COVID-19 Impact on County-Level Employment in the United States

Abstract

The coronavirus disease 2019 (COVID-19) pandemic suddenly shocked the United States at a time when the economy was operating at full employment (Dubina, 2017). The economic cost of the COVID-19 pandemic and efforts to contain it was severe with the United States national unemployment rate rising 0.9 percentage points in March 2020 and 10.3 percentage points in April 2020 (BLS, 2020). During this period, employment fell across all industries with the greatest losses in leisure and hospitality. Chetty (2020) suggests that the severe decline in economic activity during this period was primarily driven by health concerns in the United States. The health consequences of COVID-19 are highly variable; Maher (2020) found that most of the cases are considered mild or moderate with 20 percent of the reported cases becoming severe. Of those hospitalized, 75 percent have an underlying medical condition (Peters, 2020). The analysis I performed shows that the 12-month percent change in employment for counties with above average rates of COVID-19 risk factors (e.g., higher population density, older populations, etc.) was less severe than counties with below average rates of COVID-19 risk factors. For example, counties with a percent of population 65 years or older that is above the national average experienced an average decline in employment of 5.2% while counties with a younger population, on average, experienced an average decline in employment of 5.9%. The aim of this paper is to fill a gap in the literature by evaluating if concentrations of COVID-19 risks in county-level populations across the United States are significant predictors of the 12-month percent change in employment. This paper also seeks to understand if the 12-month percent change in employment is better predicted by NPIs, COVID-19 Cases, or concentrations of COVID-19 risks in county-level populations. These goals are accomplished by showing that concentrations of COVID-19 risks in county-level populations are significant predictors of the 12-month percent change in employment; however, it is shown that the effect of concentrations of COVID-19 risks on employment are small. Furthermore, the findings in this paper show that the severe decline in economic activity is not primarily driven by health concerns; rather, Non-Pharmaceutical Intervention (NPI) adoption is the strongest predictor of the 12-month percent change in employment. This is accomplished by estimating the sensitivity of employment to concentrations of COVID-19 risks in county-level populations across the United States during the COVID-19 era, while controlling for Non-Pharmaceutical Intervention (NPI) adoption, overall economic activity, concentrations in leisure and hospitality employment, COVID-19 infection rates, and rurality as measured by Rural Urban Continuum Codes (RUCCs).