智能化对中国劳动力市场的影响——基于就业广度和强度的分析

随着自动化、智能化技术的不断发展,越来越多的工作岗位可能被机器和人工智能所替代。本文将美国劳工部标准职业代码与中国职业代码相匹配,基于Frey and Osborne(2017)对美国各种职业被智能化替代概率的估计结果,估算了中国各职业被智能化替代的概率,并在此基础上计算了城市层面的被替代指标。接下来,利用多个年份的人口普查和家庭调查微观数据以及欧盟的机器人使用数据,本文在城市层面和个人层面估计了智能化对就业广度(就业人数)和就业强度(工作时长)的影响。研究发现,智能化对中国劳动就业产生了明显的替代作用,一方面减少了就业人数的增长,另一方面却增加了在职劳动力的工作时间,分样本分析发现女性、低教育劳动者、大龄劳动者、移民等劳动力市场中相对脆弱的群体所受的冲击更大。

The Impact of Automation and Artificial Intelligence on China's Labor Market:Quantity and Intensity of Employment

Automation and artificial intelligence (AI) are major trends in the workplace that have significantly improved production efficiency. Through the Internet+, big data, and cloud computing, AI has sparked a global technological revolution that has changed the traditional social order. While automation and AI have a positive impact on economic growth, they also negatively affect many traditional occupations. Many jobs may be replaced by automation. Studies show that the decline in the employment and wages of low-and medium-skilled workers can be attributed to the application of automation and AI. The rapid development of the AI industry and the aging of the population have drawn attention to how automation and AI affect China's labor market. Unlike the previous three technological revolutions in which machines and equipment replaced manual labor, automation and AI are being integrated into the production process. This not only requires machines to approximate human dexterity but, more importantly, it means that machines are gradually developing cognitive abilities similar to those of a human. This transformation will have a significant impact on the labor market. In China, the impact of automation and AI will be more pronounced, partly because China is at the forefront of AI development, and partly because of its large population and labor-intensive industrial structure. However, little research (especially quantitative research) examines how China's labor market will be affected by automation and AI. This study estimates the effect of automation and AI on China's labor market and suggests relevant countermeasures. Based on Frey and Osborne's (2017) estimation of the probability of computerization for 702 detailed occupations, this paper estimates the probability that each occupation in China will be replaced by automation. Based on these estimates, we use data from latest Census and China Family Panel Studies to estimate the automation-substitution probability by city, and gauge the macro impact of automation on China's labor market. Finally, we empirically examine the effect of automation on labor market outcomes, such as employment at the city level and individual working hours. The results show that automation has a significant and negative impact on employment but a positive impact on working hours. The effect is larger among vulnerable groups in the labor market, such as women, those with a low level of education, the elderly, and migrants. The contributions of this paper are as follows. First, it shows how the labor market in developing countries is affected by automation. Second, it estimates the replacement probability for each occupation in China, and comprehensively assesses the possible substitution effect of automation. Third, it estimates the impact of automation on the quantity and intensity of employment and conducts a heterogeneity analysis using labor force characteristics, to provide a comprehensive assessment of the impact of automation. Finally, the paper uses data on EU robots as a proxy for AI in China to examine the impact of AI on the labor market from different perspectives. Although this study focuses on showing correlations rather than inferring causality, it is nonetheless informative about the impact of automation and AI on China's labor market, and has important policy implications. While China promotes the development of new technologies such as AI, it needs to address their potential negative impact on the labor market. First, this impact needs to be comprehensively assessed, because its effect will differ across industries and workers. Second, more attention needs to be paid to vulnerable groups in the labor market (e.g., women, low-educated workers, older workers, and migrants). Efforts to improve their labor skills and human capital through vocational training are needed to alleviate the negative impact of automation and AI on them. Finally, attention needs to be paid to the impact of automation and AI on the welfare of workers, particularly the polarization of income and social class. Technological progress needs to be harnessed to promote economic development while improving the welfare of workers and maintaining social equity.