Changes in the association between county industrial composition and working-age mortality from 2000 to 2022

Rising working-age mortality in recent decades was partly due to declines in the manufacturing industry. Less is known about how working-age mortality was associated with concurrent changes in other industries and whether those associations fluctuated over time alongside exogenous shocks to labor markets. This study draws on the precarious work literature to assess temporally specific associations between industry composition and mortality, which likely operate through direct effects on workers and spillover effects on communities. We estimate associations between county-level industry composition and mortality among working-age adults (ages 25-64) within multiple subperiods capturing shocks to labor markets and mortality (e.g., China Shock, Great Recession, opioid epidemic, and COVID-19) across 2000 to 2022. We use Bayesian Hierarchical Modeling to account for spatial and temporal patterns, net of other county-level characteristics. Results corroborate and extend prior findings that shares of certain industries are associated with mortality rates. Results further show that the size and direction of the associations fluctuated over time for some industries and causes of death. Higher shares of agricultural employment were associated with lower mortality across time, especially in latter periods, while the mortality benefits associated with higher shares of manufacturing employment waned in the latter period. Higher shares of service industry employment predicted higher mortality, especially during the pandemic. Associations between shares of employment in the professional service and mining/construction industries and mortality were mixed. The findings underscore the need to broaden the focus beyond the manufacturing industry to better understand trends and disparities in working-age mortality.