Spatial Dependencies in the Relationship between Automation and Migrant Worker Employment: Evidence from Chinese Cities
Description
This dataset includes all the data used to replicate: "Spatial Dependencies in the Relationship between Automation and Migrant Worker Employment: Evidence from Chinese Cities", published in Economic Modelling, 2026. The empirical content of the paper comprises eight tables and two appendices, along with four charts, each corresponding to the empirical analysis. The data on migrant workers is sourced from the China Migrant Dynamic Monitoring Survey (CMDS) organised by the National Health Commission of China. The robot data comes from the dataset of China’s industrial robots by industry, published by the International Federation of Robotics (IFR). Data on the number of employees across sectors in China and in each city are sourced from the China Statistical Yearbook and the China City Statistical Yearbook. The data for control variables are sourced from the China City Statistical Yearbook. The latitude and longitude coordinates, as well as the city's adjacency relationships, are available on the National Geographic Information System website (http://bzdt.ch.mnr.gov.cn/). This study examines the impact of industrial automation on rural migrant employment in China, with a focus on spatial spillover effects. It provides new insights into how automation influences labor markets beyond local boundaries. Using city-level data from 2011 to 2018 and industrial robot adoption as a measure of automation, we find that automation significantly reduces local rural migrant employment while generating positive spillover effects in neighbouring cities. These effects vary by migrants’ skills, tasks, industries, migration types, age, and marital status. Mechanism analyses indicate that automation promotes high-tech enterprise clustering and skill upgrading, creating skill premiums and labour outflows. At the same time, it also strengthens industrial links and structural similarity across neighboring cities. These effects are conducive to positive spillovers. The findings inform inter-regional policies aimed at stabilising rural migrant employment and well-being amid ongoing technological advancements. Keywords: Automation, Artificial Intelligence; Industrial robots; Migration; China; Spatial Econometrics; Spillover effects; Rural Economy
Files
Steps to reproduce
The empirical content comprises eight tables and two appendices, along with four charts, each corresponding to the empirical analysis in the paper. The data on migrant workers is sourced from the China Migrant Dynamic Monitoring Survey (CMDS) organised by the National Health Commission of China. The robot data comes from the dataset of China’s industrial robots by industry, published by the International Federation of Robotics (IFR). Data on the number of employees across sectors in China and in each city are sourced from the China Statistical Yearbook and the China City Statistical Yearbook. The data for control variables are sourced from the China City Statistical Yearbook. The latitude and longitude coordinates, as well as the city's adjacency relationships, are available on the National Geographic Information System website (http://bzdt.ch.mnr.gov.cn/). 1. Before executing the .do-file please download the following user-written Stata commands (using: ssc install [name of command] or findit [command]): spatwmat, xsmle, xtmoran, splagvar, spatdiag, logout 2. The replication folder contains all the codes for the empirical procedure in this paper. For Table 1, please run the code below: # Bookmark 1 # Table 1 Descriptive Statistics. For Table 2, please run the code below: # Bookmark 2 # Table 2 Global Moran's I index. For Figure 1, please run the code below: # Bookmark 3 # Figure 1. Moran's I scatter plot of the density of industrial robots and the employment of rural migrant workers For Tables 3 and 4, please run the code below: # Bookmark # Table 3 Baseline Results and Table 4 Spatial Effect Decomposition. For Table 5, please run the code below: # Bookmark # Table 5 Robustness Analyses. For Table 6, please run the code below: # Bookmark # Table 6 Heterogeneity Analyses I For Table 7, please run the code below: # Bookmark # Table 7 Heterogeneity Analyses II. For Table 8, please run the code below: # Bookmark # Table 8 Transmission Mechanisms. For Table A1, please run the code below: # Bookmark # Table A1. Spatial Panel Model Selection For Table A2, please run the code below: # Bookmark # Table A2. Descriptive Statistics for the Different Spatial Weight Matrices