A New Way to Quantify County-Level Deindustrialization in the United States

• In Event: U.S. Trade Politics

Abstract

American manufacturing jobs have been decreasing over the past few decades, but at the county level, deindustrialization patterns are more variable. While each county’s deindustrialization path is unique, there are archetypes among them. Traditionally, existing research has used a rate of change over some time period to measure county-level deindustrialization trends. This misses some nuance; the trajectory of job loss over time is not always linear. I use a novel time series clustering method to categorize communities, evaluating the trends in manufacturing jobs over 51 years for the 2,316 counties for which data are available. I find that six distinct clusters represent the major deindustrialization trajectories. While manufacturing jobs as a proportion of all jobs have decreased in most counties, this is not true of all counties, and declines have been quite sharp in some counties, giving them less time to transition to a post-industrial economy.

Author

• Emily Lucinda Kurtz