The Measurement of Total Factor Productivity in High-Tech Industry—Stochastic Frontier Model Based on Translog

在创新驱动发展战略背景下，本文基于随机前沿分析框架构建超越对数生产函数模型，采用中国24个省(市、区) 2013~2022年面板数据，对高技术产业TFP增长效率及其源泉展开实证研究。结果表明：研究期内高技术产业TFP年均增长率达11.9%，呈现显著提升趋势，但其增长主要依赖劳动、资本与技术要素的协同投入驱动；技术进步对TFP增长的贡献率达65%以上，构成增长核心动力源；资源配置效率的持续失衡显著抑制了技术进步效能的充分释放，形成“要素投入驱动增强–配置效率拖累加剧”的双重特征；自然间断点分级法显示，省域间TFP差异呈现东部创新密集区与中西部追赶区的梯度分布格局。由此提出相应政策建议。Under the background of innovation-driven development strategy, this paper constructs the translog production function model based on the stochastic frontier analysis framework, and uses the panel data of 24 provinces (municipalities, autonomous regions) in China from 2013 to 2022 to carry out empirical research on the TFP growth efficiency and its sources in high-tech industries. The results show that: during the study period, the average annual growth rate of TFP in high-tech industry is 11.9%, showing a significant upward trend, but its growth is mainly driven by the synergistic input of labor, capital and technology factors; Technological progress contributed more than 65% to TFP growth, constituting the core driving force of growth. The continuous imbalance of resource allocation efficiency significantly inhibits the full release of technological progress efficiency, forming the dual characteristics of “enhanced factor input drive and intensified drag of allocation efficiency”; The classification method of natural discontinuity points shows that the TFP difference between provinces shows a gradient distribution pattern in the eastern innovation-intensive areas and the catch-up areas in the central and western regions. Therefore, the corresponding policy recommendations are put forward.