Technological advancement challenges labor market
Spectators watch a smart welding robot perform at the 2019 World Robot Exhibition in Beijing. Photo: CFP
At present, a new wave of scientific and technological revolution and industrial transformation represented by artificial intelligence (AI) is on the rise. With the accumulation of big data, the innovation of theoretical algorithms, computing power’s enhancement and improved network infrastructure, technological progress has entered a new stage featuring smart technologies.
As automation driven by AI transforms scientific and technological progress into higher productivity and economic growth, it also has a profound impact on the labor market. In fact, automation technology has already become a key factor affecting changing global employment structures.
Many studies show that workers in conventional and low-skilled jobs are most affected by technological advances. Automation is likely to exacerbate unemployment, job polarization, and income inequality, notably by increasing the share of jobs toward the top and bottom ends of the market in terms of skills and income distribution, with a sharp decline in the share of jobs in the middle.
According to the International Federation of Robotics, the number of robots used on the Chinese mainland soared from 37,000 in 2009 to 780,000 in 2019. In 2019, 140,500 new robots were installed in China, accounting for nearly 40% of the world’s total increase. This means China has more robots per 1,000 manufacturing workers than developed countries such as the United Kingdom and France.
As the largest developing country, China has always upheld that “employment is pivotal to people’s wellbeing.” The impact of automation technologies such as robots on China’s labor market is an important academic and policy topic of discussion, which is of great significance for maintaining long-term stability of employment rates and preventing the exacerbation of income inequality.
Employment structure
In the face of automation technology’s rapid development, the wide application of industrial robots in the past decade has triggered concerns that workers will be soon replaced by intelligent machines. For the sake of analysis, we will divide China’s working-age population, based on the likelihood of being replaced by automation, into five categories: unconventional knowledge work, conventional knowledge work, unconventional operational work, conventional operational work, and the non-employed.
Based on the above classification of employment types, we investigated the dynamic changes in China’s employment structure using data from the national population census and the 1% population sample survey from 1990 to 2015. The results show that the proportion of conventional operational work in the workforce dropped sharply from 57% to 32%, whereas conventional knowledge work surged from 8% to 19%, and the non-employed increased from 16% to 31%. Meanwhile, the share of unconventional operational work remained almost unchanged.
This indicates that in the process of technological change, the labor force that moved from conventional operational work did not take on unconventional operational work, which is unlikely to be replaced by automation technology in large numbers, but mostly became unemployed or switched to conventional knowledge jobs.
In theory, if the increased share of non-employed people is mainly due to an increase in the working age population’s schooling time or retirement rate, that would translate into an increase in human capital or enjoyment of leisure, which would not cause social problems.
However, statistics show that the main reason for the continuous increase in the non-employed group’s proportion is unemployment or withdrawal from the labor market due to a lack of suitable jobs, which is closely related to the mismatch between employment structures and workers’ skills, caused by technological progress.
Judging from China’s employment structure trends from 1990 to 2015, there is still a large space for structural optimization in response to future technological progress. Over the same 25-year period, the United States experienced a decline in the share of conventional jobs and a slow rise in that of unconventional jobs. Specifically, the proportion of conventional knowledge jobs decreased from 19.6% to 16.1%, and the proportion of conventional operational jobs decreased from 21.0% to 15.1%. At the same time, the proportion of unconventional knowledge jobs increased from 24.7% to 28.2%, the proportion of unconventional operational jobs from 9.6% to 12.3%, and the proportion of non-employed people from 25.2% to 28.3%.
In comparison, the proportion of conventional jobs in China is still quite high, and the growth of unconventional jobs has been stagnant for a long time, making China vulnerable to technological progress represented by robots and AI.
Impact of robots
From 2009 to 2019, the application of industrial robots in China increased at an average annual rate of 36%. According to a report released by the Chinese Institute of Electronics at the World Robot Conference in 2019, China has been the largest market for the application of industrial robots for seven straight years. Its profound impact on China’s labor market cannot be underestimated.
Economic theory suggests that automation, such as robots, has two opposite effects on employment. The substitution effect of direct replacement of workers in jobs, as well as the productivity effect of technological progress improving production efficiency, could lead to an increase in demand for workers.
The productivity effect is mainly composed of two parts. One is the decrease of production costs and the improvement of product quality after the application of robots. As the industry’s market size expands, the demand for labor that has not been replaced or is complementary with robots will go up. The other component is a productivity spillover effect between industries that use robots and complementary industries. Therefore, industries that do not use robots may also see a decrease in costs and output expansion, thus an increase in labor demand.
In the end, the impact of robots on employment depends on which is dominant: the substitution effect or the productivity effect.
We assessed the impact of robots on employment, wages, and income inequality at the city level, taking into account differences in the industrial structure prior to the arrival of robots and robot usage’s technological feasibility in various industries. The results show that from 2009 to 2017, robots drove up China’s job opportunities on the whole, but the substitution effect of robots increased year by year, and negative impact of robots on workers’ wages has already appeared in the manufacturing industry. In addition, the impact of robots on the service industry is fully demonstrated in the employment of producer services in complementary industries.
As we can see, the impact of robots on employment and wages showcases heterogeneity among different labor groups, which is related to the degree of skill substitution of different groups and the elasticity of labor supply. Overall, negative impacts on low-income groups are more serious. Estimates of income inequality within cities show that the use of robotics significantly widens the income gap.
In the long run, the growth of robot application is of great value to alleviate labor supply shortages in China’s era of population aging, and is of great importance to improve labor productivity.
In the short term, however, young people and less-educated workers face a severe employment shock, while women and migrant populations suffer more substantial wage losses. For these lower-income groups, the rapid growth of robot application is more likely to exacerbate income inequality. Therefore, it is necessary to distinguish between different labor groups and formulate targeted policies in response to technological progress.
‘Technological unemployment’
In the process of economic structure upgrading in China, attention must be paid to alleviating the negative impact of automation technologies such as robots on workers.
To start with, educational content and methods need to transform with the change of automation technology to enhance the complementarity of labor skills with robots. To this end, the government needs to vigorously develop vocational education, through financial subsidies, tax relief and other ways, to mobilize the enthusiasm of enterprises to carry out skills training, so as to make workers’ career transitions smooth.
As automation technology matures at an accelerating rate, young people face a greater risk of “technical unemployment” in the future job market. In this light, focus should be shifted to R&D capacities, robot maintenance skills, or interpersonal communication and innovation which robots cannot replace, and this should be the toehold of education.
Second, through relevant fiscal, tax, and social security policies, the government needs to better protect low-income groups, and improve the market competition mechanism to promote work incentives for high-income groups. Going forward, social security coverage needs to be expanded, with special attention paid to key areas and groups vulnerable to technological progress.
At the same time, tax and fee cuts should be adopted to stimulate enterprises’ innovation vitality, reduce their burdens, uplift enterprises to become the main players in automation technology, and encourage them to expand their scale through transformation and upgrading to create more high-quality jobs.
Finally, flexible labor market policies should be in place to cope with short-term employment substitution risks, so as to promote sustainable development of talent dividends and high-quality economic growth. In the short term, the negative impact of unemployment on workers and their families can be mitigated by reducing working hours and promoting flexible employment. In the long term, it is crucial to remove the household registration system and ownership barriers to allow free movement of labor, ensure fair distribution of education resources, and develop lifelong learning habits among the public, as ways to maintain China’s edge in talent and in the fourth wave of industrial revolution.
Ge Peng and Zhao Zhong are from the School of Labor and Human Resources at Renmin University of China.
Edited by YANG XUE