AI Insight
This study analyzed 2,446 worker-year observations from China (2012-2018) to examine how AI exposure affects workers' occupational mobility and job quality. The research found that AI exposure increases job switching, but the effects differ by type: labor-augmenting AI (which enhances worker capabilities) promotes transitions across broader occupational categories and better job outcomes, while labor-saving AI (which replaces workers) creates barriers to major career transitions, particularly for less-educated workers. Workers exposed to labor-saving AI in their original jobs experienced longer working hours and lower skill-match satisfaction after switching occupations.
Why it matters
The findings demonstrate that different types of AI technology have distinct effects on workers' career trajectories, with labor-saving automation creating particular challenges for vulnerable workers. This suggests policymakers and organizations should provide targeted reskilling programs and career support, especially for less-educated workers in occupations with high labor-saving AI exposure.
Understand the Science
This study examines how exposure to artificial intelligence (AI) shapes Chinese workers’ occupational mobility and how the consequences differ across AI’s labor-saving and labor-augmenting channels. Using 2,446 worker-year observations from the China Labor Force Dynamics Survey 2012–2018, we measure occupation-level AI exposure as the semantic similarity between Chinese task descriptions and AI patent texts and decompose it along routine and non-routine task lines. Causal identification uses an instrumental variable that interacts 2010 baseline occupation-level exposure with lagged cumulative U.S. patent stocks from the USPTO AI Patent Dataset. The 2SLS estimates yield three findings. First, a one-log-unit increase in total AI exposure raises minor-group switching by 3.6 percentage points and intermediate-group switching by 3.4 points, with no significant effect at the major-group level. In the decomposed estimates, only the labor-augmenting component has significant positive coefficients beyond the minor-group margin, reaching 4.6 points at the intermediate and 6.9 points at the major-group distance. The labor-saving coefficients are statistically indistinguishable from zero in the full-sample switching-incidence IV estimates. The short-distance response is concentrated among women, and among less-educated workers labor-saving exposure predicts lower and labor-augmenting exposure higher major-group mobility, suggesting a labor-saving-related barrier to long-distance transitions. Second, among switchers, labor-augmenting exposure exhibits positive but attenuated origin–destination sorting in all three switching samples. Labor-saving sorting is positive in the minor-group sample, statistically indistinguishable from zero in the intermediate-group sample, and negative in the major-group sample. Third, among switchers, greater labor-saving exposure in the origin occupation is associated with longer hours in the minor- and intermediate-group samples and with lower skill-match satisfaction in all three samples, whereas favorable associations with origin labor-augmenting exposure are confined to specific samples. The study extends organizational-psychology research by linking instrumented AI exposure to occupational switching as an observable career-adaptation behavior and to the fit and demand conditions of destination jobs. The findings suggest targeted reskilling and organizational support for less-educated workers in occupations with high labor-saving exposure.