Robot arm
Robot arm
Robot adoption has skyrocketed in China in the last decade. New research finds that this exposure has led to a decline in employment and wages, influencing workers’ training and retirement decisions. How can developing countries prepare themselves for the artificial intelligence revolution?
Historically, major technological revolutions—from mechanisation during the Industrial Revolution to the rise of information and communication technologies—have profoundly reshaped labour markets. While some workers and regions benefited, others faced job displacement and wage stagnation, particularly those engaged in routine or easily codifiable tasks (Autor et al. 2003). The current wave of automation, driven by advancements in robotics and artificial intelligence, is often described as both faster and more pervasive, with the potential to affect a broader range of occupations and geographies (Susskind 2020, Acemoglu and Johnson 2023). Unlike previous transformations that unfolded gradually over decades, the speed and scale of current innovations present urgent challenges for policymakers and workers alike (Brynjolfsson and McAfee 2014).
The impact of robots on global labour markets
Global studies have attempted to quantify automation’s potential impact. The risk of automation varies significantly across countries due to differences in occupational structures, tasks, and institutions (Arntz et al. 2016, Gihleb et al. 2022, 2023). According to McKinsey Global Institute (2017), up to 800 million workers worldwide could be displaced by automation by 2030, particularly in developing economies where labour-saving technologies are rapidly diffusing. The International Labour Organization has raised concerns about job polarisation and social exclusion in the Global South, noting that younger workers and women may be disproportionately affected (ILO 2016). Nedelkoska and Quintini (2018) further show that automation risk correlates with low educational attainment and limited digital skills. Current generative artificial intelligence (AI) and other technologies have the potential to automate work activities that absorb 60-70% of employees’ time today (McKinsey Global Institute 2023).
In emerging economies, where formal education and reskilling systems are weaker, the risk of exclusion or displacement is higher, especially for youth and informal workers. Without adequate job creation, automation, digital technologies, and labour-saving innovations, global inequality could exacerbate further. As a result, developing countries may encounter novel policy dilemmas and significant economic trade-offs, notably between the productivity enhancements brought by automation and potential increases in economic inequality and social unrest. Consequently, the impacts of robotisation in emerging economies are expected to surpass those observed thus far in developed nations (Schlogl and Sumner 2018).
Previous research has already pointed out the risks associated with premature deindustrialisation, highlighting that automation may disrupt income convergence and limit developing countries' capacity to leverage their lower labour costs for economic growth. However, only few studies have directly assessed how workers in emerging markets are adapting or how their labour market trajectories are evolving in response to robotic automation. Addressing this empirical gap is crucial to developing equitable policies that harness productivity gains while minimising social dislocation. There is limited understanding of the specific impacts of industrial robots within developing economies.
Robot adoption has skyrocketed in China, driven by the automative and electronic sectors
In Giuntella et al. (2025), we analyse the effects of industrial robot exposure on the labour market in China—a country that has rapidly expanded investments in automation. Since President Xi Jinping's 2014 call for a ‘robot revolution’, China has aggressively pursued robotics and automation, committing substantial financial resources in its latest Five-Year Plan (2016-2020). Provinces such as Guangdong have allocated massive funds toward industrial robots and automation technologies. Research has shown that state-owned and politically connected firms disproportionately benefit from these subsidies (Cheng 2019).
China’s ambition is to position itself as a global leader in robotics, competing directly with established players such as Germany, Japan, and US. Indeed, since 2013, China has become the world's largest market for industrial robots in terms of annual purchases and total robots installed, though per capita levels remain lower than advanced economies. Automation is viewed as a crucial strategy to address rising labour costs, demographic aging, and intensified global competition. Nonetheless, the implications for millions of Chinese workers could be severe, as approximately 77% of Chinese jobs face high susceptibility to automation (Frey and Osborne 2017, Manyika 2017). Prominent cases such as Foxconn's replacing over 400,000 jobs with robots, and Zhejiang Province replacing two million workers with robots underscore these risks.
Figure 1 illustrates the incredible growth in robot adoption in China over the last few years, driven by the automotive and electronics sectors (Figure 2). Previous research has analysed the factors explaining the variation in robot adoption across Chinese firms, documenting how innovation subsidies are preferentially allocated to state-owned and politically connected firms (Cheng et al. 2019). These studies documented the adoption of robots by Chinese manufacturers using aggregate industry-level and firm-level data (Cheng et al. 2019), examined how robot adoption may create skill-biased development in firms’ employment structure (Tang et al. 2021), and analysed the effects of labour costs on the adoption of robots (Fan et al. 2021). However, the mechanisms behind these adjustments have received less attention, particularly in emerging economies.
Figure 1: The growth of robots in China
The growth of robots in China
Note: Data is drawn from the International Federation of Robotics.
Figure 2: Robots penetration by sector
Robots penetration by sector
Note: Data is drawn from the International Federation of Robotics.
Robot exposure has had negative effects on employment and wages in China
Our study leverages data from the International Federation of Robotics (IFR), combined with longitudinal individual data from the China Family Panel Studies (CFPS) (2010-2016), to analyse robot exposure effects. Adapting the identification approach from Acemoglu and Restrepo (2020) to a Chinese context, we document significant negative impacts on employment and wages. Specifically, we find that a one standard deviation increase in robot exposure reduces employment probabilities by 5 percentage points (-6.7% relative to the mean), increases exits from the labour market by 1 percentage point (+12.6%), and raises unemployment reporting by 4 percentage points (0.15 standard deviations). Hourly wages decline by about 8%, although total annual earnings remain unaffected as exposed workers tend to extend their working hours by roughly 8%. These effects are notably pronounced among low-skilled, male, prime-age, and older workers.
A key contribution of our study is examining how individuals respond dynamically to automation, revealing a notable increase in early retirement among older workers, while younger workers tend to seek technical or vocational training to remain competitive. Using CFPS data on training and retirement decisions, we find that robot exposure significantly increases participation in technical or work-related training—particularly among younger workers aged 16–44—while having no effect on ideological or political training. In contrast, robot exposure increases the likelihood of early retirement, especially among older workers: a one standard deviation rise in robot exposure raises early retirement by 0.2 percentage points, or about 50% relative to the mean. Overall, these findings suggest that workers adjust to automation either by upgrading their skills or exiting the labour force altogether.
Implications for the role of robots in labour policy
Our findings contribute directly to the emerging research predominantly focused on advanced economies. Few studies have explored the direct effects of robot exposure in emerging markets, particularly China. Existing research predominantly analyses either firm-level robot adoption and its skill-biased implications for employment structures, or the indirect effects of foreign robot adoption through trade and offshoring channels (Cheng et al. 2019, Fan 2021, Kugler et al. 2020). By employing longitudinal microdata on Chinese workers, our research directly measures individual labour market outcomes and adjustments, such as training participation and retirement decisions, providing valuable insights into the broader labour market responses to robot-induced disruptions.
Future research should focus on whether exposure to robots is affecting educational and career choices of young adults in developing economies so far characterised by heavy specialisation in manufacturing industries. Whether long-run productivity gains translate to employment growth, or how the labour market effects of automation and digitisation shape political outcomes are important questions that require further scientific investigation.
References
Acemoglu, D, and S Johnson (2023), "Power and progress: Our thousand-year struggle over technology and prosperity."
Acemoglu, D, and P Restrepo (2020), “Robots and jobs: Evidence from US labor markets,” Journal of Political Economy, 128(6): 2188–2244.
Arntz, M, T Gregory, and U Zierahn (2016), “The risk of automation for jobs in OECD countries: A comparative analysis,” OECD.
Autor, D H, F Levy, and R J Murnane (2003), “The skill content of recent technological change: An empirical exploration,” The Quarterly Journal of Economics, 118(4): 1279–1333.
Brynjolfsson, E, and A McAfee (2014), "The second machine age: Work, progress, and prosperity in a time of brilliant technologies," WW Norton & Company.
Cheng, H, R Jia, D Li, and H Li (2019), “The rise of robots in China,” Journal of Economic Perspectives, 33(2): 71–88.
Fan, H, Y Hu, and L Tang (2021), “Labor costs and the adoption of robots in China,” Journal of Economic Behavior & Organization, 186: 608–631.
Frey, C B, and M A Osborne (2017), “The future of employment: How susceptible are jobs to computerisation?” Technological Forecasting and Social Change, 114: 254–280.
Gihleb, R, O Giuntella, L Stella, and T Wang (2022), “Industrial robots, workers’ safety, and health,” Labour Economics, 78: 102205.
Gihleb, R, O Giuntella, L Stella, and T Wang (2023), “Keeping workers safe in the automation revolution,” The Brookings Institution.
Giuntella, O, Y Lu, and T Wang (2025), “How do workers adjust to robots? Evidence from China,” The Economic Journal, 135(666): 637–652.
Graetz, G, and G Michaels (2018), “Robots at work,” Review of Economics and Statistics, 100(5): 753–768.
International Labour Organization (ILO) (2016), "ASEAN in transformation: How technology is changing jobs and enterprises."
Kugler, A D, L Kugler, M Morales, and S Saavedra (2020), “US robots and their impacts in the tropics: Evidence from Colombian labor markets,” Unpublished manuscript.
Manyika, J, M Chui, M Miremadi, J Bughin, K George, P Willmott, and M Dewhurst (2017), “A future that works: AI, automation, employment, and productivity,” McKinsey Global Institute.
McKinsey Global Institute (2017), "Jobs lost, jobs gained: Workforce transitions in a time of automation."
McKinsey Global Institute (2023), "The economic potential of generative AI: The next productivity frontier."
Nedelkoska, L, and G Quintini (2018), "Automation, skills use and training," OECD.
Schlogl, L, and A Sumner (2018), “The rise of the robot reserve army: Automation and the future of economic development, work, and wages in developing countries,” Center for Global Development.
Susskind, D (2020), "A world without work: Technology, automation and how we should respond," Penguin UK.
Tang, C, K Huang, and Q Liu (2021), “Robots and skill-biased development in employment structure: Evidence from China,” Economics Letters, 205: 109960.