Working Paper: NBER ID: w27999
Authors: Jeffrey Clemens; Drew McNichols; Joseph J. Sabia
Abstract: The long-run costs and benefits of social insurance expansions may not be realized until a program has been in place through a cycle of boom, bust, and recovery. In the case of the Affordable Care Act (ACA), the arrival of the program's inaugural bust and recovery have been hastened by the COVID-19 pandemic. In this context, our analysis begins by developing two facts. First, during the pre-pandemic boom, we show that the ACA's effects had largely stabilized by 2016. Second, we develop a new fact involving variations in the ACA's effects across industries. Specifically, we show that the ACA’s effects differed dramatically across industries with lower versus higher levels of pre-ACA insurance coverage, and that this difference cannot be explained by differences in workers’ incomes or other observable characteristics, nor by geographic differences in pre-ACA uninsured rates. Finally, we set the stage for pre-committed analyses of the ACA's effects over the remainder of the current cycle of boom, bust, and recovery. In so doing, we seek to advance the use of pre-committed research designs in observational settings.
Keywords: Affordable Care Act; Medicaid; COVID-19; Insurance Coverage; Social Insurance
JEL Codes: H51; H53; I13
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| ACA's Medicaid expansions (I18) | increases in Medicaid coverage (I18) |
| ACA's Medicaid expansions (I18) | decreases in the uninsured rate (I13) |
| low-education individuals in states with Medicaid expansions (I24) | smaller increases in probability of being uninsured during the pandemic (G52) |
| COVID-19 economic downturn (F44) | greater declines in employer insurance offerings in states with Medicaid expansions (G52) |
| COVID-19 economic downturn (F44) | greater declines in employer insurance offerings in heavily impacted industries with Medicaid expansions (I18) |