7) were reverse-coded prior to aggregation. In our study, interpersonal trust and political trust function as mediating variables. Interpersonal trust is further categorized into two distinct types: stranger trust and trust in close relatives and neighbours. The CFPS data on stranger trust includes questions such as "How much do you trust Individuals from other countries?", "How much do you trust strangers?"; while questions on trust in close relatives and neighbours include "How much do you trust your parents?" and "How much do you trust your neighbours?" The CFPS data on political trust primarily comprises questions regarding the extent of trust placed in local government officials. The scoring scale employed for all trust surveys ranges from 0 to 10, where 0 denotes a high level of distrust and 10 denotes a high level of trust. The concept of subjective class identity is proposed as the moderating variable in our study. Subjective class identity was measured by asking respondents to rate their social status within their local community on a scale from 1 (very low) to 5 (very high). Furthermore, the study's control variables encompass gender (1 = male, 0 = female), age, personal income (logarithmically transformed), educational attainment (measured on an 8-point scale ranging from 1 = illiterate/semi-illiterate to 8 = doctoral degree), hukou registration (1 = non-agricultural, 0 = agricultural), residence (1 = urban, 0 = rural), and marital status (1 = married, 0 = unmarried), which are standard demographic controls identified in previous studies (Graham & Chattopadhyay, 2013 ; Rodríguez-Pose & Maslauskaite, 2012 ). To rigorously test the cross-period causal effect and address potential endogeneity, we employ the Lagged Fixed Effects (LFE) model. The LFE model incorporates the dependent variable lagged by one period (Yt-1) alongside the individual fixed effects (ui). This approach offers a significant advantage by simultaneously controlling for unobserved time-invariant individual heterogeneity and crucial dynamic endogeneity, which arises when the outcome variable influences the predictor in the subsequent period. Therefore, the LFE model provides a more robust and conservative causal estimate than simple OLS or standard Fixed Effects models when analysing the dynamic relationship between internet use and subjective well-being over time. Furthermore, to enhance statistical power and ensure the reliability of our findings, we utilized the Bootstrap method with 5,000 resamples to calculate the 95% confidence intervals for all indirect effects in the mediation and moderated mediation analyses. Results Descriptive Statistics The statistical characteristics of the variables suggest several notable patterns in the sample distribution (see Table 1). Internet usage has a mean score of 1.69, with its positive skewness (0.554) indicating a relatively dispersed usage pattern that leans slightly toward the lower end of the 1-to-3 scale. The mean score for subjective class identity is 0.80; its strong negative skewness (-1.683) across a wide range (-8.7 to 2.5) implies that while the average sits lower numerically, a heavy concentration of respondents identify with the higher end of their perceived class strata, offset by a long tail of individuals reporting significantly lower subjective status. Table 1 Descriptive Statistics for Each Variable Variable N Average Std. Dev. Skewness Min Max Subjective class identity 14328 0.80 0.96 -1.683 -8.7 2.5 Family and Neighborhood trust 14328 13.60 5.47 -1.395 1.0 19.0 Stranger trust 14328 7.48 4.39 0.035 1.0 19.0 Political trust 14328 5.38 2.64 -0.185 0.0 10.0 Subjective well-being 14328 3.90 0.99 -0.658 1.0 5.0 Internet Use 14328 1.69 0.73 0.55 1.0 3.0 Age 14328 51.92 13.62 -0.05 14.0 93.0 Gender 14328 0.56 0.50 -0.24 0.0 1.0 Education 14328 2.27 1.69 0.84 0.0 8.0 Hukou registration 14328 0.23 0.42 1.27 0.0 1.0 Residence 14328 0.50 0.50 -0.01 0.0 1.0 Marriage 14328 0.87 0.34 -2.17 0.0 1.0 ln Personal Income 14328 1.02 0.63 1.02 0 6.54 Regarding the trust measures, family and neighborhood trust (mean 13.60) is highly concentrated toward the top of the 19-point scale, supported by a strong negative skew (-1.395). This suggests that a majority of respondents tend to report high levels of trust within their close social circles. In contrast, stranger trust (mean 7.48) exhibits a more dispersed and highly symmetrical distribution (skewness 0.035), which reflects considerable variation and general caution in public attitudes toward trusting strangers. Political trust (mean 5.38) falls squarely within the middle range, with a slight negative skew (-0.185) showing a mild tendency toward institutional recognition. Furthermore, subjective well-being has a high mean of 3.90 on a 5-point scale, confirming that the overall reported happiness level in the surveyed group is generally elevated. Table 2 Pearson Correlation Matrix of Key Study Variables Variable Family/Neighbor Trust Stranger Trust Political Trust Subjective Class Well-being Internet Use Family/Neighbor Trust 1.000 Stranger Trust 0.138 *** 1.000 Political Trust -0.040 *** 0.230 *** 1.000 Subjective Class 0.135 *** 0.105 *** -0.049 *** 1.000 Well-being -0.089 *** 0.031 *** 0.237 *** -0.028 *** 1.000 Internet Use 0.137 *** 0.111 *** -0.043 *** 0.409 *** -0.073 *** 1.000 *p<0.05, **p<0.01, ***p<0.001 Table 2 presents the Pearson correlation coefficient matrix, revealing the complex structural associations among the core variables and laying the empirical foundation for subsequent mechanism analyses (see Table 2). Among the results, the most salient relationship exists between subjective class identity and internet use, which exhibit a strong positive correlation (r = 0.409, p < 0.001). This confirms that a higher perceived social stratum is a crucial precondition or correlate for individual digital engagement. Regarding the primary variable of interest, internet use demonstrates a heterogeneous association with different dimensions of trust. It is significantly and positively correlated with both family/neighbor trust (r = 0.137, p < 0.001) and stranger trust (r = 0.111, p < 0.001), yet it shows a negative correlation with political trust (r = -0.043, p < 0.001). This suggests that while internet usage may expand horizontal social networks and generalized trust, it might subtly erode vertical institutional trust. Furthermore, subjective well-being exhibits a highly differentiated pattern when associated with these trust dimensions. It is most strongly and positively correlated with political trust (r=0.237,p<0.001), indicating that institutional confidence is a vital psychological anchor for individual well-being. Conversely, well-being is significantly negatively correlated with family and neighborhood trust (r=-0.089,p<0.001), preliminarily hinting at the potential emotional burden or eroding effect of traditional acquaintance-based social ties. Notably, internet use itself is significantly negatively correlated with subjective well-being (r=-0.073,p<0.001), providing initial correlational support for the "well-being paradox" in the digital age. Collectively, these results demonstrate that the nexus among internet use, multi-dimensional trust, and well-being is intricately interwoven, characterized by a complex structure of coexisting positive network expansions and negative psychological impacts. Lagged Fixed Effects Model Statistics Our analysis aims to delve into complex cross-period causal effects and initially confirmed that core variables such as subjective well-being and social trust exhibit significant temporal persistence. To precisely capture this dynamic adjustment characteristic while strictly controlling for all time-invariant individual unobserved heterogeneity, we constructed a Lagged Fixed Effects model (LFE). Furthermore, to rigorously isolate the effects of interest, the model incorporates the 2016 baseline levels of all respective variables, alongside a comprehensive set of demographic controls including age, gender, education, hukou registration, place of residence, and marital status By utilizing the LFE model, we rigorously eliminated biases caused by all time-invariant individual characteristics, and the multi-period lagged structure (X2018→M2020→Y2022) ensures the temporal precedence required for causal inference. Therefore, employing this rigorous Lagged Fixed Effects approach, we proceeded with an in-depth empirical analysis of the net, cross-period causal effects of internet usage and various forms of trust on subjective well-being. H1b and H2 are supported, while H1a, H4, and H6 are not supported. Table 3 Lagged Effects of Internet Use on Trust and Well-being Model Coeff. Std. Error t p 95% CI [Lower, Upper] R 2 Internet Use (2018)→Family and Neighborhood trust (2020) 0.178 0.131 1.351 0.177 [-0.080, 0.435] 0.051 Internet Use (2018)→Stranger trust (2020) 0.460 *** 0.108 4.272 0.000 [0.249, 0.672] 0.039 Internet Use (2018)→Political trust (2020) -0.020 0.058 -0.339 0.735 [-0.134, 0.094] 0.113 Internet Use (2018)→Subjective well-being (2022) -0.043 * 0.022 -1.966 0.049 [-0.085, -0.001] 0.071 *p<0.05, **p<0.01, ***p<0.001 Note: Higher values for internet use indicate higher frequency due to reverse-coding. The lagged analysis examines the longitudinal impact of Internet use (measured in 2018) on different dimensions of social trust (measured in 2020) and subjective well-being (measured in 2022). The results reveal divergent cross-period effects of Internet engagement on individuals' social attitudes and psychological outcomes (see Table 3). First, the model demonstrates a highly significant and robust positive lagged effect of prior Internet use on stranger trust (Coeff. = 0.460, p < 0.001). This suggests that increased internet usage significantly fosters a broader generalized trust toward strangers over time. Conversely, the analysis reveals a statistically significant, albeit small, negative longitudinal impact on subjective well-being (Coeff. = -0.043, p < 0.05). This indicates that higher levels of Internet use in 2018 are associated with a slight decline in individuals' reported subjective well-being four years later in 2022. In contrast, earlier Internet use does not exert any significant delayed influence on localized or institutional trust. Specifically, the lagged effects on family and neighbourhood trust (Coeff. = 0.178, p = 0.177) and political trust (Coeff. = -0.020, p = 0.735) are statistically insignificant, with both of their 95% confidence intervals crossing zero. This implies that while Internet use reshapes attitudes toward strangers and personal well-being, it does not meaningfully alter individuals' baseline trust in their immediate social circles or the political system over time. Table 4 Lagged Effects of Social Trust Dimensions on Well-being Dependent Variable (2020) Coeff. Std. Error t p 95% CI [Lower, Upper] R 2 Family & Neighbor(2020) Trust→ Subjective well-being (2022) -0.006 * 0.003 -2.230 0.026 [-0.012, -0.001] 0.085 Stranger Trust (2020) → Subjective well-being (2022) -0.006 0.004 -1.602 0.109 [-0.012, 0.001] 0.085 Political Trust (2020) → Subjective well-being (2022) 0.046 *** 0.006 7.223 0.000 [0.034, 0.059] 0.085 *p<0.05, **p<0.01, ***p<0.001 Table 4 presents the findings from the lagged analysis examining how different dimensions of social trust in 2020 impact subjective well-being in 2022 (see Table 4). The results reveal that various forms of trust exert divergent longitudinal effects on an individual's future psychological outcomes. Most notably, political trust demonstrates a highly significant and robust positive lagged effect on subsequent subjective well-being (Coeff. = 0.046, p < 0.001). This indicates that individuals who reported higher levels of trust in political institutions in 2020 experienced significantly higher subjective well-being two years later. Conversely, family and neighbor trust exhibits a slight but statistically significant negative longitudinal impact on future well-being (Coeff. = -0.006, p = 0.026). This suggests that higher levels of trust concentrated within localized, primary social ties are associated with a marginal decrease in reported well-being over time. Finally, the lagged effect of stranger trust on subjective well-being is not statistically significant (Coeff. = -0.006, p = 0.109). Because the 95% confidence interval crosses zero, there is no conclusive evidence in this model to suggest that generalized trust toward strangers has a delayed, long-term impact on personal well-being. Overall, the model (which explains 8.5% of the variance, R 2 = 0.085) underscores that social trust is not monolithic in its psychological outcomes. Institutional (political) trust serves as a strong positive predictor for long-term well-being, whereas localized (family and neighbor) trust shows a minor negative association, and generalized (stranger) trust shows no significant temporal effect. Table 5 Mediation Analysis of the Effect of Internet Use on Subjective Well-being Effect Pathway Coeff. S.E. 95% CI (Lower, Upper) p Indirect Effect (via Family & Neighbor Trust) -0.005 * 0.003 (-0.011,-0.000) 0.038 Indirect Effect (via Stranger Trust) -0.001 0.004 (-0.008,0.007) 0.838 Indirect Effect (via Political Trust) -0.002 0.003 (-0.009,0.005) 0.602 *p<0.05, **p<0.01, ***p<0.001 Table 5 presents the Bootstrap mediation analysis evaluating the indirect pathways through which Internet use affects subjective well-being via three distinct dimensions of social trust (see Table 5). Among the examined mediators, only family and neighbour trust exhibits a statistically significant indirect effect (Coeff. = -0.005, p = 0.038). Although the direct lagged effect of internet use on family trust was statistically marginal (as shown in Table 3), the rigorous Bootstrap mediation test detected a significant, albeit weak, indirect pathway. In contrast, the indirect pathways through stranger trust (Coeff. = -0.001, p = 0.838) and political trust (Coeff. = -0.002, p = 0.602) are not statistically significant, as both of their 95% confidence intervals cross zero. H5 is supported, while H3 and H7 are not supported. These findings indicate that a portion of the adverse effect of internet use on well-being is indirectly channelled through its association with localized family and neighbourhood trust, whereas generalized and institutional trust do not serve as significant mediating mechanisms in this specific relationship. Table 6 Moderating Effect of Subjective Class Identity on Stranger Trust Mediator (M) Moderator (Subjective Class Identity) Level Est. 95% CI (Lower, Upper) p Stranger trust Low Subjective Class Identity 0.005 (-0.003,0.014) 0.185 High Subjective Class Identity 0.011 (0.004,0.019) 0.000 * *p<0.05, **p<0.01, ***p<0.001 Further analysis demonstrates that the internet's capacity to foster stranger trust is subject to significant boundary conditions (see Table 6 & Figure 2 ). For individuals with a low subjective class identity, internet use fails to significantly foster stranger trust (Est. = 0.005, p = 0.185). However, as an individual's subjective class identity level increases, this effect becomes highly significant (Est. = 0.011, p < 0.001), fully supporting H8. This demonstrates that the internet's potential to expand bridging social capital is selectively unlocked by a higher perceived social standing. Crucially, however, as indicated by the non-significant mediation pathway in Table 5, this unequally distributed bridging capital ultimately hits a dead end. Even for higher-status groups who successfully accumulate digital stranger trust, these weak ties lack the substantive emotional support necessary to effectively translate into subjective well-being. In conclusion, these findings confirm that an individual's subjective class level serves as a critical boundary condition. It fundamentally determines whether internet use can effectively promote the conversion of positive, generalized social capital and ultimately impact an individual's long-term subjective well-being. Discussion Our study rigorously examined the long-term and dynamic effects of internet use on subjective well-being (see Fig. 3 for the specific action pathway). Our findings move beyond the simple "foster or erode" debate by confirming that the effect of Internet on well-being is highly conditional and realized through heterogeneous pathways of social trust. By precisely revealing these distinct mechanisms and boundary conditions, our results redefine the Chinese pathway of social capital in the digital society and offer critical insights for social governance. Taken together, our research offers a sophisticated, multi-mechanism integrative explanatory framework for the long-term Internet and Well-being Paradox. The overall effect of the internet on well-being is negative. We also found the negative mechanism of the cost of family and neighbourhood trust erodes the well-being of all individuals. In the moderated mediation model, the internet's potential to foster generalized trust among strangers is only accessible to those with a high subjective class identity. This theoretical framework moves beyond a simple benefits-versus-harms debate, emphasizing the critical role of social structure and subjective stratification in the conversion of digital capital, and ultimately reveals the structural asymmetry of digital technology's impact on well-being. Advancing Hierarchical Political Trust Theory Our study highlights potential boundary conditions regarding the conversion of interpersonal trust into political trust in the digital age, thereby deepening the understanding of the hierarchical government trust pattern specific to the Chinese context. Existing theories have focused on the potential for a bottom-up conversion of micro-level interpersonal trust to macro-level institutional trust. For instance, Putnam suggests that high levels of trust among social members can be converted into support for the political system (Putnam, 1966 ; Norris, 2001 ). However, within China's singular social and digital ecology, the direct impact of the internet on political trust remains an area demanding theoretical discussion. Our findings extend theoretical assumptions about the simple spill over of social capital across public and private spheres, fostering deep dialogue with Lianjiang Li's work on hierarchical government trust (L Li,2012). Results show internet use fails to significantly enhance this form of trust. Our longitudinal model finds no significant evidence that general internet use enhances this macro-level institutional trust. This null finding suggests that even if the internet cultivates generalized trust, social capital formed through atomized digital connections may not naturally spill over to cross public-private boundaries and influence public authority (Uslaner, 2004 ; King et al., 2013 ; Bolsover, 2017 ; Kang & Zhu, 2021 ; Jhang, 2022 ; Zhan et al., 2025 ). This prompts reflection that within China's digital governance landscape, effectively activating and guiding the positive effects of social capital to better serve macro‑governance objectives remains an important ongoing topic. Extending Social Capital Theory On the dimension of strong ties, we confirmed the hidden costs of bonding social capital, thus extending the understanding of traditional social capital theory. Family and neighbourhood trust are typically regarded as classic examples of bonding social capital that support individuals' emotional needs and resource acquisition(D. S. Brisson & Usher, 2005 ; D. Brisson, 2009 ). While classical social capital theory consistently promotes the positive role of strong ties, whether these ties incur hidden costs in the digital age, remains a theoretical issue awaiting resolution. Moreover, previous studies are not enough to explain whether such costs are amplified within China's unique differential pattern of association. While the direct lagged effect of internet use on family trust appears statistically subtle in the aggregate model, our rigorous mediation analysis detects a significant, albeit weak, indirect pathway. This provides preliminary evidence suggesting that family and neighbourhood trust may possess the potential to act as a negative mechanism in the digital well-being paradox. It points to the possibility that, rather than unequivocally enhancing well-being, digitally intensified bonding social capital runs the risk of being alienated into a psychological burden. In China's internet ecosystem, where social software like WeChat are highly prevalent, the state of being always online may lead to over-connection (Davidow, 2011; Zhou, 2019 ). High levels of trust in a digital context can potentially translate into a heavy burden of human relations, the pressure of constant social presence, and intensified social norm constraints (Gargiulo & Ertug, 2006 ; Van Bruyssel et al., 2024 ). Therefore, by utilizing longitudinal data, our study reveals that bonding social capital may have become a negative mechanism in the paradox of digital age well-being, thereby supplementing the literature on the dark side of social capital. Although internet use successfully fosters stranger trust, this form of bridging social capital does not significantly translate into improved long-term subjective well-being. Despite the internet’s capacity to generate weak ties, the lack of substantial support and emotional depth in such capital prevents the mediation mechanism from being confirmed. This overall pattern does not represent a simple failure of mediation. Instead, our moderation hypothesis (H8) reveals a deep structural paradox. The transformation of internet use into bridging trust is restricted by individuals’ subjective social status and benefits only those with higher perceived class standing. Even so, such resource accumulation does not extend to improvements in well-being. The bridging social capital formed online operates only as a structural resource rather than an emotional resource. It fails to generate lasting improvements in subjective well-being, regardless of class identity. Expanding Social Comparison Theory Our research focuses on the potential of bridging social capital to promote well-being in the digital age, and introduces subjective class identity as a boundary condition for this analysis. The traditional technology empowerment hypothesis tends to suggest that digital tools enhance social capital (Choi et al., 2022 ). However, this hypothesis did not adequately explain how social stratification limits individuals' ability to convert digital resources into positive well-being outcomes(van Deursen & Helsper, 2015 ). To answer this question, our study finds that the different levels of subjective class identity interact with internet use on well-being, which engages in a profound dialogue with social comparison theory. Our finding addresses the debate of whether digital connectivity is a true social equalizer or merely a re-manifestation of social inequality on internet usage (Hargittai, 2011 ; Katz & Gonzalez, 2016 ). We argue that subjective class identity acts as a filter for social comparison. For those with high class identity, they tend to view the internet as a tool for acquiring heterogeneous information and expanding weak ties. Conversely, individuals with low class identity may be more susceptible to feelings of relative deprivation when encountering idealized lives presented online. Furthermore, while our broad measurement of internet use does not capture specific online activities, these statistical disparities can be contextualized within the structural reality of China’s digital economy. In this broader context, researchers have increasingly noted that the 'stranger connections' accessible to lower-status groups are often precarious weak ties shaped by algorithmic exploitation (e.g., the digital gig economy). This dual disadvantage of psychological deprivation and the inferior structural quality of their digital ties ultimately inhibits their capacity to convert online interactions into beneficial generalized trust. Therefore, our study emphasizes that subjective class identity significantly influences the conversion efficiency of digital capital, providing an important perspective for understanding social inequality in the digital age. Practical Implications The Lagged Fixed Effects model applied to our four-year panel data establishes robust causal identification, providing a solid foundation for targeted interventions. Enhancing national well-being requires synergistic, actor-specific efforts targeting distinct empirical mechanisms: First, directly addressing the negative mediating path of family and neighbourhood trust, digital platform developers must mitigate the psychological burden of over-connectivity. Developers (e.g., WeChat) should design architectural boundaries by incorporating features such as "off-duty" modes and granular notification controls. These technical adjustments empower users to manage heavy reciprocal obligations and limit constant social surveillance. Second, recognizing that digital bridging capital is structurally unequal yet emotionally insufficient for well-being, interventions must move beyond mere connectivity. While local governments should provide capacity-building to help lower-status groups safely navigate online networks to reduce resource stratification, society as a whole must actively rebuild offline, physically situated community engagements. Relying solely on the internet to foster generalized trust risks trapping individuals in emotionally hollow digital networks. Third, responding to the persistent barrier preventing interpersonal ties from converting into institutional trust, policymakers must actively optimize digital governance. Since political trust is the strongest long-term predictor of well-being, governments cannot rely on natural social capital spillover. Policymakers should establish highly responsive, trackable online public feedback mechanisms to ensure citizens' digital grievances receive substantive institutional responses, thereby consolidating this structural cornerstone of well-being. Limitation and Future Directions Our study confirms important empirical connections, yet the digital sphere invites further methodological and theoretical development. Subsequent research could improve the granularity of measurement by using fine-grained indicators of online purpose to pinpoint effective activities for fostering social trust. Future studies may also benefit from utilizing sophisticated serial mediation models incorporating psychological variables to fully reveal the intricate social comparison effects behind subjective class identity's moderating role. Finally, comparative analyses involving countries with differing social capital structures and governance regimes would be valuable to assess the external validity and boundary conditions of our findings, which are currently grounded in the unique socio-political context of China. Conclusion Our study confirms a conditional and heterogeneous dual-mechanism framework that resolves the digital age well-being paradox. The research reveals that the internet's influence is structurally asymmetric: it generates a negative path by amplifying the hidden psychological costs of family and neighbourhood trust. These findings advance social capital theory and social comparison theory by defining the costs and boundary conditions of digital capital conversion, ultimately providing a rigorous, structural explanation for the unequal distribution of digital benefits in the Chinese context. Declarations Author Contributions Qing Wang (WQ) and Xuebo Zhang (ZXB) conceived and designed the study. WQ performed the data extraction, conducted the statistical analyses (including the Lagged Fixed Effects modeling), and drafted the initial manuscript. ZXB supervised the entire research process, provided critical methodological guidance, and extensively revised the manuscript for important intellectual content. Both authors read and approved the final manuscript. Ethical Considerations Not applicable Consent to participate Not applicable Consent for publication: Not applicable Declaration of conflicting interest: No potential conflict of interest was reported by the authors. Funding statement: This research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors. Data Availability: The datasets analyzed during the current study, derived from the China Family Panel Studies (CFPS), are available at the following URL: https://docs.google.com/spreadsheets/d/1eJx-wyAqlR11zrJS3qgZiL2mmteI2Ehp/edit?usp=sharing&ouid=107616459617005785709&rtpof=true&sd=true. The original raw data are publicly available from the Institute of Social Science Survey (ISSS) at Peking University. References Bagherianziarat, A., & Hamplová, D. (2025). Life Satisfaction and Subjective/Objective Socioeconomic Status in European Countries: Does Affluence Matter? Fudan Journal of the Humanities and Social Sciences . https://doi.org/10.1007/s40647-025-00449-0 Becchetti, L., Florio, E., & Mancini, S. (2021). Internet Exposure and Social Capital. SSRN Electronic Journal . https://doi.org/10.2139/ssrn.3923612 Blank, G., & Groselj, D. (2014). Dimensions of Internet use: Amount, variety, and types. Information, Communication & Society , 17 (4), 417～435. https://doi.org/10.1080/1369118X.2014.889189 Bolsover, G. (2017). Harmonious communitarianism or a rational public sphere: A content analysis of the differences between comments on news stories on Weibo and Facebook. Asian Journal of Communication , 27 (2), 115～133. https://doi.org/10.1080/01292986.2016.1214743 Brisson, D. (2009). Testing the Relationship of Formal Bonding, Informal Bonding, and Formal Bridging Social Capital on Key Outcomes for Families in Low-Income Neighborhoods. Journal of Sociology & Social Welfare , 36 , 167. Brisson, D. S., & Usher, C. L. (2005). Bonding social capital in low‐income neighborhoods. Family Relations, 54(5), 644-653. Büchi, M., Festic, N., & Latzer, M. (2018). How social well-being is affected by digital inequalities. International Journal of Communication, 12, 21. Li, L. (2012). Differential government trust (Hierarchical trust in government). Twenty-First Century, (131), 108–114. Cai, H., & Hou, J. (2014). Subjective class identity and social trust: An empirical analysis based on CGSS 2010 data. Sociological Studies , (6), 20-42. https://doi.org/10.19934/j.cnki.issn1002-5936.2014.06.002. Cárcaba, A., Arrondo, R., & González, E. (2022). Does good local governance improve subjective well-being? European Research on Management and Business Economics , 28 (2), 100192. https://doi.org/10.1016/j.iedeen.2021.100192 Cardozo, G. T. (2023). Virtually united: The role of the internet in building social capital. AWARI , 4 , 1～12. https://doi.org/10.47909/awari.605 Choi, B., Park, H., & Joo, S. (2022). Digital technology use and social capital: The moderating effect of social class. Innovation in Aging, 6(Supplement_1), 478. Chen, Y., & Yang, Y. Y. (2025). Downward subjective class identification and relative deprivation among the middle-income group: An empirical study based on CSS2021. Social Sciences in China , 45(3), 123-146. https://doi.org/10.12422/cass.2025.0003. Coleman, J. S. (1988). Social capital in the creation of human capital. American journal of sociology, 94, S95-S120. Davidow, W. H. (2011). Overconnected: The Promise and Threat of the Internet . Open Road Media. De Hesselle, L. C., & Montag, C. (2024). Effects of a 14-day social media abstinence on mental health and well-being: Results from an experimental study. BMC Psychology , 12 (1), 141. https://doi.org/10.1186/s40359-024-01611-1 Diener, E., Oishi, S., & Lucas, R. E. (2003). Personality, Culture, and Subjective Well-Being: Emotional and Cognitive Evaluations of Life. Annual Review of Psychology , 54 (1), 403～425. https://doi.org/10.1146/annurev.psych.54.101601.145056 Diener, E., Suh, E. M., Lucas, R. E., & Smith, H. L. (1999). Subjective well-being: three decades of progress. Psychological bulletin, 125(2), 276. DiMaggio, P., Hargittai, E., Neuman, W. R., & Robinson, J. P. (2001). Social Implications of the Internet. Annual Review of Sociology , 27 (Volume 27, 2001), 307～336. https://doi.org/10.1146/annurev.soc.27.1.307 Fei, X. T. (1992). From the soil: The foundations of Chinese society . Sanlian Shudian (Joint Publishing Company). Fu, X. (2018). The Contextual Effects of Political Trust on Happiness: Evidence from China. Social Indicators Research , 139 (2), 491～516. https://doi.org/10.1007/s11205-017-1721-2 Gargiulo, M., & Ertug, G. (2006). The Dark Side of Trust. Handbook of Trust Research . Edward Elgar Publishing. https://www.elgaronline.com/edcollchap/9781843767541.00016.xml Graham, C., & Chattopadhyay, S. (2013). Gender and well-being around the world. International Journal of Happiness and Development . https://www.inderscienceonline.com/doi/10.1504/IJHD.2013.055648 Hall, J. A., Xing, C., Ross, E. M., & Johnson, R. M. (2021). Experimentally manipulating social media abstinence: Results of a four-week diary study. Media Psychology , 24 (2), 259～275. https://doi.org/10.1080/15213269.2019.1688171 Hamilton, K., Helliwell, J. F., & Woolcock, M. (2016). Social Capital, Trust and Well-being in the Evaluation of Wealth (Working Paper No. 22556). National Bureau of Economic Research. https://doi.org/10.3386/w22556 Hargittai, E. (2001). Second-Level Digital Divide: Mapping Differences in People’s Online Skills (No. arXiv:cs/0109068). arXiv. https://doi.org/10.48550/arXiv.cs/0109068 Hargittai, E. (2011). The Digital Reproduction of Inequality. 收入 The Inequality Reader (2 edtion). Routledge. Hollebeek, L. D., Kumar, V., Srivastava, R. K., Lim, W. M., & Urbonavicius, S. (2025). Guidelines for theory selection: The IMPACT framework. Psychology & Marketing, 42(11), 2789–2806. https://doi.org/10.1002/mar.70009 Horak, S., Klein, A., & Li, X. (2022). Can general and particularistic types of trust mix? Advancing the trust (dis-) continuity debate in a Chinese context. International Journal of Emerging Markets , 17 (5), 1273～1291. https://doi.org/10.1108/IJOEM-07-2020-0745 Howard, P. N. (2010). The Digital Origins of Dictatorship and Democracy: Information Technology and Political Islam . Oxford University Press. Jhang, F.-H. (2022). Bonding social capital and political trust: The mediating role of trust in public figures in Taiwan. The Social Science Journal , 59 (1), 71～84. https://doi.org/10.1016/j.soscij.2019.07.009 Kang, S., & Zhu, J. (2021). Do People Trust the Government More? Unpacking the Distinct Impacts of Anticorruption Policies on Political Trust. Political Research Quarterly , 74 (2), 434～449. https://doi.org/10.1177/1065912920912016 Katz, V. S., & Gonzalez, C. (2016). Toward Meaningful Connectivity: Using Multilevel Communication Research to Reframe Digital Inequality. Journal of Communication , 66 (2), 236～249. https://doi.org/10.1111/jcom.12214 King, G., Pan, J., & Roberts, M. E. (2013). How Censorship in China Allows Government Criticism but Silences Collective Expression. American Political Science Review , 107 (2), 326～343. https://doi.org/10.1017/S0003055413000014 Li, R., Xu, W., Nie, H., & Zhang, W. (2025). The Longitudinal Relationship Between Close Friendship and Subjective Well-Being: The Chain Mediation Model of Interpersonal Trust and Perceived Social Support. Behavioral Sciences , 15 (4), 480. https://doi.org/10.3390/bs15040480 Lim, W. M. (2026). Guidelines for theory development: Typologies for theoretical novelty and interestingness. Martínez, L. M., Estrada, D., & Prada, S. I. (2019). Mental health, interpersonal trust and subjective well-being in a high violence context. SSM - Population Health , 8 , 100423. https://doi.org/10.1016/j.ssmph.2019.100423 Mathy, R. M., & Cooper, A. (2003). The duration and frequency of Internet use in a nonclinical sample: Suicidality, behavioral problems, and treatment histories. Psychotherapy: Theory, Research, Practice, Training , 40 (1～2), 125～135. https://doi.org/10.1037/0033-3204.40.1-2.125 McCrae, N., Gettings, S., & Purssell, E. (2017). Social Media and Depressive Symptoms in Childhood and Adolescence: A Systematic Review. Adolescent Research Review , 2 (4), 315～330. https://doi.org/10.1007/s40894-017-0053-4 McLean, G., Krey, N., & Barhorst, J. B. (2025). Revealing the double‐edged sword: Introducing the Technology and Consumer Well‐being Paradox Model. Psychology & Marketing , 42 (1), 5～20. https://doi.org/10.1002/mar.22110 Miao, X. L. (2023). Interconnection and mutual trust: Online social capital and generalized trust in the digital era. Open Times, (6), 102-116+8. Norris, P. (2001). Making democracies work: social capital and civic engagement in 47 societies. Harvard University, John F. Kennedy School of Government Working Paper Series, (rwp01-036). Norris, P. (2011). Democratic Deficit: Critical Citizens Revisited . Cambridge University Press. https://doi.org/10.1017/CBO9780511973383 Portes, A. (2000). Social Capital: Its Origins and Applications in Modern. Knowledge and Social Capital: Foundations and Applications , 43. Putnam, R. D. (2000). Bowling alone: America’s declining social capital. In Culture and politics: A reader (pp. 223-234). New York: Palgrave Macmillan US. Putnam, R. D. (1966). Political Attitudes and the Local Community. American Political Science Review , 60 (3), 640～654. https://doi.org/10.2307/1952976 Ratan, R. A., Chung, J. E., Shen, C., Williams, D., & Poole, M. S. (2010). Schmoozing and Smiting: Trust, Social Institutions, and Communication Patterns in an MMOG. Journal of Computer-Mediated Communication , 16 (1), 93～114. https://doi.org/10.1111/j.1083-6101.2010.01534.x Rodríguez-Pose, A., & Maslauskaite, K. (2012). Can policy make us happier? Individual characteristics, socio-economic factors and life satisfaction in Central and Eastern Europe. Cambridge Journal of Regions, Economy and Society , 5 (1), 77～96. https://doi.org/10.1093/cjres/rsr038 Rotter, J. B. (1980). Interpersonal trust, trustworthiness, and gullibility. American Psychologist , 35 (1), 1～7. https://doi.org/10.1037/0003-066X.35.1.1 Seabrook, E. M., Kern, M. L., & Rickard, N. S. (2016). Social networking sites, depression, and anxiety: A systematic review. JMIR Mental Health, 3(4), e50. https://doi.org/10.2196/mental.5842 Shensa, A., Sidani, J. E., Escobar-Viera, C. G., Switzer, G. E., Primack, B. A., & Choukas-Bradley, S. (2020). Emotional support from social media and face-to-face relationships: Associations with depression risk among young adults. Journal of Affective Disorders , 260 , 38～44. https://doi.org/10.1016/j.jad.2019.08.092 Silchenko, K. (2025). Technology Dualities and Paradoxes in Digital Well-Being: Analysis of Discourses of Digital Well-Being Market Offerings. Journal of Macromarketing , 45 (4), 550～573. https://doi.org/10.1177/02761467251356660 Uslaner, E. M. (2002). The Moral Foundations of Trust (SSRN Scholarly Paper No. 824504). Social Science Research Network. https://doi.org/10.2139/ssrn.824504 Uslaner, E. M. (2004). Trust, Civic Engagement, and the Internet. Political Communication , 21 (2), 223～242. https://doi.org/10.1080/10584600490443895 Van Bruyssel, S., De Wolf, R., & Vanden Abeele, M. (2024). From bliss to burden: An ethnographic inquiry into how social, material and individual obstacles to digital well-being shape everyday life. New Media & Society , 14614448241288159. https://doi.org/10.1177/14614448241288159 van Deursen, A. J. A. M., & Helsper, E. J. (2015). The Third-Level Digital Divide: Who Benefits Most from Being Online? https://doi.org/10.1108/S2050-206020150000010002 Weber, M. (1968). Economy and society: An outline of interpretative sociology (G. Roth & C. Wittich, Eds.). University of California Press. (Original work published 1922) Yang, Y., Zeng, D., & Yang, F. (2022). Internet Use and Subjective Well-Being of the Elderly: An Analysis of the Mediating Effect Based on Social Capital. International Journal of Environmental Research and Public Health , 19 (19), 12087. https://doi.org/10.3390/ijerph191912087 Yap, S.-F., & Lim, W. M. (2024). A Paradox Theory of Social Media Consumption and Child Well-Being. Australasian Marketing Journal , 32 (1), 65～75. https://doi.org/10.1177/14413582221139492 Yoon, S. (2014). Does social capital affect SNS usage? A look at the roles of subjective well-being and social identity. Computers in Human Behavior , 41 , 295～303. https://doi.org/10.1016/j.chb.2014.09.043 You, Y., & Wang, Z. (2020). The Internet, political trust, and regime types: A cross-national and multilevel analysis. Japanese Journal of Political Science , 21 (2), 68～89. https://doi.org/10.1017/S1468109919000203 Zhan, S., Wang, Y., & Yang, Y. (2025). How Internet Use Affects Hierarchical Political Trust pattern?—Empirical analysis based on CSS2021 data. Public Administration and Policy Review , 14 (2), 87. Zhang, D., Shi, J., & Feng, Z. (2025). The dark side of the association between internet use and older adults’ subjective wellbeing: The role of subjective social class. BMC Geriatrics , 25 (1), 63. https://doi.org/10.1186/s12877-024-05606-w Zhang, C. M., & Li, Y. Q. (2023). Internet use, social trust, and residents' well-being: A study based on CFPS data. Rural Finance Research, (5), 70-80. Retrieved from https://www.cqvip.com/qk/93444X/202305/ Zhou, B. (2019). Fear of missing out, feeling of acceleration, and being permanently online: A survey study of university students’ use of mobile apps in China. Chinese Journal of Communication . https://www.tandfonline.com/doi/abs/10.1080/17544750.2018.1523803 Additional Declarations No competing interests reported. Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 08 May, 2026 Reviews received at journal 06 May, 2026 Reviewers agreed at journal 16 Apr, 2026 Reviews received at journal 16 Apr, 2026 Reviewers agreed at journal 14 Apr, 2026 Reviewers agreed at journal 03 Apr, 2026 Reviewers invited by journal 31 Mar, 2026 Editor invited by journal 30 Mar, 2026 Editor assigned by journal 27 Mar, 2026 Submission checks completed at journal 27 Mar, 2026 First submitted to journal 25 Mar, 2026 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-9225549","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":617288228,"identity":"b27d377e-a4ea-4675-a8d0-571d54b90489","order_by":0,"name":"Qing Wang","email":"","orcid":"","institution":"Fudan University","correspondingAuthor":false,"prefix":"","firstName":"Qing","middleName":"","lastName":"Wang","suffix":""},{"id":617288229,"identity":"03c5e132-6007-4364-a2db-f146b25be8b6","order_by":1,"name":"Xuebo Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA6UlEQVRIie3PsWrDMBCA4TMCeRHNqpBgv4JMoJOHPMqJgKYGOmoI1CFFHurQtUOgr9Cxo4vBWdR01ag8QrZ0a/cG29k66Jvv5+4AguAfounX0aPIkzR+PnnUq/7khjMivFazrKpnwtu2P0k4o2NvG1m4+e34+EgGHDbZ1lwaFa1fQGlZUBiVT9idTA8opMlJPClaJ9+nwO3nW3cCKPB3C412H8ZJS0HwZX9SS9MwcAt6Lw0ZkPC7rEDbcHCKwrCEtQtArURWWcLRtqz3l7Tc7L/PIn94javodNarZFRuu5M/2HXjQRAEwUU/gj5NVNlbWGMAAAAASUVORK5CYII=","orcid":"","institution":"South China Normal University","correspondingAuthor":true,"prefix":"","firstName":"Xuebo","middleName":"","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2026-03-25 16:26:18","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-9225549/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9225549/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":106238894,"identity":"193ae896-9266-40ed-b798-2bcecfe04b7f","added_by":"auto","created_at":"2026-04-06 14:30:49","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":67028,"visible":true,"origin":"","legend":"\u003cp\u003eTheoretical Model\u003c/p\u003e\n\u003cp\u003eNote: Control variables including age, gender, education, hukou registration, residence, and marriage are included in the analyses but omitted from this path diagram for visual clarity.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-9225549/v1/d7d12ea44de706864a4f2b53.png"},{"id":106238892,"identity":"e51be53f-27e2-48b4-99b6-d10867ab72fa","added_by":"auto","created_at":"2026-04-06 14:30:49","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":114355,"visible":true,"origin":"","legend":"\u003cp\u003eModerating effect of subjective class identity on the relationship between internet use and stranger trust\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-9225549/v1/a08a9693837d30200ce64fa3.png"},{"id":106238893,"identity":"dc8d1c04-ebfb-4ccc-98ce-b6ead9bc5bd3","added_by":"auto","created_at":"2026-04-06 14:30:49","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":85931,"visible":true,"origin":"","legend":"\u003cp\u003eActual model\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-9225549/v1/a472e6d5bfc1ed3c43c00fff.png"},{"id":106402710,"identity":"8eb12344-64b2-4cb9-8c42-532396870255","added_by":"auto","created_at":"2026-04-08 09:12:36","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1154482,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9225549/v1/9a99db0b-1eca-4d14-a626-da55ad83011d.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"The Conditional Effects of Internet Use on Well-being: Heterogeneous Trust Mechanisms and the Boundary of Subjective Social Status","fulltext":[{"header":"Introduction","content":"\u003cp\u003eThe pervasive integration of the internet into public life has sparked intense academic debate regarding its impact on subjective well-being (SWB). While the digital environment empowers connectivity, it concurrently risks alienating individuals, underscoring a persistent \u003cem\u003edigital well-being paradox\u003c/em\u003e (Diener et al., \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e1999\u003c/span\u003e; Diener et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2003\u003c/span\u003e; Yap \u0026amp; Lim, \u003cspan citationid=\"CR59\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; McLean et al., \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Silchenko, \u003cspan citationid=\"CR52\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This scholarly divergence suggests that the transition from digital connectivity to psychological well-being is not linear, but rather governed by complex, often contradictory, mediating mechanisms.\u003c/p\u003e \u003cp\u003eTo unravel this paradox, extant research frequently utilizes Social Capital Theory, positing that internet use fosters social ties and trust, subsequently enhancing SWB. However, a critical limitation is the tendency to treat trust as a monolithic construct. This unidimensional view overlooks the profound theoretical insight that trust is inherently heterogeneous. This paper addresses this underdeveloped issue, demonstrating that trust heterogeneity can actually flip the sign of internet use effects. Treating trust as a single mediator conceals both the benefits and the psychological costs of digital engagement, particularly within China\u0026rsquo;s unique differential trust structure (Fei, \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e1992\u003c/span\u003e; Li, \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eAccordingly, this study explores two core questions: (1) How does internet usage influence SWB through the distinct mediating pathways of stranger (bridging), family-neighbour (bonding), and political (institutional) trust? (2) How does subjective class identity condition these mechanisms?\u003c/p\u003e \u003cp\u003eThis study is theory-driven, with social capital theory serving as the focal lens for explaining how internet use shapes trust-based resources. Differential trust pattern theory complements this lens by clarifying why trust is not a single construct in China and why trust directed at close ties, generalized others, and institutions operate through distinct mechanisms. Social comparison theory further explains why subjective class identity conditions whether internet-enabled exposure translates into beneficial trust. Theory selection strictly follows the IMPACT logic (Hollebeek et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2025\u003c/span\u003e): interestingness lies in explaining why the same digital behaviour generates both social gain and psychological cost; matching aligns each trust dimension with its theoretical mechanism; parsimony limits the framework to the necessary trust pathways; applicability links findings to concrete well-being concerns; conceptual rigor specifies clear boundary conditions; and testability anchors each claim to panel-based empirical tests.\u003c/p\u003e \u003cp\u003eUtilizing four waves of longitudinal data from the China Family Panel Studies (2016\u0026ndash;2022) and employing a lagged fixed-effects approach, this research estimates cross-period associations while mitigating unobserved time-invariant confounding. Moving beyond the limitations of cross-sectional scale data, we introduce subjective class identity as a critical boundary condition.\u003c/p\u003e \u003cp\u003eGuided by the typologies of theory development (Lim, \u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e2026\u003c/span\u003e), this study contributes through theoretical modification and extension. First, we extend social capital theory by uncovering the hidden costs of digital connectivity. We critically demonstrate that internet-enhanced bonding trust (family and neighbours) acts as a psychological burden in a digitally mediated environment, highlighting the cultural specificity of strong ties. Second, we expand boundary conditions by demonstrating a structural paradox in digital bridging trust. While higher subjective class identity enables individuals to successfully convert internet use into stranger trust, this digitally acquired capital remains emotionally superficial, failing to deliver SWB gains across all class strata. Theoretical interestingness arises in revealing the illusion of digital social capital, as the internet reinforces structural inequalities in network building, yet these networks ultimately prove hollow for deep psychological fulfilment. Third, we contribute to political trust literature by clarifying the boundaries of digital social capital. We reveal that digitally formed interpersonal trust struggles to cross structural barriers to enhance institutional (political) trust, clarifying which trust pathways are actually available for policy leverage.\u003c/p\u003e"},{"header":"Literature Review","content":"\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eInternet Usage and Subjective Well-Being\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eAs internet technology increasingly permeates public life, its impact on subjective well-being (SWB) is highly complex and multifaceted. The internet offers significant benefits by enhancing political transparency, daily efficiency, and interpersonal connections (Becchetti et al., 2021; B\u0026uuml;chi et al., 2018; Yang et al., 2022; Yoon, 2014; Zhang \u0026amp; Li, 2023). It facilitates instantaneous, low-cost communication that transcends time and space, provides broad access to information, and fosters diverse community support, particularly for marginalized groups (DiMaggio et al., 2001; Cardozo, 2023).\u003c/p\u003e\n\u003cp\u003eHowever, these benefits are countered by documented risks. Excessive internet use is associated with fragmented attention, social comparison, and cyberbullying, which can persistently erode mental health and trigger anxiety and loneliness (McCrae et al., 2017; Shensa et al., 2020). For instance, internet-addicted groups show pronounced levels of depression and suicidal ideation (Seabrook et al., 2016).\u003c/p\u003e\n\u003cp\u003eWhile digital engagement serves as a valuable resource for connectivity, it simultaneously acts as a catalyst for social comparison and psychological burden (De Hesselle \u0026amp; Montag, 2024; Mathy \u0026amp; Cooper, 2003). Consequently, positing a strictly unidirectional effect on SWB is theoretically insufficient. Aligning with the digital well-being paradox, the direct effect of internet use on well-being may be deeply contested. Therefore, rather than assuming a uniform outcome, this study proposes the following competing hypotheses to capture the baseline relationship before introducing our trust-based mechanisms:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypothesis 1a\u003c/strong\u003e (The Enhancement Hypothesis): Internet usage is positively associated with residents\u0026apos; subjective well-being.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eHypothesis 1b\u003c/strong\u003e (The Burden Hypothesis): Internet usage is negatively associated with residents\u0026apos; subjective well-being.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eInternet Usage and Heterogeneous Interpersonal Trust: Bridging Gains and Bonding Costs\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe classic distinction between particularistic (bonding) trust and universalistic (bridging) trust is central to understanding Chinese society\u0026apos;s differential trust pattern (Fei, 1992). Universalistic trust extends to generalized others and non-kin (strangers), forming bridging social capital, while particularistic trust is based on strong kinship or geographical ties (family and neighbors), forming bonding social capital (Coleman, 1988; Portes, 2000; Putnam, 2000; Hall et al., 2021).\u003c/p\u003e\n\u003cp\u003eWhile traditional frameworks often assume the internet uniformly enhances interpersonal trust, differentiating these trust targets reveals a divergent mechanism. Regarding bridging trust, the prevailing view indicates a positive association. The internet functions as a low-cost, boundary-spanning tool that effectively reduces communication barriers, allowing individuals to connect with heterogeneous groups and accumulate stranger trust (Ratan et al., 2010). Because this expanded social network provides novel social support and fulfills psychological needs, it translates into higher subjective well-being (SWB) (Mart\u0026iacute;nez et al., 2019; Li et al., 2025).\u003c/p\u003e\n\u003cp\u003eTherefore, we hypothesize:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH2:\u003c/strong\u003e Internet usage positively influences stranger trust.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH3:\u0026nbsp;\u003c/strong\u003eInternet usage exerts a positive indirect effect on subjective well-being through the mediation of stranger trust.\u003c/p\u003e\n\u003cp\u003eConversely, the impact of internet use on bonding trust (family and neighbors) may paradoxically manifest as a hidden psychological cost. In the digital age, ubiquitous social platforms (such as WeChat in China) facilitate perpetual connectivity, which can continuously reinforce and intensify interactions within primary social circles (Davidow, 2011). While traditional social capital theory assumes strong ties are unequivocally beneficial, digital \u0026quot;over-connection\u0026quot; fundamentally alters this dynamic. The constant online presence blurs personal boundaries, subjecting individuals to inescapable social surveillance, rigid reciprocal norms, and the heavy emotional labor required to maintain these strong ties (Van Bruyssel et al., 2024; Zhou, 2019). Consequently, rather than providing liberating social support, this digitally intensified bonding trust transforms into a psychological burden (Rotter, 1980; Horak et al., 2022). This inescapable network of localized obligations and constant social presence ultimately diminishes overall life satisfaction. Thus, we propose:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH4:\u003c/strong\u003e Internet usage positively influences family and neighbourhood trust.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH5:\u0026nbsp;\u003c/strong\u003eInternet usage exerts a negative indirect effect on subjective well-being through the mediation of family and neighbourhood trust.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eInternet Usage, Political Trust, and Subjective Well-Being\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe widespread adoption of internet technology also exerts a complex influence on public political (institutional) trust. While the internet facilitates information flow, higher usage frequently exposes individuals to unverified critical discourse, governance flaws, or social inequalities, which can incrementally erode institutional confidence (You \u0026amp; Wang, 2020).\u003c/p\u003e\n\u003cp\u003eCross-national analyses consistently show that frequent internet use can widen the gap between heightened democratic expectations and perceived government performance, negatively impacting institutional trust (Howard, 2010; Norris, 2011).\u003c/p\u003e\n\u003cp\u003ePolitical trust is a well-documented macro-level driver of SWB. High levels of institutional trust reduce perceived societal uncertainty by strengthening expectations of institutional safeguards and life stability (C\u0026aacute;rcaba et al., 2022). It reflects public confidence in policy response mechanisms; when citizens believe their concerns are systematically addressed, their well-being significantly increases (Hamilton et al., 2016). Therefore, if internet usage compromises political trust, it subsequently removes a critical pillar of subjective well-being. Based on this mechanism, we propose:\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH6:\u0026nbsp;\u003c/strong\u003eInternet usage negatively influences political trust.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH7:\u0026nbsp;\u003c/strong\u003eInternet usage exerts a negative indirect effect on subjective well-being through the mediation of political trust.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eSubjective Class Identity as a Boundary Condition\u0026nbsp;\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eSubjective class identity refers to an individual\u0026apos;s self-perception and emotional identification with their position in the social hierarchy (Weber, 1968). This identity serves as a critical socio-psychological filter regulating this conversion (Hargittai, 2001).\u003c/p\u003e\n\u003cp\u003eIndividuals with a high subjective class identity generally possess greater psychological security, cognitive resources, and optimism (Uslaner, 2002). This enables them to use the internet strategically, approaching unfamiliar online environments with confidence to cultivate generalized trust and extract bridging resources (Miao, 2023). In contrast, those with lower subjective class identity often experience higher baseline vulnerability and resource anxieties (Zhang et al., 2025). For them, the digital environment may appear riskier, limiting their capacity or willingness to transform digital connectivity into genuine stranger trust (CAI \u0026amp; HOU, 2014; CHEN \u0026amp; YANG, 2025; Bagherianziarat \u0026amp; Hamplov\u0026aacute;, 2025) \u0026nbsp;Therefore, subjective class identity acts as a boundary condition, determining whether online exposure successfully translates into beneficial forms of generalized trust.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eH8:\u003c/strong\u003e Subjective class identity positively moderates the relationship between internet usage and stranger trust; specifically, the positive effect of internet use on stranger trust is stronger for individuals with higher subjective class identity.\u003c/p\u003e\n\u003cp\u003eTherefore, we construct the following theoretical model (see Figure 1 ):\u003c/p\u003e"},{"header":"Method","content":"\u003cp\u003eThe data for our study are sourced from the China Family Panel Studies (CFPS), a nationwide longitudinal fixed-sample tracking survey conducted by the Institute of Social Science Surveys (ISSS) at Peking University. This project systematically collects microdata at the individual, household, and community levels to monitor and analyse dynamic changes across multiple dimensions of Chinese society. Its findings primarily provide empirical evidence for academic discourse and public policy formulation. Specific survey topics encompass employment status, educational attainment, family structure and interactions, population mobility, and physical and mental health conditions. To maintain panel data continuity and variable consistency, exclusion method was applied to address missing values. The resulting valid sample comprises 14,328 observations from 3,582 residents tracked longitudinally across four waves of data: 2016, 2018, 2020, and 2022.\u003c/p\u003e \u003cp\u003eSubjective well-being is the dependent variable in our study. The CFPS data poses the question, \"To what extent are you satisfied with your life as a whole?\" The scale ranges from 1 to 5, with 1 representing 'very dissatisfied' and 5 representing 'very satisfied'. Responses are scored on a scale of 1 to 5, with 1 representing the weakest response and 5 representing the strongest.\u003c/p\u003e \u003cp\u003eInternet usage is the core independent variable in our research. In the context of surveying internet usage, the CFPS data typically poses the following questions: \"Do you use mobile internet?\" and \"Do you use computer internet?\" The values assigned to the responses were as follows: 1 for \"yes\" responses and 0 for \"no\" responses. Furthermore, the CFPS data inquired about the frequency of public consumption, daily activities, and shopping on internet platforms. The following scale was utilized: 1\u0026thinsp;=\u0026thinsp;Almost daily, 2\u0026thinsp;=\u0026thinsp;3\u0026ndash;4 times per week, 3\u0026thinsp;=\u0026thinsp;1\u0026ndash;2 times per week, 4\u0026thinsp;=\u0026thinsp;2\u0026ndash;3 times per month, 5\u0026thinsp;=\u0026thinsp;Once per month, 6\u0026thinsp;=\u0026thinsp;Once every few months, 7\u0026thinsp;=\u0026thinsp;Never. The integration of these scores with the initial two questions resulted in the formation of a novel variable pertaining to internet usage (Blank \u0026amp; Groselj, \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2014\u003c/span\u003e). To ensure that higher numerical values represent a higher frequency of digital engagement, the original frequency scales (1\u0026ndash;7) were reverse-coded prior to aggregation.\u003c/p\u003e \u003cp\u003eIn our study, interpersonal trust and political trust function as mediating variables. Interpersonal trust is further categorized into two distinct types: stranger trust and trust in close relatives and neighbours. The CFPS data on stranger trust includes questions such as \"How much do you trust Individuals from other countries?\", \"How much do you trust strangers?\"; while questions on trust in close relatives and neighbours include \"How much do you trust your parents?\" and \"How much do you trust your neighbours?\" The CFPS data on political trust primarily comprises questions regarding the extent of trust placed in local government officials. The scoring scale employed for all trust surveys ranges from 0 to 10, where 0 denotes a high level of distrust and 10 denotes a high level of trust.\u003c/p\u003e \u003cp\u003eThe concept of subjective class identity is proposed as the moderating variable in our study. Subjective class identity was measured by asking respondents to rate their social status within their local community on a scale from 1 (very low) to 5 (very high). Furthermore, the study's control variables encompass gender (1\u0026thinsp;=\u0026thinsp;male, 0\u0026thinsp;=\u0026thinsp;female), age, personal income (logarithmically transformed), educational attainment (measured on an 8-point scale ranging from 1\u0026thinsp;=\u0026thinsp;illiterate/semi-illiterate to 8\u0026thinsp;=\u0026thinsp;doctoral degree), hukou registration (1\u0026thinsp;=\u0026thinsp;non-agricultural, 0\u0026thinsp;=\u0026thinsp;agricultural), residence (1\u0026thinsp;=\u0026thinsp;urban, 0\u0026thinsp;=\u0026thinsp;rural), and marital status (1\u0026thinsp;=\u0026thinsp;married, 0\u0026thinsp;=\u0026thinsp;unmarried), which are standard demographic controls identified in previous studies (Graham \u0026amp; Chattopadhyay, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Rodr\u0026iacute;guez-Pose \u0026amp; Maslauskaite, \u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2012\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo rigorously test the cross-period causal effect and address potential endogeneity, we employ the Lagged Fixed Effects (LFE) model. The LFE model incorporates the dependent variable lagged by one period (Yt-1) alongside the individual fixed effects (ui). This approach offers a significant advantage by simultaneously controlling for unobserved time-invariant individual heterogeneity and crucial dynamic endogeneity, which arises when the outcome variable influences the predictor in the subsequent period. Therefore, the LFE model provides a more robust and conservative causal estimate than simple OLS or standard Fixed Effects models when analysing the dynamic relationship between internet use and subjective well-being over time. Furthermore, to enhance statistical power and ensure the reliability of our findings, we utilized the Bootstrap method with 5,000 resamples to calculate the 95% confidence intervals for all indirect effects in the mediation and moderated mediation analyses.\u003c/p\u003e"},{"header":"Results","content":"\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eDescriptive Statistics\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eThe statistical characteristics of the variables suggest several notable patterns in the sample distribution (see Table 1). Internet usage has a mean score of 1.69, with its positive skewness (0.554) indicating a relatively dispersed usage pattern that leans slightly toward the lower end of the 1-to-3 scale. The mean score for subjective class identity is 0.80; its strong negative skewness (-1.683) across a wide range (-8.7 to 2.5) implies that while the average sits lower numerically, a heavy concentration of respondents identify with the higher end of their perceived class strata, offset by a long tail of individuals reporting significantly lower subjective status.\u003c/p\u003e\n\u003cp\u003eTable 1 Descriptive Statistics for Each Variable\u003c/p\u003e\n\u003ctable style=\"border: none; width: 100%;\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eVariable\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eN\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eAverage\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eStd. Dev.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eSkewness\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMin\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMax\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eSubjective class identity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.80\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.96\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-1.683\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-8.7\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2.5\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eFamily and Neighborhood trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e13.60\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5.47\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-1.395\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e19.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eStranger trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e7.48\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4.39\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.035\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e19.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003ePolitical trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5.38\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2.64\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.185\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e10.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eSubjective well-being\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e3.90\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.99\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.658\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e5.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eInternet Use\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.73\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.55\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e3.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eAge\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e51.92\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e13.62\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.05\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e93.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eGender\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.56\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.24\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eEducation\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e2.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.69\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.84\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e8.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eHukou registration\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.23\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.42\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.27\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eResidence\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.50\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.01\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eMarriage\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.87\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.34\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-2.17\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.0\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eln Personal Income\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e14328\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.63\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.02\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e6.54\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003eRegarding the trust measures, family and neighborhood trust (mean 13.60) is highly concentrated toward the top of the 19-point scale, supported by a strong negative skew (-1.395). This suggests that a majority of respondents tend to report high levels of trust within their close social circles. In contrast, stranger trust (mean 7.48) exhibits a more dispersed and highly symmetrical distribution (skewness 0.035), which reflects considerable variation and general caution in public attitudes toward trusting strangers. Political trust (mean 5.38) falls squarely within the middle range, with a slight negative skew (-0.185) showing a mild tendency toward institutional recognition. Furthermore, subjective well-being has a high mean of 3.90 on a 5-point scale, confirming that the overall reported happiness level in the surveyed group is generally elevated.\u003c/p\u003e\n\u003cp\u003eTable 2 Pearson Correlation Matrix of Key Study Variables\u003c/p\u003e\n\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\" width=\"558\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003eVariable\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003eFamily/Neighbor Trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eStranger Trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003ePolitical Trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eSubjective Class\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003eWell-being\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003eInternet Use\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003eFamily/Neighbor Trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003eStranger Trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.138\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003ePolitical Trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e-0.040\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e0.230\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003eSubjective Class\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.135\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e0.105\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.049\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003eWell-being\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e-0.089\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e0.031\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e0.237\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.028\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\u003cbr\u003e\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd style=\"width: 106px;\"\u003e\n \u003cp\u003eInternet Use\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 92px;\"\u003e\n \u003cp\u003e0.137\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e0.111\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.043\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e0.409\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 76px;\"\u003e\n \u003cp\u003e-0.073\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd style=\"width: 57px;\"\u003e\n \u003cp\u003e1.000\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e*p\u0026lt;0.05, **p\u0026lt;0.01, ***p\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003eTable 2 presents the Pearson correlation coefficient matrix, revealing the complex structural associations among the core variables and laying the empirical foundation for subsequent mechanism analyses (see Table 2). Among the results, the most salient relationship exists between subjective class identity and internet use, which exhibit a strong positive correlation (r = 0.409, p \u0026lt; 0.001). This confirms that a higher perceived social stratum is a crucial precondition or correlate for individual digital engagement.\u003c/p\u003e\n\u003cp\u003eRegarding the primary variable of interest, internet use demonstrates a heterogeneous association with different dimensions of trust. It is significantly and positively correlated with both family/neighbor trust (r = 0.137, p \u0026lt; 0.001) and stranger trust (r = 0.111, p \u0026lt; 0.001), yet it shows a negative correlation with political trust (r = -0.043, p \u0026lt; 0.001). This suggests that while internet usage may expand horizontal social networks and generalized trust, it might subtly erode vertical institutional trust.\u003c/p\u003e\n\u003cp\u003eFurthermore, subjective well-being exhibits a highly differentiated pattern when associated with these trust dimensions. It is most strongly and positively correlated with political trust (r=0.237,p\u0026lt;0.001), indicating that institutional confidence is a vital psychological anchor for individual well-being. Conversely, well-being is significantly negatively correlated with family and neighborhood trust (r=-0.089,p\u0026lt;0.001), preliminarily hinting at the potential emotional burden or eroding effect of traditional acquaintance-based social ties. Notably, internet use itself is significantly negatively correlated with subjective well-being (r=-0.073,p\u0026lt;0.001), providing initial correlational support for the \u0026quot;well-being paradox\u0026quot; in the digital age.\u003c/p\u003e\n\u003cp\u003eCollectively, these results demonstrate that the nexus among internet use, multi-dimensional trust, and well-being is intricately interwoven, characterized by a complex structure of coexisting positive network expansions and negative psychological impacts.\u003c/p\u003e\n\u003ch2\u003e\u003cstrong\u003e\u003cem\u003eLagged Fixed Effects Model Statistics\u003c/em\u003e\u003c/strong\u003e\u003c/h2\u003e\n\u003cp\u003eOur analysis aims to delve into complex cross-period causal effects and initially confirmed that core variables such as subjective well-being and social trust exhibit significant temporal persistence. To precisely capture this dynamic adjustment characteristic while strictly controlling for all time-invariant individual unobserved heterogeneity, we constructed a Lagged Fixed Effects model (LFE). Furthermore, to rigorously isolate the effects of interest, the model incorporates the 2016 baseline levels of all respective variables, alongside a comprehensive set of demographic controls including age, gender, education, hukou registration, place of residence, and marital status\u003c/p\u003e\n\u003cp\u003eBy utilizing the LFE model, we rigorously eliminated biases caused by all time-invariant individual characteristics, and the multi-period lagged structure (X2018\u0026rarr;M2020\u0026rarr;Y2022) ensures the temporal precedence required for causal inference. Therefore, employing this rigorous Lagged Fixed Effects approach, we proceeded with an in-depth empirical analysis of the net, cross-period causal effects of internet usage and various forms of trust on subjective well-being. H1b and H2 are supported, while H1a, H4, and H6 are not supported.\u003c/p\u003e\n\u003cp\u003eTable 3 Lagged Effects of Internet Use on Trust and Well-being\u003c/p\u003e\n\u003ctable style=\"width: 4.4e+2pt;border: none;\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eCoeff.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eStd. Error\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003et\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e95% CI [Lower, Upper]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eInternet Use (2018)\u0026rarr;Family and Neighborhood trust (2020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.178\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.131\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e1.351\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.177\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[-0.080, 0.435]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.051\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eInternet Use (2018)\u0026rarr;Stranger trust (2020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.460\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.108\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e4.272\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[0.249, 0.672]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eInternet Use (2018)\u0026rarr;Political trust (2020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.058\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.339\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.735\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[-0.134, 0.094]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.113\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eInternet Use (2018)\u0026rarr;Subjective well-being (2022)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.043\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.022\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-1.966\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.049\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[-0.085, -0.001]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.071\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e*p\u0026lt;0.05, **p\u0026lt;0.01, ***p\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003eNote: Higher values for internet use indicate higher frequency due to reverse-coding.\u003c/p\u003e\n\u003cp\u003eThe lagged analysis examines the longitudinal impact of Internet use (measured in 2018) on different dimensions of social trust (measured in 2020) and subjective well-being (measured in 2022). The results reveal divergent cross-period effects of Internet engagement on individuals\u0026apos; social attitudes and psychological outcomes (see Table 3).\u003c/p\u003e\n\u003cp\u003eFirst, the model demonstrates a highly significant and robust positive lagged effect of prior Internet use on stranger trust (Coeff. = 0.460, p \u0026lt; 0.001). This suggests that increased internet usage significantly fosters a broader generalized trust toward strangers over time.\u003c/p\u003e\n\u003cp\u003eConversely, the analysis reveals a statistically significant, albeit small, negative longitudinal impact on subjective well-being (Coeff. = -0.043, p \u0026lt; 0.05). This indicates that higher levels of Internet use in 2018 are associated with a slight decline in individuals\u0026apos; reported subjective well-being four years later in 2022.\u003c/p\u003e\n\u003cp\u003eIn contrast, earlier Internet use does not exert any significant delayed influence on localized or institutional trust. Specifically, the lagged effects on family and neighbourhood trust (Coeff. = 0.178, p = 0.177) and political trust (Coeff. = -0.020, p = 0.735) are statistically insignificant, with both of their 95% confidence intervals crossing zero. This implies that while Internet use reshapes attitudes toward strangers and personal well-being, it does not meaningfully alter individuals\u0026apos; baseline trust in their immediate social circles or the political system over time.\u003c/p\u003e\n\u003cp\u003eTable 4 Lagged Effects of Social Trust Dimensions on Well-being\u003c/p\u003e\n\u003ctable style=\"width: 4.2e+2pt;border: none;\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eDependent Variable (2020)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eCoeff.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eStd. Error\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003et\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e95% CI [Lower, Upper]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eR\u003csup\u003e2\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eFamily \u0026amp; Neighbor(2020) Trust\u0026rarr; Subjective well-being (2022)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.006\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-2.230\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.026\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[-0.012, -0.001]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.085\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eStranger Trust (2020) \u0026rarr; Subjective well-being (2022)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-1.602\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.109\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[-0.012, 0.001]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.085\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003ePolitical Trust (2020) \u0026rarr; Subjective well-being (2022)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.046\u003csup\u003e***\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.006\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e7.223\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e[0.034, 0.059]\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.085\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e*p\u0026lt;0.05, **p\u0026lt;0.01, ***p\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003eTable 4 presents the findings from the lagged analysis examining how different dimensions of social trust in 2020 impact subjective well-being in 2022 (see Table 4). The results reveal that various forms of trust exert divergent longitudinal effects on an individual\u0026apos;s future psychological outcomes.\u003c/p\u003e\n\u003cp\u003eMost notably, political trust demonstrates a highly significant and robust positive lagged effect on subsequent subjective well-being (Coeff. = 0.046, p \u0026lt; 0.001). This indicates that individuals who reported higher levels of trust in political institutions in 2020 experienced significantly higher subjective well-being two years later.\u003c/p\u003e\n\u003cp\u003eConversely, family and neighbor trust exhibits a slight but statistically significant negative longitudinal impact on future well-being (Coeff. = -0.006, p = 0.026). This suggests that higher levels of trust concentrated within localized, primary social ties are associated with a marginal decrease in reported well-being over time.\u003c/p\u003e\n\u003cp\u003eFinally, the lagged effect of stranger trust on subjective well-being is not statistically significant (Coeff. = -0.006, p = 0.109). Because the 95% confidence interval crosses zero, there is no conclusive evidence in this model to suggest that generalized trust toward strangers has a delayed, long-term impact on personal well-being.\u003c/p\u003e\n\u003cp\u003eOverall, the model (which explains 8.5% of the variance, R\u003csup\u003e2\u003c/sup\u003e = 0.085) underscores that social trust is not monolithic in its psychological outcomes. Institutional (political) trust serves as a strong positive predictor for long-term well-being, whereas localized (family and neighbor) trust shows a minor negative association, and generalized (stranger) trust shows no significant temporal effect.\u003c/p\u003e\n\u003cp\u003eTable 5 Mediation Analysis of the Effect of Internet Use on Subjective Well-being\u003c/p\u003e\n\u003ctable style=\"width: 100%;border: none;\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eEffect Pathway\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eCoeff.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eS.E.\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e95% CI (Lower, Upper)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003ep\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eIndirect Effect (via Family \u0026amp; Neighbor Trust)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.005\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e(-0.011,-0.000)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.038\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eIndirect Effect (via Stranger Trust)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e(-0.008,0.007)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.838\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eIndirect Effect (via Political Trust)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e-0.002\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.003\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e(-0.009,0.005)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.602\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e\n\u003cp\u003e*p\u0026lt;0.05, **p\u0026lt;0.01, ***p\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003eTable 5 presents the Bootstrap mediation analysis evaluating the indirect pathways through which Internet use affects subjective well-being via three distinct dimensions of social trust (see Table 5). Among the examined mediators, only family and neighbour trust exhibits a statistically significant indirect effect (Coeff. = -0.005, p = 0.038). Although the direct lagged effect of internet use on family trust was statistically marginal (as shown in Table 3), the rigorous Bootstrap mediation test detected a significant, albeit weak, indirect pathway. In contrast, the indirect pathways through stranger trust (Coeff. = -0.001, p = 0.838) and political trust (Coeff. = -0.002, p = 0.602) are not statistically significant, as both of their 95% confidence intervals cross zero. H5 is supported, while H3 and H7 are not supported.\u003c/p\u003e\n\u003cp\u003eThese findings indicate that a portion of the adverse effect of internet use on well-being is indirectly channelled through its association with localized family and neighbourhood trust, whereas generalized and institutional trust do not serve as significant mediating mechanisms in this specific relationship.\u003c/p\u003e\n\u003cp\u003eTable 6 Moderating Effect of Subjective Class Identity on Stranger Trust\u003c/p\u003e\n\u003cdiv align=\"center\"\u003e\n \u003ctable style=\"width: 100%;border: none;\"\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eMediator (M)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eModerator (Subjective Class Identity) Level\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003eEst.\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003e95% CI\u003c/strong\u003e\u003c/p\u003e\n \u003cp\u003e\u003cstrong\u003e(Lower, Upper)\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e\u003cstrong\u003ep\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd rowspan=\"2\"\u003e\n \u003cp\u003eStranger trust\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003eLow Subjective Class Identity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e(-0.003,0.014)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.185\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd\u003e\n \u003cp\u003eHigh Subjective Class Identity\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.011\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e(0.004,0.019)\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd\u003e\n \u003cp\u003e0.000\u003csup\u003e*\u003c/sup\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e*p\u0026lt;0.05, **p\u0026lt;0.01, ***p\u0026lt;0.001\u003c/p\u003e\n\u003cp\u003eFurther analysis demonstrates that the internet\u0026apos;s capacity to foster stranger trust is subject to significant boundary conditions (see Table 6 \u0026amp; Figure 2 ). For individuals with a low subjective class identity, internet use fails to significantly foster stranger trust (Est. = 0.005, p = 0.185). However, as an individual\u0026apos;s subjective class identity level increases, this effect becomes highly significant (Est. = 0.011, p \u0026lt; 0.001), fully supporting H8. This demonstrates that the internet\u0026apos;s potential to expand bridging social capital is selectively unlocked by a higher perceived social standing.\u003c/p\u003e\n\u003cp\u003eCrucially, however, as indicated by the non-significant mediation pathway in Table 5, this unequally distributed bridging capital ultimately hits a dead end. Even for higher-status groups who successfully accumulate digital stranger trust, these weak ties lack the substantive emotional support necessary to effectively translate into subjective well-being.\u003c/p\u003e\n\u003cp\u003eIn conclusion, these findings confirm that an individual\u0026apos;s subjective class level serves as a critical boundary condition. It fundamentally determines whether internet use can effectively promote the conversion of positive, generalized social capital and ultimately impact an individual\u0026apos;s long-term subjective well-being.\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eOur study rigorously examined the long-term and dynamic effects of internet use on subjective well-being (see Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e for the specific action pathway). Our findings move beyond the simple \"foster or erode\" debate by confirming that the effect of Internet on well-being is highly conditional and realized through heterogeneous pathways of social trust. By precisely revealing these distinct mechanisms and boundary conditions, our results redefine the Chinese pathway of social capital in the digital society and offer critical insights for social governance.\u003c/p\u003e \u003cp\u003eTaken together, our research offers a sophisticated, multi-mechanism integrative explanatory framework for the long-term Internet and Well-being Paradox. The overall effect of the internet on well-being is negative. We also found the negative mechanism of the cost of family and neighbourhood trust erodes the well-being of all individuals. In the moderated mediation model, the internet's potential to foster generalized trust among strangers is only accessible to those with a high subjective class identity. This theoretical framework moves beyond a simple benefits-versus-harms debate, emphasizing the critical role of social structure and subjective stratification in the conversion of digital capital, and ultimately reveals the structural asymmetry of digital technology's impact on well-being.\u003c/p\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eAdvancing Hierarchical Political Trust Theory\u003c/h2\u003e \u003cp\u003eOur study highlights potential boundary conditions regarding the conversion of interpersonal trust into political trust in the digital age, thereby deepening the understanding of the hierarchical government trust pattern specific to the Chinese context. Existing theories have focused on the potential for a bottom-up conversion of micro-level interpersonal trust to macro-level institutional trust. For instance, Putnam suggests that high levels of trust among social members can be converted into support for the political system (Putnam, \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e1966\u003c/span\u003e; Norris, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2001\u003c/span\u003e). However, within China's singular social and digital ecology, the direct impact of the internet on political trust remains an area demanding theoretical discussion.\u003c/p\u003e \u003cp\u003eOur findings extend theoretical assumptions about the simple spill over of social capital across public and private spheres, fostering deep dialogue with Lianjiang Li's work on hierarchical government trust (L Li,2012). Results show internet use fails to significantly enhance this form of trust. Our longitudinal model finds no significant evidence that general internet use enhances this macro-level institutional trust. This null finding suggests that even if the internet cultivates generalized trust, social capital formed through atomized digital connections may not naturally spill over to cross public-private boundaries and influence public authority (Uslaner, \u003cspan citationid=\"CR54\" class=\"CitationRef\"\u003e2004\u003c/span\u003e; King et al., \u003cspan citationid=\"CR33\" class=\"CitationRef\"\u003e2013\u003c/span\u003e; Bolsover, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Kang \u0026amp; Zhu, \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2021\u003c/span\u003e; Jhang, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Zhan et al., \u003cspan citationid=\"CR62\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). This prompts reflection that within China's digital governance landscape, effectively activating and guiding the positive effects of social capital to better serve macro‑governance objectives remains an important ongoing topic.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eExtending Social Capital Theory\u003c/h2\u003e \u003cp\u003eOn the dimension of strong ties, we confirmed the hidden costs of bonding social capital, thus extending the understanding of traditional social capital theory. Family and neighbourhood trust are typically regarded as classic examples of bonding social capital that support individuals' emotional needs and resource acquisition(D. S. Brisson \u0026amp; Usher, \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2005\u003c/span\u003e; D. Brisson, \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2009\u003c/span\u003e). While classical social capital theory consistently promotes the positive role of strong ties, whether these ties incur hidden costs in the digital age, remains a theoretical issue awaiting resolution. Moreover, previous studies are not enough to explain whether such costs are amplified within China's unique differential pattern of association.\u003c/p\u003e \u003cp\u003eWhile the direct lagged effect of internet use on family trust appears statistically subtle in the aggregate model, our rigorous mediation analysis detects a significant, albeit weak, indirect pathway. This provides preliminary evidence suggesting that family and neighbourhood trust may possess the potential to act as a negative mechanism in the digital well-being paradox. It points to the possibility that, rather than unequivocally enhancing well-being, digitally intensified bonding social capital runs the risk of being alienated into a psychological burden. In China's internet ecosystem, where social software like WeChat are highly prevalent, the state of being always online may lead to over-connection (Davidow, 2011; Zhou, \u003cspan citationid=\"CR65\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). High levels of trust in a digital context can potentially translate into a heavy burden of human relations, the pressure of constant social presence, and intensified social norm constraints (Gargiulo \u0026amp; Ertug, \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2006\u003c/span\u003e; Van Bruyssel et al., \u003cspan citationid=\"CR55\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Therefore, by utilizing longitudinal data, our study reveals that bonding social capital may have become a negative mechanism in the paradox of digital age well-being, thereby supplementing the literature on the dark side of social capital.\u003c/p\u003e \u003cp\u003eAlthough internet use successfully fosters stranger trust, this form of bridging social capital does not significantly translate into improved long-term subjective well-being. Despite the internet\u0026rsquo;s capacity to generate weak ties, the lack of substantial support and emotional depth in such capital prevents the mediation mechanism from being confirmed. This overall pattern does not represent a simple failure of mediation. Instead, our moderation hypothesis (H8) reveals a deep structural paradox. The transformation of internet use into bridging trust is restricted by individuals\u0026rsquo; subjective social status and benefits only those with higher perceived class standing. Even so, such resource accumulation does not extend to improvements in well-being. The bridging social capital formed online operates only as a structural resource rather than an emotional resource. It fails to generate lasting improvements in subjective well-being, regardless of class identity.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eExpanding Social Comparison Theory\u003c/h2\u003e \u003cp\u003eOur research focuses on the potential of bridging social capital to promote well-being in the digital age, and introduces subjective class identity as a boundary condition for this analysis. The traditional technology empowerment hypothesis tends to suggest that digital tools enhance social capital (Choi et al., \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). However, this hypothesis did not adequately explain how social stratification limits individuals' ability to convert digital resources into positive well-being outcomes(van Deursen \u0026amp; Helsper, \u003cspan citationid=\"CR56\" class=\"CitationRef\"\u003e2015\u003c/span\u003e).\u003c/p\u003e \u003cp\u003eTo answer this question, our study finds that the different levels of subjective class identity interact with internet use on well-being, which engages in a profound dialogue with social comparison theory. Our finding addresses the debate of whether digital connectivity is a true social equalizer or merely a re-manifestation of social inequality on internet usage (Hargittai, \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2011\u003c/span\u003e; Katz \u0026amp; Gonzalez, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). We argue that subjective class identity acts as a filter for social comparison. For those with high class identity, they tend to view the internet as a tool for acquiring heterogeneous information and expanding weak ties. Conversely, individuals with low class identity may be more susceptible to feelings of relative deprivation when encountering idealized lives presented online. Furthermore, while our broad measurement of internet use does not capture specific online activities, these statistical disparities can be contextualized within the structural reality of China\u0026rsquo;s digital economy. In this broader context, researchers have increasingly noted that the 'stranger connections' accessible to lower-status groups are often precarious weak ties shaped by algorithmic exploitation (e.g., the digital gig economy).\u003c/p\u003e \u003cp\u003eThis dual disadvantage of psychological deprivation and the inferior structural quality of their digital ties ultimately inhibits their capacity to convert online interactions into beneficial generalized trust. Therefore, our study emphasizes that subjective class identity significantly influences the conversion efficiency of digital capital, providing an important perspective for understanding social inequality in the digital age.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003ePractical Implications\u003c/h2\u003e \u003cp\u003eThe Lagged Fixed Effects model applied to our four-year panel data establishes robust causal identification, providing a solid foundation for targeted interventions. Enhancing national well-being requires synergistic, actor-specific efforts targeting distinct empirical mechanisms:\u003c/p\u003e \u003cp\u003eFirst, directly addressing the negative mediating path of family and neighbourhood trust, digital platform developers must mitigate the psychological burden of over-connectivity. Developers (e.g., WeChat) should design architectural boundaries by incorporating features such as \"off-duty\" modes and granular notification controls. These technical adjustments empower users to manage heavy reciprocal obligations and limit constant social surveillance.\u003c/p\u003e \u003cp\u003eSecond, recognizing that digital bridging capital is structurally unequal yet emotionally insufficient for well-being, interventions must move beyond mere connectivity. While local governments should provide capacity-building to help lower-status groups safely navigate online networks to reduce resource stratification, society as a whole must actively rebuild offline, physically situated community engagements. Relying solely on the internet to foster generalized trust risks trapping individuals in emotionally hollow digital networks.\u003c/p\u003e \u003cp\u003eThird, responding to the persistent barrier preventing interpersonal ties from converting into institutional trust, policymakers must actively optimize digital governance. Since political trust is the strongest long-term predictor of well-being, governments cannot rely on natural social capital spillover. Policymakers should establish highly responsive, trackable online public feedback mechanisms to ensure citizens' digital grievances receive substantive institutional responses, thereby consolidating this structural cornerstone of well-being.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eLimitation and Future Directions\u003c/h2\u003e \u003cp\u003eOur study confirms important empirical connections, yet the digital sphere invites further methodological and theoretical development. Subsequent research could improve the granularity of measurement by using fine-grained indicators of online purpose to pinpoint effective activities for fostering social trust. Future studies may also benefit from utilizing sophisticated serial mediation models incorporating psychological variables to fully reveal the intricate social comparison effects behind subjective class identity's moderating role. Finally, comparative analyses involving countries with differing social capital structures and governance regimes would be valuable to assess the external validity and boundary conditions of our findings, which are currently grounded in the unique socio-political context of China.\u003c/p\u003e \u003c/div\u003e"},{"header":"Conclusion","content":"\u003cp\u003eOur study confirms a conditional and heterogeneous dual-mechanism framework that resolves the digital age well-being paradox. The research reveals that the internet's influence is structurally asymmetric: it generates a negative path by amplifying the hidden psychological costs of family and neighbourhood trust. These findings advance social capital theory and social comparison theory by defining the costs and boundary conditions of digital capital conversion, ultimately providing a rigorous, structural explanation for the unequal distribution of digital benefits in the Chinese context.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eQing Wang (WQ) and Xuebo Zhang (ZXB) conceived and designed the study. WQ performed the data extraction, conducted the statistical analyses (including the Lagged Fixed Effects modeling), and drafted the initial manuscript. ZXB supervised the entire research process, provided critical methodological guidance, and extensively revised the manuscript for important intellectual content. Both authors read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthical Considerations\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent for publication:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNot applicable\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eDeclaration of conflicting interest:\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eNo potential conflict of interest was reported by the authors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding statement:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis research received no specific grant from any funding agency in the public, commercial, or not-for-profit sectors.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability:\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets analyzed during the current study, derived from the China Family Panel Studies (CFPS), are available at the following URL: https://docs.google.com/spreadsheets/d/1eJx-wyAqlR11zrJS3qgZiL2mmteI2Ehp/edit?usp=sharing\u0026amp;ouid=107616459617005785709\u0026amp;rtpof=true\u0026amp;sd=true. The original raw data are publicly available from the Institute of Social Science Survey (ISSS) at Peking University.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBagherianziarat, A., \u0026amp; Hamplov\u0026aacute;, D. (2025). Life Satisfaction and Subjective/Objective Socioeconomic Status in European Countries: Does Affluence Matter? \u003cem\u003eFudan Journal of the Humanities and Social Sciences\u003c/em\u003e. https://doi.org/10.1007/s40647-025-00449-0\u003c/li\u003e\n\u003cli\u003eBecchetti, L., Florio, E., \u0026amp; Mancini, S. (2021). Internet Exposure and Social Capital. \u003cem\u003eSSRN Electronic Journal\u003c/em\u003e. https://doi.org/10.2139/ssrn.3923612\u003c/li\u003e\n\u003cli\u003eBlank, G., \u0026amp; Groselj, D. (2014). Dimensions of Internet use: Amount, variety, and types. \u003cem\u003eInformation, Communication \u0026amp; Society\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(4), 417～435. https://doi.org/10.1080/1369118X.2014.889189\u003c/li\u003e\n\u003cli\u003eBolsover, G. (2017). Harmonious communitarianism or a rational public sphere: A content analysis of the differences between comments on news stories on Weibo and Facebook. \u003cem\u003eAsian Journal of Communication\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(2), 115～133. https://doi.org/10.1080/01292986.2016.1214743\u003c/li\u003e\n\u003cli\u003eBrisson, D. (2009). Testing the Relationship of Formal Bonding, Informal Bonding, and Formal Bridging Social Capital on Key Outcomes for Families in Low-Income Neighborhoods. \u003cem\u003eJournal of Sociology \u0026amp; Social Welfare\u003c/em\u003e, \u003cem\u003e36\u003c/em\u003e, 167.\u003c/li\u003e\n\u003cli\u003eBrisson, D. S., \u0026amp; Usher, C. L. (2005). Bonding social capital in low‐income neighborhoods. Family Relations, 54(5), 644-653.\u003c/li\u003e\n\u003cli\u003eB\u0026uuml;chi, M., Festic, N., \u0026amp; Latzer, M. (2018). How social well-being is affected by digital inequalities. International Journal of Communication, 12, 21.\u003c/li\u003e\n\u003cli\u003eLi, L. (2012). Differential government trust (Hierarchical trust in government). Twenty-First Century, (131), 108\u0026ndash;114.\u003c/li\u003e\n\u003cli\u003eCai, H., \u0026amp; Hou, J. (2014). Subjective class identity and social trust: An empirical analysis based on CGSS 2010 data. \u003cem\u003eSociological Studies\u003c/em\u003e, (6), 20-42. https://doi.org/10.19934/j.cnki.issn1002-5936.2014.06.002.\u003c/li\u003e\n\u003cli\u003eC\u0026aacute;rcaba, A., Arrondo, R., \u0026amp; Gonz\u0026aacute;lez, E. (2022). Does good local governance improve subjective well-being? \u003cem\u003eEuropean Research on Management and Business Economics\u003c/em\u003e, \u003cem\u003e28\u003c/em\u003e(2), 100192. https://doi.org/10.1016/j.iedeen.2021.100192\u003c/li\u003e\n\u003cli\u003eCardozo, G. T. (2023). Virtually united: The role of the internet in building social capital. \u003cem\u003eAWARI\u003c/em\u003e, \u003cem\u003e4\u003c/em\u003e, 1～12. https://doi.org/10.47909/awari.605\u003c/li\u003e\n\u003cli\u003eChoi, B., Park, H., \u0026amp; Joo, S. (2022). Digital technology use and social capital: The moderating effect of social class. Innovation in Aging, 6(Supplement_1), 478.\u003c/li\u003e\n\u003cli\u003eChen, Y., \u0026amp; Yang, Y. Y. (2025). Downward subjective class identification and relative deprivation among the middle-income group: An empirical study based on CSS2021. \u003cem\u003eSocial Sciences in China\u003c/em\u003e, 45(3), 123-146. https://doi.org/10.12422/cass.2025.0003.\u003c/li\u003e\n\u003cli\u003e\u003cem\u003eColeman, J. S. (1988). Social capital in the creation of human capital. American journal of sociology, 94, S95-S120.\u003c/em\u003eDavidow, W. H. (2011). \u003cem\u003eOverconnected: The Promise and Threat of the Internet\u003c/em\u003e. Open Road Media.\u003c/li\u003e\n\u003cli\u003eDe Hesselle, L. C., \u0026amp; Montag, C. (2024). Effects of a 14-day social media abstinence on mental health and well-being: Results from an experimental study. \u003cem\u003eBMC Psychology\u003c/em\u003e, \u003cem\u003e12\u003c/em\u003e(1), 141. https://doi.org/10.1186/s40359-024-01611-1\u003c/li\u003e\n\u003cli\u003eDiener, E., Oishi, S., \u0026amp; Lucas, R. E. (2003). Personality, Culture, and Subjective Well-Being: Emotional and Cognitive Evaluations of Life. \u003cem\u003eAnnual Review of Psychology\u003c/em\u003e, \u003cem\u003e54\u003c/em\u003e(1), 403～425. https://doi.org/10.1146/annurev.psych.54.101601.145056\u003c/li\u003e\n\u003cli\u003e\u003cem\u003eDiener, E., Suh, E. M., Lucas, R. E., \u0026amp; Smith, H. L. (1999). Subjective well-being: three decades of progress. Psychological bulletin, 125(2), 276.\u003c/em\u003e\u003c/li\u003e\n\u003cli\u003eDiMaggio, P., Hargittai, E., Neuman, W. R., \u0026amp; Robinson, J. P. (2001). Social Implications of the Internet. \u003cem\u003eAnnual Review of Sociology\u003c/em\u003e, \u003cem\u003e27\u003c/em\u003e(Volume 27, 2001), 307～336. https://doi.org/10.1146/annurev.soc.27.1.307\u003c/li\u003e\n\u003cli\u003eFei, X. T. (1992). \u003cem\u003eFrom the soil: The foundations of Chinese society\u003c/em\u003e. Sanlian Shudian (Joint Publishing Company).\u003c/li\u003e\n\u003cli\u003eFu, X. (2018). The Contextual Effects of Political Trust on Happiness: Evidence from China. \u003cem\u003eSocial Indicators Research\u003c/em\u003e, \u003cem\u003e139\u003c/em\u003e(2), 491～516. https://doi.org/10.1007/s11205-017-1721-2\u003c/li\u003e\n\u003cli\u003eGargiulo, M., \u0026amp; Ertug, G. (2006). The Dark Side of Trust. \u003cem\u003eHandbook of Trust Research\u003c/em\u003e. Edward Elgar Publishing. https://www.elgaronline.com/edcollchap/9781843767541.00016.xml\u003c/li\u003e\n\u003cli\u003eGraham, C., \u0026amp; Chattopadhyay, S. (2013). Gender and well-being around the world. \u003cem\u003eInternational Journal of Happiness and Development\u003c/em\u003e. https://www.inderscienceonline.com/doi/10.1504/IJHD.2013.055648\u003c/li\u003e\n\u003cli\u003eHall, J. A., Xing, C., Ross, E. M., \u0026amp; Johnson, R. M. (2021). Experimentally manipulating social media abstinence: Results of a four-week diary study. \u003cem\u003eMedia Psychology\u003c/em\u003e, \u003cem\u003e24\u003c/em\u003e(2), 259～275. https://doi.org/10.1080/15213269.2019.1688171\u003c/li\u003e\n\u003cli\u003eHamilton, K., Helliwell, J. F., \u0026amp; Woolcock, M. (2016). \u003cem\u003eSocial Capital, Trust and Well-being in the Evaluation of Wealth\u003c/em\u003e (Working Paper No. 22556). National Bureau of Economic Research. https://doi.org/10.3386/w22556\u003c/li\u003e\n\u003cli\u003eHargittai, E. (2001). \u003cem\u003eSecond-Level Digital Divide: Mapping Differences in People\u0026rsquo;s Online Skills\u003c/em\u003e (No. arXiv:cs/0109068). arXiv. https://doi.org/10.48550/arXiv.cs/0109068\u003c/li\u003e\n\u003cli\u003eHargittai, E. (2011). The Digital Reproduction of Inequality. 收入 \u003cem\u003eThe Inequality Reader\u003c/em\u003e (2 edtion). Routledge.\u003c/li\u003e\n\u003cli\u003eHollebeek, L. D., Kumar, V., Srivastava, R. K., Lim, W. M., \u0026amp; Urbonavicius, S. (2025). Guidelines for theory selection: The IMPACT framework. Psychology \u0026amp; Marketing, 42(11), 2789\u0026ndash;2806. https://doi.org/10.1002/mar.70009\u003c/li\u003e\n\u003cli\u003eHorak, S., Klein, A., \u0026amp; Li, X. (2022). Can general and particularistic types of trust mix? Advancing the trust (dis-) continuity debate in a Chinese context. \u003cem\u003eInternational Journal of Emerging Markets\u003c/em\u003e, \u003cem\u003e17\u003c/em\u003e(5), 1273～1291. https://doi.org/10.1108/IJOEM-07-2020-0745\u003c/li\u003e\n\u003cli\u003eHoward, P. N. (2010). \u003cem\u003eThe Digital Origins of Dictatorship and Democracy: Information Technology and Political Islam\u003c/em\u003e. Oxford University Press.\u003c/li\u003e\n\u003cli\u003eJhang, F.-H. (2022). Bonding social capital and political trust: The mediating role of trust in public figures in Taiwan. \u003cem\u003eThe Social Science Journal\u003c/em\u003e, \u003cem\u003e59\u003c/em\u003e(1), 71～84. https://doi.org/10.1016/j.soscij.2019.07.009\u003c/li\u003e\n\u003cli\u003eKang, S., \u0026amp; Zhu, J. (2021). Do People Trust the Government More? Unpacking the Distinct Impacts of Anticorruption Policies on Political Trust. \u003cem\u003ePolitical Research Quarterly\u003c/em\u003e, \u003cem\u003e74\u003c/em\u003e(2), 434～449. https://doi.org/10.1177/1065912920912016\u003c/li\u003e\n\u003cli\u003eKatz, V. S., \u0026amp; Gonzalez, C. (2016). Toward Meaningful Connectivity: Using Multilevel Communication Research to Reframe Digital Inequality. \u003cem\u003eJournal of Communication\u003c/em\u003e, \u003cem\u003e66\u003c/em\u003e(2), 236～249. https://doi.org/10.1111/jcom.12214\u003c/li\u003e\n\u003cli\u003eKing, G., Pan, J., \u0026amp; Roberts, M. E. (2013). How Censorship in China Allows Government Criticism but Silences Collective Expression. \u003cem\u003eAmerican Political Science Review\u003c/em\u003e, \u003cem\u003e107\u003c/em\u003e(2), 326～343. https://doi.org/10.1017/S0003055413000014\u003c/li\u003e\n\u003cli\u003eLi, R., Xu, W., Nie, H., \u0026amp; Zhang, W. (2025). The Longitudinal Relationship Between Close Friendship and Subjective Well-Being: The Chain Mediation Model of Interpersonal Trust and Perceived Social Support. \u003cem\u003eBehavioral Sciences\u003c/em\u003e, \u003cem\u003e15\u003c/em\u003e(4), 480. https://doi.org/10.3390/bs15040480\u003c/li\u003e\n\u003cli\u003eLim, W. M. (2026). Guidelines for theory development: Typologies for theoretical novelty and interestingness.\u003c/li\u003e\n\u003cli\u003eMart\u0026iacute;nez, L. M., Estrada, D., \u0026amp; Prada, S. I. (2019). Mental health, interpersonal trust and subjective well-being in a high violence context. \u003cem\u003eSSM - Population Health\u003c/em\u003e, \u003cem\u003e8\u003c/em\u003e, 100423. https://doi.org/10.1016/j.ssmph.2019.100423\u003c/li\u003e\n\u003cli\u003eMathy, R. M., \u0026amp; Cooper, A. (2003). The duration and frequency of Internet use in a nonclinical sample: Suicidality, behavioral problems, and treatment histories. \u003cem\u003ePsychotherapy: Theory, Research, Practice, Training\u003c/em\u003e, \u003cem\u003e40\u003c/em\u003e(1～2), 125～135. https://doi.org/10.1037/0033-3204.40.1-2.125\u003c/li\u003e\n\u003cli\u003eMcCrae, N., Gettings, S., \u0026amp; Purssell, E. (2017). Social Media and Depressive Symptoms in Childhood and Adolescence: A Systematic Review. \u003cem\u003eAdolescent Research Review\u003c/em\u003e, \u003cem\u003e2\u003c/em\u003e(4), 315～330. https://doi.org/10.1007/s40894-017-0053-4\u003c/li\u003e\n\u003cli\u003eMcLean, G., Krey, N., \u0026amp; Barhorst, J. B. (2025). Revealing the double‐edged sword: Introducing the Technology and Consumer Well‐being Paradox Model. \u003cem\u003ePsychology \u0026amp; Marketing\u003c/em\u003e, \u003cem\u003e42\u003c/em\u003e(1), 5～20. https://doi.org/10.1002/mar.22110\u003c/li\u003e\n\u003cli\u003eMiao, X. L. (2023). Interconnection and mutual trust: Online social capital and generalized trust in the digital era. Open Times, (6), 102-116+8.\u003c/li\u003e\n\u003cli\u003eNorris, P. (2001). Making democracies work: social capital and civic engagement in 47 societies. Harvard University, John F. Kennedy School of Government Working Paper Series, (rwp01-036).\u003c/li\u003e\n\u003cli\u003eNorris, P. (2011). \u003cem\u003eDemocratic Deficit: Critical Citizens Revisited\u003c/em\u003e. Cambridge University Press. https://doi.org/10.1017/CBO9780511973383\u003c/li\u003e\n\u003cli\u003ePortes, A. (2000). Social Capital: Its Origins and Applications in Modern. \u003cem\u003eKnowledge and Social Capital: Foundations and Applications\u003c/em\u003e, 43.\u003c/li\u003e\n\u003cli\u003ePutnam, R. D. (2000). Bowling alone: America\u0026rsquo;s declining social capital.\u003cem\u003e In Culture and politics: A reader (pp. 223-234). New York: Palgrave Macmillan US. \u003c/em\u003e\u003c/li\u003e\n\u003cli\u003ePutnam, R. D. (1966). Political Attitudes and the Local Community. \u003cem\u003eAmerican Political Science Review\u003c/em\u003e, \u003cem\u003e60\u003c/em\u003e(3), 640～654. https://doi.org/10.2307/1952976\u003c/li\u003e\n\u003cli\u003eRatan, R. A., Chung, J. E., Shen, C., Williams, D., \u0026amp; Poole, M. S. (2010). Schmoozing and Smiting: Trust, Social Institutions, and Communication Patterns in an MMOG. \u003cem\u003eJournal of Computer-Mediated Communication\u003c/em\u003e, \u003cem\u003e16\u003c/em\u003e(1), 93～114. https://doi.org/10.1111/j.1083-6101.2010.01534.x\u003c/li\u003e\n\u003cli\u003eRodr\u0026iacute;guez-Pose, A., \u0026amp; Maslauskaite, K. (2012). Can policy make us happier? Individual characteristics, socio-economic factors and life satisfaction in Central and Eastern Europe. \u003cem\u003eCambridge Journal of Regions, Economy and Society\u003c/em\u003e, \u003cem\u003e5\u003c/em\u003e(1), 77～96. https://doi.org/10.1093/cjres/rsr038\u003c/li\u003e\n\u003cli\u003eRotter, J. B. (1980). Interpersonal trust, trustworthiness, and gullibility. \u003cem\u003eAmerican Psychologist\u003c/em\u003e, \u003cem\u003e35\u003c/em\u003e(1), 1～7. https://doi.org/10.1037/0003-066X.35.1.1\u003c/li\u003e\n\u003cli\u003eSeabrook, E. M., Kern, M. L., \u0026amp; Rickard, N. S. (2016). Social networking sites, depression, and anxiety: A systematic review. JMIR Mental Health, 3(4), e50. https://doi.org/10.2196/mental.5842\u003c/li\u003e\n\u003cli\u003eShensa, A., Sidani, J. E., Escobar-Viera, C. G., Switzer, G. E., Primack, B. A., \u0026amp; Choukas-Bradley, S. (2020). Emotional support from social media and face-to-face relationships: Associations with depression risk among young adults. \u003cem\u003eJournal of Affective Disorders\u003c/em\u003e, \u003cem\u003e260\u003c/em\u003e, 38～44. https://doi.org/10.1016/j.jad.2019.08.092\u003c/li\u003e\n\u003cli\u003eSilchenko, K. (2025). Technology Dualities and Paradoxes in Digital Well-Being: Analysis of Discourses of Digital Well-Being Market Offerings. \u003cem\u003eJournal of Macromarketing\u003c/em\u003e, \u003cem\u003e45\u003c/em\u003e(4), 550～573. https://doi.org/10.1177/02761467251356660\u003c/li\u003e\n\u003cli\u003eUslaner, E. M. (2002). The Moral Foundations of Trust (SSRN Scholarly Paper No. 824504). Social Science Research Network. https://doi.org/10.2139/ssrn.824504\u003c/li\u003e\n\u003cli\u003eUslaner, E. M. (2004). Trust, Civic Engagement, and the Internet. \u003cem\u003ePolitical Communication\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(2), 223～242. https://doi.org/10.1080/10584600490443895\u003c/li\u003e\n\u003cli\u003eVan Bruyssel, S., De Wolf, R., \u0026amp; Vanden Abeele, M. (2024). From bliss to burden: An ethnographic inquiry into how social, material and individual obstacles to digital well-being shape everyday life. \u003cem\u003eNew Media \u0026amp; Society\u003c/em\u003e, 14614448241288159. https://doi.org/10.1177/14614448241288159\u003c/li\u003e\n\u003cli\u003evan Deursen, A. J. A. M., \u0026amp; Helsper, E. J. (2015). \u003cem\u003eThe Third-Level Digital Divide: Who Benefits Most from Being Online?\u003c/em\u003e https://doi.org/10.1108/S2050-206020150000010002\u003c/li\u003e\n\u003cli\u003eWeber, M. (1968). \u003cem\u003eEconomy and society: An outline of interpretative sociology\u003c/em\u003e (G. Roth \u0026amp; C. Wittich, Eds.). University of California Press. (Original work published 1922)\u003c/li\u003e\n\u003cli\u003eYang, Y., Zeng, D., \u0026amp; Yang, F. (2022). Internet Use and Subjective Well-Being of the Elderly: An Analysis of the Mediating Effect Based on Social Capital. \u003cem\u003eInternational Journal of Environmental Research and Public Health\u003c/em\u003e, \u003cem\u003e19\u003c/em\u003e(19), 12087. https://doi.org/10.3390/ijerph191912087\u003c/li\u003e\n\u003cli\u003eYap, S.-F., \u0026amp; Lim, W. M. (2024). A Paradox Theory of Social Media Consumption and Child Well-Being. \u003cem\u003eAustralasian Marketing Journal\u003c/em\u003e, \u003cem\u003e32\u003c/em\u003e(1), 65～75. https://doi.org/10.1177/14413582221139492\u003c/li\u003e\n\u003cli\u003eYoon, S. (2014). Does social capital affect SNS usage? A look at the roles of subjective well-being and social identity. \u003cem\u003eComputers in Human Behavior\u003c/em\u003e, \u003cem\u003e41\u003c/em\u003e, 295～303. https://doi.org/10.1016/j.chb.2014.09.043\u003c/li\u003e\n\u003cli\u003eYou, Y., \u0026amp; Wang, Z. (2020). The Internet, political trust, and regime types: A cross-national and multilevel analysis. \u003cem\u003eJapanese Journal of Political Science\u003c/em\u003e, \u003cem\u003e21\u003c/em\u003e(2), 68～89. https://doi.org/10.1017/S1468109919000203\u003c/li\u003e\n\u003cli\u003eZhan, S., Wang, Y., \u0026amp; Yang, Y. (2025). How Internet Use Affects Hierarchical Political Trust pattern?\u0026mdash;Empirical analysis based on CSS2021 data. \u003cem\u003ePublic Administration and Policy Review\u003c/em\u003e, \u003cem\u003e14\u003c/em\u003e(2), 87.\u003c/li\u003e\n\u003cli\u003eZhang, D., Shi, J., \u0026amp; Feng, Z. (2025). The dark side of the association between internet use and older adults\u0026rsquo; subjective wellbeing: The role of subjective social class. \u003cem\u003eBMC Geriatrics\u003c/em\u003e, \u003cem\u003e25\u003c/em\u003e(1), 63. https://doi.org/10.1186/s12877-024-05606-w\u003c/li\u003e\n\u003cli\u003eZhang, C. M., \u0026amp; Li, Y. Q. (2023). Internet use, social trust, and residents\u0026apos; well-being: A study based on CFPS data. Rural Finance Research, (5), 70-80. Retrieved from https://www.cqvip.com/qk/93444X/202305/\u003c/li\u003e\n\u003cli\u003eZhou, B. (2019). Fear of missing out, feeling of acceleration, and being permanently online: A survey study of university students\u0026rsquo; use of mobile apps in China. \u003cem\u003eChinese Journal of Communication\u003c/em\u003e. https://www.tandfonline.com/doi/abs/10.1080/17544750.2018.1523803\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"trust heterogeneity, internet usage, interpersonal trust, political trust, lagged fixed effects","lastPublishedDoi":"10.21203/rs.3.rs-9225549/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9225549/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe increasing integration of the internet into public life has ignited an intense academic debate regarding whether digital technology fundamentally fosters or erodes social capital and subjective well-being. To address this complexity, our study rigorously utilized the Lagged Fixed Effects (LFE) model and multi-period panel data from the China Family Panel Studies (CFPS) to examine the long-term, cross-period causal structure of internet use and well-being. The research confirms a highly conditional and heterogeneous dual-track mechanism. While internet use generates psychological costs via family and neighbourhood trust (bonding capital), its ability to foster stranger trust (bridging capital) is strictly contingent upon subjective class identity. However, even for individuals with higher class recognition who successfully accumulate this bridging trust, it ultimately lacks the emotional depth to translate into long-term subjective well-being. Crucially, the study found a persistent barrier preventing this interpersonal trust from converting into institutional trust, despite political trust being the key driver of long-term well-being. By uncovering these asymmetric mechanisms and boundary conditions, this research significantly advances social capital theory and social comparison theory in the digital context. The findings provide crucial evidence for understanding the digital well-being paradox, offering targeted insights for public governance aimed at mitigating the unequal distribution of digital benefits across different social strata.\u003c/p\u003e","manuscriptTitle":"The Conditional Effects of Internet Use on Well-being: Heterogeneous Trust Mechanisms and the Boundary of Subjective Social Status","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-04-06 14:30:45","doi":"10.21203/rs.3.rs-9225549/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2026-05-08T04:58:18+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-07T02:13:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"268593603215973895838871491789092399167","date":"2026-04-17T00:26:35+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-04-16T17:14:00+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"261613022718282953947491832087067077037","date":"2026-04-14T11:22:46+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"99754351595156220224516763686302602641","date":"2026-04-03T17:55:44+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2026-03-31T11:45:11+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2026-03-30T09:34:04+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2026-03-27T11:11:46+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2026-03-27T11:11:28+00:00","index":"","fulltext":""},{"type":"submitted","content":"BMC Psychology","date":"2026-03-25T16:06:15+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"bmc-psychology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"psyo","sideBox":"Learn more about [BMC Psychology](http://bmcpsychology.biomedcentral.com/)","snPcode":"","submissionUrl":"","title":"BMC Psychology","twitterHandle":"BMC_series","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"BMC Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"90aa9a11-d5c3-489f-8581-9d971e9e2ddc","owner":[],"postedDate":"April 6th, 2026","published":true,"recentEditorialEvents":[{"type":"decision","content":"Revision requested","date":"2026-05-08T04:58:18+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2026-05-07T02:13:35+00:00","index":32,"fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2026-05-18T13:23:15+00:00","versionOfRecord":[],"versionCreatedAt":"2026-04-06 14:30:45","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9225549","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9225549","identity":"rs-9225549","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}