Robots Don't Pay Taxes: Deindustrialization and Fiscal Decline

Abstract

Deindustrialization has fundamentally reshaped the economic geography of the United States. Between 1993 and 2007 alone, increasing automation—the use of industrial robots to perform tasks done by human workers—led to the loss of upwards of 750,000 jobs, primarily in the industrial Midwest and Northeast. Prior research demonstrates the social consequences of manufacturing’s decline extend beyond its impact on workers to undermine the health and economic prospects of entire communities. But through which mechanisms? This study examines the impact of increasing automation on local government finance. Exploiting spatial variation in the adoption of industrial robots, we find each additional robot per 1,000 workers is associated with a 10 percent relative decline in local government own-source revenues, a decline that is only partially offset by intergovernmental transfers. Moreover, we find that each robot per 1,000 workers leads to an 8 percent decline in K-12 education spending and a 30 percent decline in health spending. Our findings provide direct evidence of an oft-theorized but rarely examined mechanism through which place directly shapes life chances. We use our theoretical framework to motivate a Fiscal Sociology of place that centers the role of fiscal structures in the production of place-based inequalities.