工作论文
当前位置:首页 > 工作论文
人工智能技术对劳动就业存在涟漪效应吗?——来自中国行业层面的经验证据
阅读全文         下载全文
TitleDoes AI Create Ripple Effect on Employment?——Empirical Evidence from the China  
作者王林辉 董直庆 赵景  
AuthorWang Linhui,Dong Zhiqing and Zhao Jing  
作者单位华东师范大学经济与管理学部 
OrganizationEast China Normal University 
作者Emaillinhuiwang73@sina.com;dongzhiqing@sina.com; 
中文关键词人工智能技术 劳动就业 涟漪效应 异质性劳动 
Key Words Artificial Intelligence Technology; Employment; Ripple Effect; Heterogeneous Labor 
内容提要前沿文献开始研究人工智能技术对所属行业的就业替代作用,却普遍忽视人工智能技术的就业创造效应及其对上下游行业就业的涟漪效应,更未关注人工智能技术对异质性劳动就业的影响差异。本文构建人工智能技术任务创造与消灭模型,推演人工智能技术对本行业和上下游行业就业的影响,探究人工智能技术的就业涟漪效应;利用行业面板数据和投入产出表,检验人工智能技术涟漪效应的存在性及其传导机制,及其对本行业和上下游行业异质性劳动就业的影响差异。结果发现:(1)人工智能技术存在涟漪效应。人工智能技术通过岗位更迭和生产率机制作用于本行业,当人工智能技术催生出劳动具有绝对生产优势的新生产任务,岗位更迭效应会提高劳动需求,而生产率效应则有利于增加就业。人工智能技术对上下游行业就业通过技术溢出、中间品需求或替代及竞合机制,产生就业挤出和岗位创造效应,但存在行业差异。(2)异质性劳动力就业的涟漪效应差异明显。人工智能技术在本行业和下游行业往往易使中高技术劳动群体受益,而对上游行业的中高技能劳动就业挤出作用强于创造效应。前向联动通过技术溢出机制实现,后向联动效应则主要依赖中间品需求传导,且对下游行业中高技能劳动就业创造作用大于对上游行业的就业替代效应。 
AbstractFrontier literatures quantitatively studied the employment substitution effect of artificial intelligence technology(AI) on industries using it, but generally ignored the employment creation effect of AI and whether there is ripple effect on the employment of upstream and downstream industries, and paid no attention to the influence difference of AI on heterogeneous labor employment. This paper builds a task creation and elimination model of AI, theoretically deduces the impact of AI on the employment of AI-using industries and their upstream and downstream industries, and explores the ripple effect of AI on employment. Using the panel data and Input-Output Table of China during 2002 and 2015, this paper tests the existence and the mechanism of AI’s ripple effect on employment, and examines the difference of influence on heterogeneous labor employment among the AI-using industries and upstream and downstream industries. The results show that: (i) AI creates ripple effect. AI affects AI-using industrial employment through job alternation change and productivity mechanism. When AI creates new production tasks in which labor has absolute advantages, the job alternation change effect and productivity effect both will increase labor demand and employment. AI produces employment crowding out and job creation effects on the employment of upstream and downstream industries through technology spillover, intermediate product demand or replacement and competition and cooperation mechanism, but there are differences among industries. (ii) AI’s ripple effect on heterogeneous employment is obviously different. AI prefers to benefit the high and medium skill labor in AI-using industries and the downstream industries, and shows stronger crowding out effect than creation effect on high and medium skill labor in the upstream industry. The forward linkage is realized through the technology spillover mechanism, while the backward linkage mainly relies on the transmission of the intermediate goods demand, and its employment creation effect on the of high-skilled labor in the downstream industries is greater than the employment substitution effect in the upstream industries. 
文章编号WP1453 
登载时间2019-12-26 
  • 主管单位:中国社会科学院     主办单位:中国社会科学院经济研究所
  • 经济研究杂志社版权所有 未经允许 不得转载     京ICP备10211437号
  • 本网所登载文章仅代表作者观点 不代表本网观点或意见 常年法律顾问:陆康(重光律师事务所)
  • 国际标准刊号 ISSN 0577-9154      国内统一刊号 CN11-1081/F       国内邮发代号 2-251        国外代号 M16
  • 地址:北京市西城区阜外月坛北小街2号   100836
  • 电话/传真:010-68034153
  • 本刊微信公众号:erj_weixin