企业数字化转型程度的统计测度：以数字技术人才需求为视角

Abstract

This paper investigates the degree of digital transformation in Chinese enterprises from the perspective of labor input, focusing on firms' demand for digital talent. The study responds to a growing debate in the literature regarding how far Chinese enterprises, particularly small and medium-sized enterprises (SMEs), have progressed in their digital transformation. Existing research predominantly relies on data from listed firms and emphasizes production-side indicators, which may overstate transformation levels and misrepresent the broader corporate landscape. To address these limitations, this paper proposes a novel labor-oriented measurement strategy incorporating job posting data from 3339187 firms in China from 2014 to 2021, obtained from a leading online recruitment platform. By applying a dictionary-based text mining approach, supplemented with machine learning and large language models for robustness checks, the study constructs a multidimensional index of firm-level digital transformation across five key domains: artificial intelligence, big data, cloud computing, blockchain, and applied digital technologies. The methodology is grounded in the literature on digital economy accounting, particularly the dual-framework approach to measuring data assets through production and labor input. The paper argues that demand for digital talent reflects not only a firm's technological capability but also its actual deployment of digital infrastructure and transformation readiness. Compared with methods based on financial reports or policy exposure, the recruitment-based approach is more timely, granular, and inclusive of firms that are unlisted, newly established, or in emerging sectors. Empirical results reveal that as of 2021, only 43.2% of Chinese enterprises had engaged in digital transformation, significantly lower than estimates based on listed companies (over 80%). This highlights the persistent digital divide, especially in SMEs, and the need to re-evaluate assessments of China's digital economy progress. Further disaggregation shows notable regional and sectoral heterogeneity, with higher transformation rates in the eastern region and technology-intensive sectors, while the central, western, and northeastern regions, as well as traditional service industries, lag behind. In terms of international perspective, China's digitalization rate, particularly among SMEs, remains below that of advanced economies, where foundational technology development is more mature and institutional support for SME digitalization is more robust. The limited digital upgrading in Chinese SMEs is attributed not only to financing and talent constraints but also to low perceived returns, fragmented strategies, and weak inter-firm knowledge spillovers. The paper situates its findings within broader concerns about inclusive technological growth and structural transformation in developing economies. The study also evaluates the policy effectiveness of China's National Big Data Pilot Zones launched in 2016. Leveraging a difference-in-differences framework and a rich panel dataset, the paper identifies a significant positive causal effect of the policy on enterprise-level digital transformation. However, the impact is mainly reflected in enhancing applied digital technologies rather than advancing underlying technologies such as AI or blockchain. This asymmetry suggests that while policies may encourage digital adoption, they have limited influence on long-term innovation, particularly in resource-constrained SMEs. Mechanism analysis reveals that digital subsidies, infrastructure investment, and local digital ecosystems are critical channels through which the policy exerts effects. This research makes three major contributions. First, it introduces a new measurement approach rooted in firms' human capital demand, which overcomes 'false positive' and 'false negative' biases inherent in textual analysis of corporate reports. Second, it expands the empirical base of digital transformation research beyond listed firms by incorporating a comprehensive sample of SMEs, which account for the majority of employment and output in China. Third, it offers the first firm-level evaluation of national digital economy policy using a labor demand-based metric, contributing to the policy assessment literature. The paper concludes with several actionable policy recommendations. These include establishing a national database to monitor digital transformation progress, designing differentiated support schemes for SMEs, narrowing regional digital gaps through targeted infrastructure investment, encouraging foundational digital technology R&D to drive long-term competitiveness, and fostering university-industry collaboration to strengthen digital talent training and improve SMEs' ability to attract and retain skilled professionals.