Does digitalization always benefit cultural, sports, and tourism enterprises quality? Unveiling the inverted U-shaped relationship from a resource and capability perspective

The study investigates whether and how digitalization improves the high-quality development of cultural, sports, and tourism enterprises, sectors that are labor-intensive and relatively technology-sparse compared with manufacturing. It highlights practical challenges in China’s industry integration and the uneven application of digital technologies, skills shortages, and mismatches with consumer expectations. Empirical research has been limited by data constraints and unsuitable measures of firm digitalization. The paper poses three research questions: (RQ1) What constitutes high-quality development for these enterprises within an integrated culture–sports–tourism framework? (RQ2) How are digital technologies used and how can digitalization be measured better? (RQ3) What is the relationship between digitalization and high-quality development and how do resources and capabilities (human capital, innovation, core competencies) shape this relationship? The authors propose that digitalization’s effect is nonlinear (inverted U-shaped), with potential benefits at moderate levels and crowding-out costs when excessive, and ground their analysis in the resource-based view and expanded core capabilities (including search, reorganization, and attention).

The review covers four strands: (1) High-quality development measurement often relies on total factor productivity (TFP) and industry-level frameworks, with limited attention to multi-industry integration and enterprise-level assessments. Drawing on manufacturing, the study argues that core dimensions for cultural, sports, and tourism enterprises should include efficiency, innovation, environmental sustainability, integration, openness, and social contribution. (2) High-quality development under integration has been explored mainly qualitatively across sports tourism, cultural tourism, and culture–tourism integration, with few enterprise-level empirical studies and limited consideration of integration dynamics. (3) Digitalization research in hospitality and tourism is largely qualitative (case studies, surveys) focusing on smart tourism/hotels and digital capabilities; data limitations hinder micro-level empirical work and robust measurement. (4) The impact of digitalization on development quality is theoretically mixed: benefits include recovery from shocks and enhanced competitiveness, but risks include workforce disruption, managerial challenges, and consumer technology overload. Overall, there is a gap in enterprise-level, data-driven evaluation of digitalization and high-quality development within an integrated industry context and in uncovering internal mechanisms involving resources and capabilities.

Design: An empirical panel study of 116 A-share listed cultural (62), sports (11), and tourism (43) enterprises in China from 2010–2020, categorized per Liang et al. (2021). Financial data from CSMAR; MD&A texts from CNRDS. High-quality development (dependent variable): Constructed an enterprise-level composite index tailored to integration of culture–sports–tourism. Six subobjectives: high-quality efficient growth, innovative development, green development, integrated development, open cooperation, and social sharing. Weights determined objectively via entropy method. Integration was explicitly incorporated using social network analysis (UCINET) based on (a) managerial interlocks (directors/supervisors/executives holding part-time roles across firms) as an undirected weighted network and (b) interfirm investment ties (securities/equity/related transactions) as a directed weighted network. Degree centrality from both networks entered the integrated development dimension. As robustness alternative dependent variables, TFP was computed via OP and LP methods. Digitalization (key explanatory variable): Built a domain-specific thesaurus of 112 keywords across strategic, technology (interactive AI, VR/AR/MR, e-navigation, 3D modeling/digital twin/BIM, intelligent sensing/5G/WiFi), and application layers (e.g., smart scenic spots, big data, cloud, platforms, industrial internet). Processed MD&A texts with custom dictionary and Jieba segmentation; computed firm-year TF-IDF weighted digitalization index: sum over keywords of ln(tf+1)*ln(N/(n(w)+1)). A time-agnostic IDF variant was used in endogeneity checks to mitigate measurement error. Controls: Major shareholder fund occupation, management expense ratio, Tobin’s Q, book-to-market, CEO duality, loss indicator, inventory share, years listed. Fixed effects for time and province. Models: Baseline fixed-effects regression with digital and digital^2 to test for nonlinearity. Mediation tested via Baron–Kenny stepwise regressions for human capital (share of employees with bachelor’s or above) and enterprise innovation (R&D intensity = R&D/sales). Dynamic effects examined using up to five-period lags. Inverted U verified via Sasabuchi–Lind–Mehlum U-test and Chow test for turning point. Endogeneity and robustness: Addressed measurement error by recomputing IDF on full-period corpus; controlled time–province and industry–province interaction fixed effects; 2SLS/LIML with two instruments: (IV1) interaction of regional digital infrastructure and firm’s lagged digitalization; (IV2) Bartik-style instrument using industry peers’ initial digitalization share times aggregate growth, and its square for the quadratic term. Additional robustness: excluding centrally administered municipalities; excluding 2020 (COVID-19); alternative dependent variables (TFP OP/LP). Heterogeneity analyses by region (eastern/central/western), firm life cycle (growth+maturity vs decline), and risk-taking (ROA volatility).

- Baseline relationship: Digitalization exhibits a significant inverted U-shaped association with high-quality development. Linear term positive and quadratic term negative across specifications with time and province fixed effects; effect remains when including controls. The turning point (extremal point) is estimated at digitalization index ≈ 24.336, well within the observed range [0, 53.316]. The majority of firms lie below the 97.8th percentile threshold and thus operate in the “moderate zone,” where digitalization promotes quality. - Persistence: Lagged models (up to 5 periods) show the inverted U-shape persists over time; the magnitude of the effect gradually weakens across longer lags. - Validity and endogeneity: Results are robust to alternative IDF computation, interaction fixed effects, and 2SLS/LIML with two instruments. First-stage instruments are significant; second-stage retains positive linear and negative quadratic effects. Weak-IV diagnostics suggest acceptable instrument strength (e.g., Anderson LM significant; Cragg–Donald F ≈ 7.019 near 10% threshold). Inverted U confirmed by Sasabuchi–Lind–Mehlum test (significant) with slopes positive at lower bound and negative at upper bound. - Robustness: Excluding municipalities and excluding 2020 (COVID-19) preserve the inverted U pattern. Using TFP (OP/LP) as the outcome also yields significant positive linear and negative quadratic terms. - Mediation by human capital: Digitalization’s effect on human capital is itself inverted U-shaped (positive linear, negative quadratic). Human capital positively predicts high-quality development and mediates the inverted U-shaped digitalization–quality link (H2a, H2b supported). - Mediation by enterprise innovation: Digitalization has an inverted U-shaped relation with R&D intensity; innovation positively predicts development quality and mediates the inverted U-shaped relationship (H3a, H3b supported). - Heterogeneity: - Region: Eastern and central regions show significant inverted U effects; in the western region, the linear term is not significant and the quadratic term is weak, implying greater potential and/or limited current impact. - Life cycle: Growth+maturity firms exhibit inverted U effects; declining firms show mainly positive linear effects (organizational rigidity dampens nonlinearity). - Risk-taking: Low risk-taking firms show inverted U; high risk-taking firms show positive linear effects, indicating greater ability to leverage digitalization without incurring diminishing returns as quickly.

The findings address RQ1 by operationalizing high-quality development for cultural–sports–tourism enterprises in an integrated framework, explicitly incorporating interfirm integration via social networks alongside efficiency, innovation, green, openness, and social sharing dimensions. For RQ2, the study proposes a domain-specific, TF-IDF-based digitalization measure using MD&A texts and a curated thesaurus spanning strategy–technology–application layers, improving sensitivity to sectoral digital practices. For RQ3, the evidence supports an inverted U-shaped relationship between digitalization and high-quality development, consistent with resource-based and core capability perspectives. Digitalization initially enhances resources and capabilities (human capital upgrading; search and reorganization capabilities for innovation) but, beyond a threshold, attention constraints, displacement/erosion of human capital, organizational strain, and technology overload can crowd out benefits. Mediation analyses substantiate human capital and innovation as channels through which nonlinearity manifests. Heterogeneity results highlight contextual dependencies: more developed regions and firms in growth/maturity phases benefit more, and higher risk tolerance can extend the beneficial range. Collectively, these insights refine theory by expanding core competencies to include search, reorganization, and attention in digital contexts and by revealing how resource/capability dynamics generate nonlinear performance effects.

This study contributes methodologically by (1) constructing an enterprise-level high-quality development index that embeds integration through network centrality and (2) devising a sector-specific, TF-IDF-based digitalization metric from MD&A texts. Theoretically, it extends core capability concepts (adding search, reorganization, attention) and demonstrates that digitalization’s impact on firm development quality is inverted U-shaped, with human capital and innovation as mediating mechanisms. Empirically, using 2010–2020 data for 116 Chinese listed cultural–sports–tourism enterprises, the study confirms robust, persistent, but gradually weakening nonlinear effects and reveals regional, life-cycle, and risk-taking heterogeneity. Policy and managerial implications include calibrating digital investments to a moderate range, building digital talent and innovation chains, tailoring regional support (e.g., strengthening western region infrastructure and capabilities), fostering platforms for integrated development, and enhancing risk management. Future research should broaden mechanisms (e.g., governance, data governance, organizational design), explore negative externalities (privacy/security risks, cultural homogenization), and refine measurement to reduce disclosure biases.

- Measurement bias: MD&A-based digitalization may reflect strategic disclosure or ostentation; TF-IDF weighting mitigates but does not eliminate this bias. - Sectoral specificity: Results reflect labor-intensive, technology-sparse characteristics; mediating roles of human capital and innovation may differ in tech-intensive settings. - Mechanism scope: Other mechanisms (e.g., governance, data security practices, organizational design) may influence the inverted U-shape but were not modeled. - Externalities: The study does not empirically assess negative impacts such as data privacy/security risks or potential erosion of local cultural characteristics. - Generalizability: Focus on Chinese A-share listed firms may limit external validity across countries or unlisted SMEs.