期刊文献+

外部技术断供、工作任务与企业劳动力需求结构——基于Transformer架构大数据文本挖掘的微观证据 被引量:1

External Technological Supply Disruptions,Work Tasks and Labor Demand Structure of Enterprises:Micro-evidence from Big Data Text Mining Based on Transformer Architecture
原文传递
导出
摘要 探讨外部冲击如何影响中国企业劳动力需求结构,对防范化解就业领域重大风险具有重要意义。本文以美国对华技术断供制裁清单的典型特征为切入点,基于工作任务分析框架,构建一个包含进口技术投入的局部均衡模型,厘清了外部技术断供通过任务配置效应和任务替代效应影响中国企业劳动力需求结构的理论机理。在此基础之上,采用两类人工智能领域较为前沿的Transformer架构大语言模型,对互联网招聘大数据实施文本挖掘,实证检验外部技术断供对中国企业劳动力需求结构的影响及其作用机制。结果表明,整体上,外部技术断供通过正向任务配置效应,显著增加中国企业对非常规劳动力的相对雇佣需求,推动企业劳动力需求结构转型升级,然而,任务替代效应并未对企业劳动力需求产生明显影响。进一步分析发现,企业在供应链配置、多元化经营和政府补贴方面存在的差异,将对外部技术断供的劳动力需求效应产生显著的异质性影响。本文的研究为理解外部冲击的经济效应提供了新视角,并对百年变局下,如何缓解结构性就业矛盾,实现高质量充分就业提供政策启示。 Due to the interwoven great rejuvenation cause and complex international landscape,exploring how external technological supply disruptions(ETSDs)affect the labor demand structure of Chinese enterprises is of great significance for preventing and defusing major risks in the employment sector.This study starts from the demand side of the labor market,refers to the basic settings of the task trade model,incorporates imported technological inputs into the work task analysis framework,and constructs a partial equilibrium model by integrating the theoretical analysis methods of relevant studies on production networks and supply chain shocks.Using the iceberg cost of imported technological inputs,this study characterizes the intensity of ETSDs that enterprises face and clarifies their impact on the labor demand structure of Chinese enterprises and their action mechanism.Further,to address the data limitations in existing studies on the measurement of labor skill structure and demand structure of Chinese enterprises,based on the recruitment text information of big data on online recruitment of listed companies and a set of scientific and standardized occupational classification standards,this study introduces artificial intelligence and deep learning algorithms and uses large language models based on Transformer architecture to quantitatively identify the main work task requirements reflected in specific recruitment positions to achieve accurate classification of enterprises’recruitment demands and lay a solid data foundation for the rational measurement of the labor skill structure and demand structure of Chinese enterprises.Based on the above theoretical analysis and data measurement,taking Chinese enterprises listed on the US sanctions list as a quasi-natural experiment of ETSDs,this study conducts an empirical test by using the difference-in-differences(DID)model combined with the micro-panel data of Chinese A-share listed companies.The research findings reveal the following.First,overall,ETSDs significantly promote the transformation and upgrading of Chinese enterprises’labor demand structure.This conclusion remains valid after a series of robustness tests,including pre-trend parallelism tests,placebo tests,negative weight tests,heterogeneous robust estimator estimations,exclusion of concurrent shocks,control for internet selection bias,replacement of core variable measurement methods,control for inherent correlations among affected enterprises,and adjustments to model specifications.Second,the mechanism analysis based on the theoretical model reveals that ETSDs drive the transformation and upgrading of enterprises’labor demand structure by adjusting task allocation strategies and increasing hiring demand for domestic non-routine labor.On the one hand,such disruptions force enterprises to shift non-routine tasks originally executed under an open-mode framework to independent execution,expanding the scope of tasks performed by domestic non-routine labor and reducing reliance on imported technological inputs.On the other hand,affected by supply disruptions,the relative price of imported technological inputs rises,prompting enterprises to substitute relatively lower-cost non-routine labor for imported inputs—a substitution that amplifies hiring demand for non-routine labor and thus alters labor demand structures.Although supply disruptions-triggered relative price changes in intermediate goods may induce affected enterprises to substitute routine tasks for non-routine ones,the inhibitory effect of task substitution on Chinese enterprises’labor demand structure remains insignificant at the current stage.Third,further analysis reveals that the impact of ETSDs on labor demand structure transformation is more pronounced in enterprises with higher upstream supply chain concentration,lower product diversity,and higher government subsidies.Based on the above research conclusions,this study argues that efforts should be made to enhance the skill quality of workers;effectively optimize the supply structure of human capital in China;and further improve government fiscal,financial,and industrial policy subsidy tools to alleviate the negative impact of ETSDs and strengthen the policy support for the transformation and upgrading of enterprises’labor demand structure.
作者 张星民 张建清 刘小慧 吴新雨 ZHANG Xingmin;ZHANG Jianqing;LIU Xiaohui;WU Xinyu(School of Economics and Management,Wuhan University)
出处 《数量经济技术经济研究》 北大核心 2025年第9期72-93,共22页 Journal of Quantitative & Technological Economics
基金 国家社会科学基金项目(19BGJ033) 西南财经大学中央高校基本科研业务专项基金(JBK2304060)的资助。
关键词 外部技术断供 工作任务 劳动力需求结构 Transformer架构 大语言模型 External Technological Supply Disruptions Work Tasks Labor Demand Structure Transformer Architecture Large Language Models
  • 相关文献

参考文献27

二级参考文献406

共引文献1117

同被引文献28

引证文献1

二级引证文献1

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部