Earticle

Firms increasingly reference artificial intelligence in annual reports to signal technological sophistication, yet whether such disclosures represent genuine innovation commitment or symbolic posturing remains unclear. This distinction carries significant theoretical and practical stakes: the proliferation of AI-related corporate communication introduces information asymmetry into capital markets, complicates the design of disclosure standards, and distorts the allocation of innovation subsidies toward firms engaging in rhetorical rather than substantive technological commitment. This study examines how AI-related corporate disclosures translate into innovation outcomes and financial performance, addressing gaps in understanding performative versus symbolic dimensions of technology communication. Drawing on signaling theory, the resource-based view, and institutional perspectives, we theorize AI disclosure operates through dual pathways: externally, by inviting stakeholder monitoring and attracting complementary resources; internally, by creating accountability mechanisms coordinating organizational attention and stimulating innovation momentum. We test this framework using comprehensive panel data from 36,200 firm-year observations of Chinese listed firms spanning 2007-2023, employing firm- and year-fixed effects regressions, propensity score matching, and mediation analyses with bootstrapped standard errors. Results demonstrate AI disclosure intensity strongly predicts subsequent invention patent output, with 10% increase in disclosure associating with approximately 2.8% increase in patent applications. Innovation partially mediates the disclosure-performance relationship, with significant positive indirect effects confirmed through bootstrapped estimates, though direct effects are negative, reflecting short-term costs before innovation benefits materialize. Critically, the disclosure-innovation effect diminishes among R&D-intensive firms, suggesting signaling substitution wherein communicative emphasis provides limited incremental value when firms already demonstrate technological commitment through observable R&D spending. Conversely, the effect strengthens by 34% in market-oriented institutional contexts—specifically among non-financially distressed firms operating under fewer regulatory constraints—indicating external accountability structures amplify conversion of symbolic communication into substantive innovation. Theoretically, we reconceptualize disclosure as performative communication catalyzing capability development when embedded in accountability structures, extending signaling theory by demonstrating communicative acts mobilize resources beyond passive information revelation. We identify absorptive capacity and institutional accountability as critical boundary conditions for symbolic-substantive conversion, specifying micro-mechanisms through which external signals become internal capabilities.

Drawing on panel data from Chinese A-share listed firms during 2013–2024, this study develops a comprehensive analytical framework of “board stability–agency costs–investment efficiency” to examine how board stability affects corporate investment efficiency and to uncover the underlying mechanisms through which this influence operates. In the context of emerging markets, where corporate governance quality and efficient resource allocation are increasingly emphasized, understanding the role of dynamic governance characteristics has become particularly important. While prior research has mainly focused on static board characteristics such as board size, independence, and diversity, relatively little attention has been devoted to the governance implications of board stability as a dynamic feature of corporate governance structures. To address this gap, the present study employs panel regression models with firm and year fixed effects to test the relationship between board stability and investment efficiency, while conducting a series of robustness checks and endogeneity tests to ensure the reliability of the empirical findings. The empirical results demonstrate that board stability significantly improves corporate investment efficiency, indicating that stable governance structures help maintain strategic continuity and enhance the effectiveness of resource allocation within firms. A stable board allows directors to accumulate firm-specific governance knowledge and industry experience, which strengthens monitoring effectiveness and improves the quality of strategic decision-making. Further analysis reveals that agency costs serve as an important transmission channel linking board stability and investment efficiency. Specifically, board stability reduces agency costs by strengthening internal monitoring and mitigating managerial opportunistic behavior, thereby indirectly enhancing investment efficiency. In addition, employee size is found to negatively moderate the relationship between board stability and investment efficiency. As organizational scale expands, increasing governance complexity, communication costs, and coordination difficulties may weaken the governance advantages associated with stable boards. This finding suggests that the effectiveness of governance mechanisms may vary depending on firm characteristics and organizational structure. This study contributes to the corporate governance literature in several ways. By introducing a dynamic governance perspective, it expands the existing research framework on board characteristics and corporate investment efficiency. The study also identifies the mediating role of agency costs and the moderating influence of employee size, thereby providing a more comprehensive understanding of the mechanisms through which board stability shapes corporate investment behavior. Finally, the findings offer empirical evidence from the Chinese institutional context, providing practical implications for improving governance mechanisms and promoting more efficient investment decisions among listed firms.

China's Long-Term Care Insurance (LTCI), established as the country's “Sixth Social Insurance” following pension, medical, work-injury, unemployment, and maternity insurance, is undergoing a critical transition from selective pilot programs toward nationwide institutionalization. Yet this expansion has brought deep structural contradictions into sharp relief: why do multi-actor coordination failures persist despite continuous technological upgrading, and why does rapid enrollment growth coexist with widening regional imbalances? Existing research has largely approached LTCI through an instrumental efficiency lens, leaving the governance mechanisms that underpin—or undermine—institutional performance poorly understood. This study addresses that gap by constructing an integrated analytical framework that links institutional logic conflicts, collaborative governance mechanisms, and public value realization, drawing on institutional logics theory (Thornton et al., 2012), collaborative governance theory (Ansell & Gash, 2008), and public value theory (Moore, 1995). Empirically, DEA-Malmquist analysis of 44 pilot cities over 2020-2024 finds that the national average technical efficiency (TE) of LTCI financing stands at only 0.738, constrained primarily by insufficient scale efficiency (SE mean: 0.818), with only 10 of 44 cities achieving DEA efficiency. From a dynamic perspective, total factor productivity (TFP) averages 0.891 across the observation periods—declining on average by approximately 10.9% per period—driven by declining technical efficiency (TEC mean: 0.887) while technological change remained stable (TC mean: 1.003). Dagum Gini decomposition further shows that inter-regional disparities account for 43.0% of total efficiency variation. These findings are interpreted through the lens of four competing institutional logics—state, market, professional, and community—whose unresolved tensions produce structural governance failure. Comparative case analysis of the Shanghai and Nantong models illustrates how different logic integration strategies generate distinct public value outcomes. Systematic evaluation along three public value dimensions reveals significant gaps: in outcome value (insufficient fiscal sustainability and persistent regional inequity), process value (low administrative transparency and limited democratic participation), and relational value (fragile multi-stakeholder trust and weak value consensus). The paper concludes that the over-reliance on technological fixes and administrative monitoring—what Power (1997) termed “audit governance”—cannot substitute for genuine collaborative governance reform. Building a public-value-oriented collaborative platform, extending commercial insurer contract cycles, establishing unified digital assessment infrastructure, and institutionalizing multi-stakeholder participation are identified as the key levers for converting logic conflicts into institutional symbiosis and consolidating the long-term foundations of China's LTCI system.

Commercial banks are increasingly expanding intermediary services, such as mutual funds and bancassurance, to reduce reliance on interest-margin-based profit models. However, identifying customers with a high probability of purchasing such products remains a significant challenge for retail banks. Although machine learning (ML) has been widely applied in banking, most existing studies focus on risk management tasks, while relatively limited attention has been paid to customer targeting, product adoption, and cross-selling for intermediary financial services. In addition, highly accurate ML models often lack the interpretability required for marketing management. To address these gaps, this study proposes an interpretable ML framework for predicting customer purchasing behavior for bank intermediary products. Using survey data from 368 retail banking customers in Hangzhou, China, this study constructs a multidimensional customer profile including demographic, financial, investment-related, behavioral, health-related, and psychological variables. Several classification models are compared, including Logistic Regression (LR), Naïve Bayes (NB), Support Vector Machine (SVM), Decision Tree (DT), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), and Neural Network (NN). The results show that XGBoost achieves the highest performance among the compared models. Furthermore, SHapley Additive exPlanations (SHAP) analysis and Partial Dependence Plots (PDPs) are employed to interpret the prediction results. The interpretability results reveal that investment experience and deposit balance are the most important predictors of purchasing behavior. Behavioral engagement variables, such as transaction frequency, credit card use, and mobile banking activity, also contribute to customer classification. The interaction analysis further indicates a compensatory pattern: active daily banking engagement can increase purchase probability even among customers with no clear investment preferences. This finding suggests that purchase decisions are shaped not only by financial capacity and prior investment experience, but also by customers’ ongoing interactions with banking services. Theoretically, this study extends ML applications in banking from risk management to proactive cross-selling. Practically, the findings provide implications for customer segmentation, precision marketing, and intermediary product recommendation. Overall, this study shows that interpretable ML can deliver both predictive accuracy and managerial insight in retail banking.

A long-standing popular narrative in China—the “Little Emperor” hypothesis—holds that children raised without siblings under the One-Child Policy (OCP) are overfed and overindulged, placing them at elevated risk of adult obesity. This paper provides new causal evidence on the long-run physical-health consequences of only-child status induced by the policy. The analysis draws on pooled CHNS data for individuals born in or after 1979 and observed in early adulthood. To address the endogeneity of fertility decisions, I instrument only-child status with the interaction of provincial fine-to-income ratios and pre-outcome community-to-province income weights, exploiting geographic and temporal variation in OCP enforcement intensity. OLS estimates reproduce the conventional positive association: with family and community controls, only children exhibit higher BMI and are more likely to be overweight. Once endogeneity is addressed, however, the sign reverses: the IV estimates suggest lower adult BMI and overweight risk among compliers, with a similarly negative but less precisely estimated effect on clinical obesity. First-stage diagnostics slightly exceed the conventional weak-instrument threshold, and a specification test strongly rejects the null that OLS and IV estimates are equal, consistent with substantial upward selection bias in the naive association. This pattern is consistent with positive selection on community affluence biasing the OLS estimates upward. Under a local average treatment effect (LATE) interpretation, the resource-concentration channel appears to dominate the indulgence channel among compliers: parents induced into single-child households by stricter enforcement direct marginal resources toward nutritional balance rather than caloric surplus. The BMI and overweight findings survive a battery of robustness checks—alternative cutoffs, clustered inference, province fixed effects, sample expansion, and a placebo test—while the obesity estimates remain negative but less robust. Gender-stratified estimates, interpreted cautiously given weak first-stage instrument strength, suggest the negative effects on BMI and overweight may be more pronounced among women, consistent with differential parental investment and stronger thinness norms for daughters in contemporary China. Overall, the OCP's resource-concentration effect may have operated as a net protective factor against higher adult BMI and overweight risk among the policy's compliers.

Although prior studies have extensively documented the momentum effect across different markets, time periods, and asset classes, most of the existing literature focuses on individual determinants of momentum returns in isolation. In particular, there is limited empirical evidence on how momentum profitability jointly varies with market conditions and investor ownership structure. Motivated by this gap, this study investigates the profitability of momentum strategies in the Chinese stock market using common stocks listed on the Shanghai Stock Exchange. In addition to the conventional winner-minus-loser (WML) strategy, we also examine alternative volatility-managed momentum strategies. Furthermore, following Cooper et al. (2004), we classify market states into up and down markets to analyze whether momentum returns depend on market conditions. Finally, we explore how momentum profitability varies with the level of institutional ownership. The empirical results can be summarized as follows. First, both the traditional momentum strategy and volatility-managed momentum strategies generate statistically significant positive returns in the Chinese stock market, with the Daniel and Moskowitz (2016) strategy exhibiting the strongest performance. Second, momentum profits are primarily driven by up-market periods, suggesting that market conditions play a crucial role in determining strategy effectiveness. Third, momentum returns are significantly higher in portfolios with greater institutional ownership, implying that institutional investors contribute more to the generation of momentum profits. These findings provide important economic implications. They suggest that incorporating risk management and market timing can enhance momentum strategy performance, and that investor composition—particularly institutional ownership—plays a critical role in explaining momentum returns. This study contributes to the literature by offering a more comprehensive understanding of momentum profitability through the joint consideration of market conditions and investor ownership structure. By highlighting the role of institutional investors, it also provides new insights into the underlying mechanism of the momentum effect. Nevertheless, this study is limited by its focus on stocks listed on the Shanghai Stock Exchange, which may restrict the representativeness of the findings, suggesting that future research should extend the analysis to include firms listed on the Shenzhen and Beijing Stock Exchanges to improve generalizability.

This study empirically investigates the impact of cross-border e-commerce (CBEC) development on the industrial employment structure in China using balanced panel data from 30 Chinese provinces from 2013 to 2021. As digitalization and global online trade rapidly expand, CBEC has become an important driver of economic transformation in China. Supported by government policies, logistics infrastructure, and digital platforms, CBEC growth is expected to influence labor allocation across industries. To measure the level of CBEC development, this study constructs a multidimensional composite index based on indicators related to scale, support environment, and development potential using the entropy weight method. A two-way fixed effects panel model is employed to estimate the effects of CBEC on employment shares in the primary, secondary, and tertiary industries while controlling for regional and temporal heterogeneity. The empirical results show that CBEC development significantly reduces the employment share of the primary industry and significantly increases the employment share of the tertiary industry, while its impact on the secondary industry is statistically insignificant. These findings indicate that CBEC facilitates labor reallocation from traditional agricultural sectors to service-oriented industries such as logistics, digital platforms, and online retail services. The results are consistent with theories of structural transformation emphasizing labor movement from low-productivity sectors to high-productivity sectors during economic development. In addition, the effects of CBEC development exhibit clear regional heterogeneity. The positive employment effects are more pronounced in eastern provinces, where digital infrastructure and market accessibility are relatively advanced, whereas the effects are weaker in central and western regions. Robustness checks using an alternative indicator of digital economy development confirm the consistency of the main findings. Overall, this study provides empirical evidence that CBEC serves as an important driver of industrial upgrading and employment structure transformation in China. The findings suggest that policies supporting digital infrastructure expansion, regional connectivity, and labor mobility are essential for promoting balanced regional development and sustainable employment growth in the digital economy era.

China's new energy vehicle (NEV) market has expanded rapidly, with projected sales of 10.95 million units in 2024, accounting for 78.2% of global NEV sales. Against the backdrop of intensifying domestic competition from brands such as BYD, Li Auto, and NIO, as well as accelerating market entry by traditional multinational automakers including Volkswagen, BMW, and Mercedes-Benz, how multinational automakers sustain brand competitiveness has become a significant research question in marketing. Tesla's operations in China present a structural paradox in which sales volume grew from 148,000 units in 2020 to 657,000 units in 2024, while market share simultaneously declined from a peak of 10.5% to 6.0%, and the mechanism through which its marketing strategies influence consumer purchase intention remains insufficiently examined. This study adopts the 4P marketing mix theory as its analytical framework and operationalizes the four dimensions as brand influence (promotion), technology level (product), price factor (price), and service quality (place). A mediation model of "4P strategy perception—competitive comparison cognition—purchase intention" is constructed, with competitive comparison cognition grounded in Hsee and Leclerc's (1998) comparative evaluation theory. A questionnaire survey was administered to Chinese NEV consumers aged 18 and above via a snowball sampling approach on the Wenjuanxing platform, yielding 320 valid responses from 18 cities across China. Hierarchical regression analysis and the Baron & Kenny mediation procedure were employed to test five research hypotheses, with mediation significance further verified by the Sobel test. The results indicate that brand influence, technology level, price factor, and service quality all exert significant positive effects on purchase intention, with brand influence showing the strongest effect (β=0.38), followed by technology level (β=0.31), price factor (β=0.19), and service quality (β=0.16). The overall model demonstrated satisfactory explanatory power (adjusted R²=0.573, F=42.61,p<0.001). Competitive comparison cognition was found to mediate the relationship between 4P strategy perception and purchase intention significantly (Sobel Z=4.23, p<0.001), accounting for 34.7% of the total effect, suggesting that consumers translate their evaluations of Tesla's marketing strategies into purchase decisions through a process of comparative assessment against competing brands rather than through direct perception alone. This study integrates the 4P framework with consumer behavior analysis and introduces competitive comparison cognition as a mediating variable, thereby extending the methodological scope of localization marketing research on multinational automakers. The findings offer empirically grounded strategic guidance for other NEV brands seeking to enter or expand within the Chinese market.

As the mobile internet and social media become deeply embedded in the tourism field, the spatial production logic of tourist destinations is undergoing a profound transformation. Through continuous content production, social media platforms have reshaped the dissemination of tourism information, profoundly influencing tourists' cognitive and emotional construction of place. Yanji, an emerging tourist city rapidly popularized by the internet in recent years, serves as a crucial case for observing the reconstruction of the sense of place within a mediatized context. Taking Yanji as a case study, this research explores how mediatized logic deeply participates in the generation and evolution of the tourism sense of place through the interactive mechanism of digital representation and embodied practice. Employing a grounded theory approach based on online texts, this study utilizes User-Generated Content (UGC) from social media platforms as analytical data. Through open, axial, and selective coding, it constructs a model of the "circulatory generation mechanism of the mediatized gaze." The findings reveal that: First, through symbolic reinforcement and algorithmic recommendation, social media continuously amplifies specific visual elements and emotional narratives of Yanji, forming a highly symbolized destination image. This achieves a pre-establishment of the sense of place before tourists even depart, shifting the locus of the gaze from geographical entities to digital media. Secondly, during the embodied practice phase, visitors validate media imagery within the physical space through ‘check-in’ activities and spatial performances. In this process, physical experiences such as overcrowding, queuing and adverse weather conditions do not diminish the sense of place; rather, they reinforce the individual’s sense of presence and experiential memory to a certain extent; Thirdly, during the interactive feedback stage, tourists incorporate their individual experiences into a social context through commenting, sharing and reposting, driving the continuous revision and reproduction of local meaning, thereby forming a dynamic process that unfolds progressively from ‘local imagery – local attachment – local identity’. Research indicates that, within the context of social media, the tourist gaze has shifted from a unidirectional visual observation to a cyclical process encompassing experience, expression and dissemination; consequently, the generation of a sense of place exhibits comprehensive characteristics of media-guided, embodied participation and interactive construction. Simultaneously, whilst the process of mediatisation enhances the visibility of a place, it also reshapes its socio-spatial structure to a certain extent. This paper contributes to a deeper understanding of the mechanisms underlying the generation of a sense of place in the digital age from a mediatisation perspective, and offers insights for the communication and spatial governance of tourist destinations.

Against the backdrop of rising external uncertainty, profound restructuring of industrial and supply chains, and the continued advancement of the unified national market, enhancing regional supply chain resilience has become an important issue in promoting high-quality development. Based on panel data from 30 regions in China over the period 2010–2024, this study constructs a comprehensive index of modern circulation system upgrading (MLU) from the perspectives of logistics infrastructure, digital circulation capability, and producer service empowerment, as well as a regional supply chain resilience (SCR) index capturing accessibility, supporting capacity, stability, and recovery ability. Within a spatial econometric framework, this paper examines the spatial association between modern circulation system upgrading and regional supply chain resilience. The results of the global Moran’s I test indicate that regional supply chain resilience exhibits significant positive spatial autocorrelation. After model specification tests, a two-way fixed effects Spatial Durbin Model (SDM) is employed for empirical estimation. To accurately capture the multidimensional interactions among provinces, an economic-geographic nested spatial weight matrix is utilized as the benchmark setting, providing a more rigorous estimation of cross-regional dependencies. The findings show that the upgrading of the modern circulation system is positively associated with local supply chain resilience and also exhibits certain spatial linkage characteristics across regions. These results suggest that improvements in logistics connectivity, digital circulation capacity, and producer-service support are closely related to regional supply chain resilience. Robustness checks based on alternative spatial weight matrices indicate that the direction of the core estimates remains generally consistent across different spatial settings. It should be noted that both the modern circulation system and regional supply chain resilience are comprehensive regional-level constructs. Some statistical indicators may reflect closely related aspects of regional economic operation. Therefore, the findings of this study should be interpreted as empirical evidence of spatial association rather than strict causal identification. This study contributes to the literature by integrating modern circulation system upgrading and supply chain resilience into a regional analytical framework and provides policy implications for improving circulation networks, strengthening cross-regional coordination, and enhancing regional supply chain resilience.