摘要
本文匹配了2009—2018年中国人工智能专利数据库和中关村企业调查数据库,聚焦于考察人工智能对包括服务业在内的企业就业的影响。研究发现:人工智能显著促进就业,这一研究结论在考虑内生性问题及稳健性检验后依然成立;微观机制方面,人工智能可以通过提升企业市场占有率进而提升企业就业水平。拓展分析表明,人工智能提升了高学历员工、年轻员工、新员工占比;人工智能的就业促进效应在不同细分技术领域和不同类型企业间存在异质性:相较于专家系统及机器人,计算机视觉、机器学习、自然语言处理的就业促进效应更明显;人工智能的就业促进效应在企业规模较大、出口规模较大,以及国有和集体所有的企业中更小。本文基于专利数据对企业人工智能技术水平进行测度,将研究样本拓展至包含服务业企业,从企业市场规模扩大的角度拓展了人工智能影响就业的微观机理,为企业实施人工智能发展战略、政府制定人工智能政策提供了经验证据。
As a strategic emerging technology in the digital economy,Artificial Intelligence(AI)is increasingly being integrated with industries,promoting enterprise productivity and contributing to the economic growth.Meanwhile,the possibility of AI replacing jobs raises concerns about its potential systemic impact on the labor market.Existing literature concerning employment effect of AI focused mainly on developed economies and obtained controversial findings.The impact of AI on developing countries,especially on China's labor market,remains to be explored.Compared with developed countries,China has a significant working-age population and its labor structure is still dominated by low-and medium-skilled labor.Therefore,dramatic advances in AI may have a stronger impact on employment in China.This paper constructs indicators of AI from the perspective of patent,attempting to study the impact of AI on employment in China through empirical analysis,and clarifies its possible mechanisms at the micro level.The results of this paper provide theoretical support for China to implement AI strategies,cope with the potential impact of AI on the labor market and achieve parallel economic growth and employment stability.This paper examines the impact of AI on employment at the micro-level by matching the AI patents and employment data of Zhongguancun enterprises from 2009 to 2018.The findings indicate that AI has a significantly positive effect on enterprise employment,which remains valid after considering the endogeneity problem and robustness tests.Micro-mechanism analysis shows that AI affects employment through the“market expansion effect.”Specifically,AI is conducive to improving enterprise innovation capacity and reducing production costs,which drives the expansion of market share,and leads to increased employment growth.The extended analysis further investigates the impact of AI on employment structure based on dimensions of skill,age and working age of labor.The results show that AI increases the proportion of highly educated,young,and new employees in enterprises.Heterogeneity analysis indicates that the effect of AI on employment differs among different types of AI technology fields and different types of enterprises.The promotion effect of computer vision,machine learning,and natural language processing on employment is greater than that of expert systems and robotics.The promotion effect of AI on employment decreases with enterprise size and export scale;and compared with private enterprises,the promotion effect of AI on employment is significantly weaker in state-owned and collective enterprises.This paper attempts to enrich existing literature from following three perspectives.Firstly,AI patent applications for enterprises are identified by keywords to evaluate their influence on employment in Zhongguancun.This approach provides a comprehensive and precise measurement of the overall AI technology level of enterprises,complementing existing literature.Secondly,this paper reveals the micromechanism of AI,which influences changes in employment,by examining the expansion of enterprises'market size based on the literature theory and micro-enterprise data.Thirdly,exploratory analysis is conducted by utilizing AI patent keyword classification and detailed employment structure information from survey data of Zhongguancun enterprises.On the one hand,this paper examines the heterogeneity of employment effects for five sub-technology fields of expert systems,robotics,computer vision,machine learning,and natural language processing.On the other hand,this paper investigates the impact of AI on employment structure in terms of skill,age and work experience,providing a more comprehensive discussion on the impact of AI on employment.
作者
尹志锋
曹爱家
郭家宝
郭冬梅
YIN Zhi-feng;CAO Ai-jia;GUO Jia-bao;GUO Dong-mei(School of Economics,Central University of Finance and Economics)
出处
《中国工业经济》
CSSCI
北大核心
2023年第5期137-154,共18页
China Industrial Economics
基金
国家社会科学基金重大项目“城乡融合与新发展格局战略联动的内在机理与实现路径研究”(批准号21&ZD085)
国家自然科学基金面上项目“中国非专利实施体的行为策略及其创新效应研究”(批准号72274231)。
关键词
人工智能
就业
专利
artificial intelligence
employment
patent
作者简介
尹志锋,中央财经大学经济学院副教授,经济学博士;曹爱家,中央财经大学经济学院博士研究生;郭家宝,中央财经大学经济学院博士研究生;通讯作者:郭冬梅,中央财经大学经济学院教授,博士生导师,理学博士,电子邮箱:guodongmeicufe@163.com。
