An Inequitable Transition: Coal Plant Closures and Local Labor Outcomes by Educational Attainment

Abstract

This thesis examines how coal-fired generator retirements affect local labor market outcomes across educational attainment groups in the United States. Using generator-level retirement data from the Energy Information Administration and individual-level American Community Survey data from 2006 to 2023, I estimate fixed effects specifications to evaluate the effects of coal generator closures and coal capacity loss on unemployment, labor force participation, usual hours worked, and real wage income. I then extend the analysis by estimating heterogeneous effects across education groups and interacting coal capacity loss with changes in non-coal generation capacity. To test for endogeneity in baseline specifications, I further implement a two-stage least squares methodology that instruments for retirements using the count of coal generators aged fifty years or older.

I find that coal-fired generator retirements worsen local labor market outcomes in affected counties, increasing unemployment and reducing labor force participation and real wage income. These effects are heterogeneous across education groups. Unemployment and labor force participation effects are concentrated among individuals with a high school degree, while wage losses appear across multiple education categories. Instrumented estimates are substantially larger than baseline estimates, suggesting that conventional fixed effects models may understate the labor market costs of coal retirements. Overall, the findings show that the labor costs of the energy transition are economically significant and unevenly distributed across education levels.