Artificial Intelligence and the Demand for Redistribution: Panel Evidence
Preprint
- 29 July 2026
- preprint
- Published by Elsevier in SSRN Electronic Journal
Abstract
As artificial intelligence advances, must support for the technology come at the welfare state's expense? In a 14-wave Japanese panel spanning ChatGPT's arrival, we find no such trade-off, on two margins: beliefs and jobs. Within person, support for AI and support for redistribution move together, a link that weakens where success is read as earned, not lucky. Across occupations, the most exposed workers --- educated and higher-income, whom income-based self-interest predicts to resist redistribution --- moved toward redistribution relative to the least exposed as AI diffused, while their AI attitudes moved like everyone else's. Overall, these results suggest that voters bundle support for the technology with demand for compensation.Keywords
This publication has 33 references indexed in Scilit:
- AI-driven income inequality and preferences for redistributionEconomic Analysis and Policy, 2025
- Automation and the future of work: How rhetoric shapes the response in policy preferencesJournal of Economic Behavior & Organization, 2021
- Misattributed blame? Attitudes toward globalization in the age of automationPolitical Science Research and Methods, 2021
- Occupational, industry, and geographic exposure to artificial intelligence: A novel dataset and its potential usesStrategic Management Journal, 2021
- Doubly robust difference-in-differences estimatorsJournal of Econometrics, 2020
- Preferences for Redistribution2011
- Preferences for redistribution in the land of opportunitiesJournal of Public Economics, 2005
- Inequality, Social Insurance, and RedistributionAmerican Political Science Review, 2001
- An Asset Theory of Social Policy PreferencesAmerican Political Science Review, 2001
- Social preferences, self-interest, and the demand for redistributionJournal of Public Economics, 2001