Regional Industrial & Technology Diversity in Sport Firms' Performance
Description
Data description: This study uses a comprehensive firm-level dataset of 11,968 sport-related firms located across Chinese cities. Firm-level information, including location, industry classification, employment, governance characteristics, and financial indicators, was obtained from the Orbis database (Moody’s Analytics) for 2022–2023. The sample includes firms engaged in diverse sport-related activities, including equipment manufacturing, digital platforms, media-related services, training, recreation, and supporting activities. Firms with fewer than five employees and those unrelated to sport activities were excluded to ensure meaningful measurement of human capital returns. Firm-level data are combined with city-level indicators from official Chinese statistical sources to capture regional economic and demographic conditions. The city is used as the primary spatial unit because it represents the level at which firms, technologies, infrastructures, and institutions interact to form regional ecosystems. The final dataset enables the analysis of how industrial cluster density, regional technology diversity, and their interaction are associated with firm-level return on human capital (ROHC) using multilevel mixed-effects models.
Files
Steps to reproduce
Steps to reproduce: Obtain access to the Orbis (Moody’s Analytics) database and extract firm-level records for sport-related firms operating in China, including firm location, industry classification, employment, financial information, governance characteristics, and firm age. Apply the sample selection criteria used in this study: retain firms founded before 2019, exclude firms unrelated to sport activities, and remove firms with fewer than five employees. The final sample should contain 11,968 firms. Merge firm-level records with city-level socioeconomic indicators obtained from official Chinese statistical sources. Construct the key variables: Calculate return on human capital (ROHC) as net income divided by total employee costs. Calculate industrial cluster density as the number of sport-related firms in each city normalized by city population. Calculate regional technology diversity using the number of distinct NACE industry codes per capita within each city. Generate the interaction term between industrial cluster density and regional technology diversity. Standardize continuous variables and conduct descriptive statistics, correlation analysis, and variance inflation factor (VIF) diagnostics. Estimate baseline models examining the independent associations of industrial cluster density and regional technology diversity with ROHC. Estimate interaction models introducing the product term between industrial cluster density and regional technology diversity. Estimate multilevel mixed-effects models with firms nested within cities and cluster standard errors at the city level. Conduct robustness analyses using alternative specifications, different analytical levels, and firm-age subsamples. Replicate reported tables and figures using the same model specifications and variable construction procedures. Because Orbis is a subscription-based proprietary database, the raw firm-level data cannot be publicly redistributed. However, the variable definitions, model specifications, and analytical procedures reported in the manuscript allow researchers with equivalent data access to reproduce the analysis.
Institutions
- Liaoning UniversityLiaoning, Shenyang