Abstract：In recent years, China’s working-age population has started to decrease, resulting in a gradual rise in labor costs. To cope with this increase, Chinese companies have widely installed digital technologies such as computers, automated equipment, and industrial robots, leading to a shift from the demographic dividend to a technology dividend. Previous studies have suggested that labor will continue to flow from the traditional agricultural sector to the modern industrial sector due to higher wages in the modern economic sector. However, recent statistical data show that while rural labor is still flowing to cities, it is not concentrated in the urban industrial sector, raising questions about the destination of this rural labor force and its relevance to the popularity of robotics.
While there is a considerable amount of literature discussing the income distribution effect of labor mobility between urban and rural areas, there is limited research on the impact of digital technologies such as robots and artificial intelligence. As the Chinese economy is divided into urban and rural areas with differences in occupational categories, there is inevitably urban-rural heterogeneity in the impact of robotics on income distribution. To provide clear answers to these questions, this study verifies the difference in the impact of robots on urban and rural wages through empirical tests based on the task-based model framework.
The study proposes four propositions. Proposition 1 asserts that robot penetration may facilitate labor mobility from rural to urban areas by increasing the relative wages of manual versus routine tasks. This effect is most pronounced when the labor force flows completely between urban and rural areas, resulting in higher wages for rural labor. However, in cases where labor is completely immobile between urban and rural areas, robot penetration may instead lower rural labor wages. When the labor force flows incompletely between urban and rural areas, the overall impact on wages depends on the relative strength of the aforementioned effects. Proposition 2 posits that the intersectoral transfer of labor within cities may help mitigate the potentially negative impact of robot penetration. Proposition 3 suggests that the impact of robot penetration is primarily evident in urban areas, where it depends on the relative strength of the complementary effect on abstract tasks and the substitution effect on routine tasks. Finally, Proposition 4 predicts that the wages of low-skilled labor in cities are likely to be reduced as a result of robot penetration.
The study utilizes data from the China Family Panel Studies (CFPS), the International Federation of Robotics (IFR), and the China Stock Market & Accounting Research Database (CSMAR) to distinguish the difference in the benefit between urban and rural samples and to verify that the difference in coefficients between samples is significant using Fisher’s Permutation Test (FPT). We merge the above three data sets according to the industry and province of the respondents, and the final sample obtained contains 2014, 2016 and 2018.
Benchmark regression results show that robot penetration has a significantly negative effect on wages. This result is consistent with the findings observed by Acemoglu et al. (2018) for the United States and Faber et al. (2020) for Mexico, where the substitution effect of robot penetration is larger than the productivity effect. After distinguishing between urban and rural samples, we find that robot penetration has a greater negative impact on urban wages, but a non-significant impact on rural wages. This finding remains robust after a series of robustness tests. To mitigate the bias arising from endogeneity, we constructed instrumental variables using the number of robots in the Czech Republic, and the results prove that our findings are robust again. Mechanistic analysis shows that robot penetration facilitates labor flow from rural to urban areas and contributes to higher wages of mobile labor. Furthermore, it allows for labor migration from manufacturing to other service sectors, leading to changes in urban-rural income distribution. While robots replace repetitive labor in manufacturing and create a “crowding-out effect” on low-skilled labor, the mobility of labor across sectors can help alleviate the negative effects of robot penetration.
This study contributes to the existing literature by providing a systematic examination of the impact of robot penetration on wages in urban and rural areas of China using individual and household survey data. Additionally, the study provides a theoretical model of the impact of robot penetration on urban and rural wages exploiting the task-based model and uncovers potential mechanisms of robot penetration affecting urban and rural wages from the perspective of cross-regional mobility and cross-sectoral transfers. Contrary to the majority of studies that are pessimistic about the employment substitution effect of robots, this study argues that robot penetration can have a positive effect on rural labor when labor can move adequately between urban and rural areas.
马述忠 吴鹏 潘钢健. 机器人渗透对城乡收入影响的异质性研究——基于劳动力流动视角[J]. 浙江大学学报(人文社会科学版), 0, (): 1-.
Ma Shuzhong Wu Peng Pan Gangjian. Research on the Heterogeneity Effect of Robot Penetration on Urban and
Rural Income: Based on the Perspective of Labor Mobility
. JOURNAL OF ZHEJIANG UNIVERSITY, 0, (): 1-.
