Skip to main content
main-content
Top

Hint

Swipe to navigate through the articles of this issue

23-09-2022

A synthetic control analysis of U.S. state level COVID-19 stay-at-home orders on new jobless claims

Authors: John Gibson, Xiaojin Sun

Published in: Journal of Economics and Finance

Login to get access
share
SHARE

Abstract

There is an ongoing debate regarding the economic consequences of public policies designed to curb public health crises, such as the COVID-19 pandemic. Many opponents of such policies claim that their economic costs may outweigh their health benefits. In this paper, we use synthetic control analysis to determine the impact of stay-at-home orders on weekly new jobless claims during the initial phase of the COVID-19 pandemic. Our analysis reveals that while new jobless claims spike following the stay-at-home orders, similar spikes are observed within our synthetic control. Specifically, we find that stay-at-home orders account for only 32 percent of the increase in new jobless claims, with the majority of the increase being driven by factors outside of the policy, such as the general spread of the virus and waning consumer confidence.
Footnotes
1
While estimating the health benefits of statewide SAH orders is outside the scope of the present paper, existing works in the literature have found that these policies carry large health benefits. For example, Friedson et al. (2020) find that California’s SAH order reduced COVID-19 cases by 125.5 to 219.7 per 100,000 individuals and lead to as many as 1,661 fewer COVID-19 deaths by April 20. Courtemanche et al. (2020) find that the state of Kentucky would have had more than ten fold the number of COVID-19 cases actually observed by April 25 if the state had relied on voluntary social distancing along.
 
2
Karpman et al. (2020) state that social isolation and physical distancing practices have led to a historic rise in unemployment insurance claims, the largest decline in retail sales on record, surging demand at food banks, and increased reports for delinquent rent and request for mortgage forbearance.
 
3
Difference-in-Differences (DiD) is one of the more frequently used methods in treatment evaluation studies. It involves a comparison between a treatment unit and a control unit before and after the treatment under a parallel trend assumption. The treatment effect is the difference between the observed outcome value and what the value would have been with parallel trends had there been no treatment. While the DiD model allows the outcome to be driven by unobserved confounders, it restricts the effects of those confounders to be time invariant (see the discussion in Abadie et al. 2010).
 
4
This yields 4 quarterly averages of jobless claims in 2018 and 2019, respectively, and the average of jobless claims in the first quarter of 2020 prior to the start date of an SAH order. We also try averaging the pre-treatment outcome biannually, annually, or over the entire pre-treatment period and our results are robust.
 
5
We choose to end our sample on April 25 because, since late April, many states started to reopen.
 
6
We present the graphs for New York and Washington only due to space limitations. All graphs are available from the authors upon request.
 
7
Among the 7 states without statewide SAH orders, three counties in Utah (Davis County with 352,000 people, Salt Lake County with 1.2 million people, and Summit County with 42,000 people) and one county in Wyoming (Jackson with 10,000 people) issued similar restrictions. As a robustness check, we exclude Utah and Wyoming from the donor pool and our results are robust. The robustness check results are available from the authors upon request.
 
Literature
go back to reference Abadie A (2021) Using synthetic controls: feasibility, data requirements, and methodological aspects. J Econ Lit 59(2):391–425 CrossRef Abadie A (2021) Using synthetic controls: feasibility, data requirements, and methodological aspects. J Econ Lit 59(2):391–425 CrossRef
go back to reference Abadie A, Diamond A, Hainmueller J (2010) Synthetic control methods for comparative case studies: Estimating the effect of California’s tobacco control program. J Am Stat Assoc 105(490):493–505 CrossRef Abadie A, Diamond A, Hainmueller J (2010) Synthetic control methods for comparative case studies: Estimating the effect of California’s tobacco control program. J Am Stat Assoc 105(490):493–505 CrossRef
go back to reference Abadie A, Gardeazabal J (2003) The economic costs of conflict: A case study of the Basque Country. Am Econ Rev 93(1):113–132 CrossRef Abadie A, Gardeazabal J (2003) The economic costs of conflict: A case study of the Basque Country. Am Econ Rev 93(1):113–132 CrossRef
go back to reference Baek C, McCrory PB, Messer TMui (2021) P Unemployment effects of stay-at-home orders: Evidence from high-frequency claims data. Rev Econ Stat 103(5):979–993 CrossRef Baek C, McCrory PB, Messer TMui (2021) P Unemployment effects of stay-at-home orders: Evidence from high-frequency claims data. Rev Econ Stat 103(5):979–993 CrossRef
go back to reference Luo S, Tsang KP (2020) China and world output impact of the Hubei lockdown during the coronavirus outbreak. Contemp Econ Pol 38(4):583–592 CrossRef Luo S, Tsang KP (2020) China and world output impact of the Hubei lockdown during the coronavirus outbreak. Contemp Econ Pol 38(4):583–592 CrossRef
go back to reference Partridge MD (1997) The dispersion of US state unemployment rates: The role of market and non-market equilibrium factors. Reg Stud 31(6):593–606 CrossRef Partridge MD (1997) The dispersion of US state unemployment rates: The role of market and non-market equilibrium factors. Reg Stud 31(6):593–606 CrossRef
Metadata
Title
A synthetic control analysis of U.S. state level COVID-19 stay-at-home orders on new jobless claims
Authors
John Gibson
Xiaojin Sun
Publication date
23-09-2022
Publisher
Springer US
Published in
Journal of Economics and Finance
Print ISSN: 1055-0925
Electronic ISSN: 1938-9744
DOI
https://doi.org/10.1007/s12197-022-09604-9

Premium Partner