COVID and Crime: An Early Empirical Look

on research by

DAVID S. ABRAMS

Professor of Law, Business Economics, and Public Policy

During the height of the COVID-19 pandemic, David S. Abrams was already compiling and analyzing empirical data to better understand the impact of the crisis on crime. Abrams, one of the country’s leading young economists working in empirical law and economics, collected data from 25 large U.S. cities and published his findings in “COVID and Crime: An Empirical Look” in the Journal of Public Economics in which he reported “a widespread immediate drop in both criminal incidents and arrests most heavily pronounced among drug crimes, theft, residential burglaries, and most violent crimes.”

Homicides and shootings showed no decline while nonresidential burglaries and car thefts in most cities increased. Abrams concludes that this data suggests “that criminal activity was displaced to locations with fewer people.”

Specifically, Abrams’ research uncovered overall crime rate drops of at least 35% in Pittsburgh, New York City, San Francisco, Philadelphia, Washington, D.C., and Chicago. He also found that the decline appears to have preceded stay-athome (SAH) orders and coincided with a decline in individual mobility.

In the paper, Abrams begins by recounting the sparse contemporary literature about COVID-19 and crime and then looks to the early 1900s to explore the impact of prior pandemics on crime. He shares the findings of a 1919 Chicago Department of Health report, which found a 38% decline in crime rates in Chicago from 1917 to the same three-week period in 1918 during which there was a lockdown.

“This is astonishingly similar to the 35% overall decline in crime in Chicago in the first 4 weeks of the 2020 pandemic, relative to the same period for the prior 5 years,” he writes.

Abrams then shares some additional background on crime and the pandemic, noting that generally “[t]he crime rate is a function of available opportunities and expected penalties.” He observes that evidence suggests that police presence and the enforcement of certain laws decreased once the pandemic began, “which would tend to increase crime rates due to the decreased expected penalty.” Due to a drop in mobility and the resulting decline in expected penalty, Abrams writes, “one should expect to see an increase in property crimes like car theft, theft from vehicles, and non-residential burglaries.” Conversely, because people were spending more time at home, a decrease in residential burglaries should be expected.

Turning to economic conditions that may influence crime rate, Abrams writes that empirical studies support the theory that a higher unemployment rate impacts property crime the most. During the pandemic, however, a “tremendous government response” in the form of the Families First Coronavirus Response Act and the Coronavirus Aid, Relief, and Economic Security Act, suggests that unemployment was “unlikely to have had a substantial impact on crime,” though he still includes the unemployment rate in some specifications.

Abrams credits the Open Data Movement with making his research possible. He has also made the collected data available to the public at https://citycrimestats.com. Mobility data (data

on individual movement) comes from Google COVID-19 Community Mobility Reports, generated from people who have turned on Location History for their Google Account and use a mobile device.

Abrams calls the change in crime from the onset of the pandemic “large and sharp enough that the impact is clear from the time series.” He notes that the crime drop began between 10-14 days before widespread SAH orders, “and is almost coincident with mobility drop.” Abrams found that the drop in all types of crime was “substantial,” and that the crime rates then remained low for the seven weeks following the SAH orders.

Specifically, total reported crime fell quickly by 23.3%. Property crime dropped 19.3%, and overall, most types of property crime fell substantially except for non-residential burglary and car theft. Theft rates also declined substantially by 28.2%, and theft from vehicles dropped by 20.3%.

All violent crime rates dropped except for homicide. Simple assault showed the largest decline (33.3%), while robbery dropped by 20.2% and aggravated assault by 15.9%. Drug crimes saw a 65% decrease (though this regression included only the 12 cities that provided drug crime data).

“There were almost certainly fewer low-level public altercations, especially associated with alcohol, as many public establishments were closed,” writes Abrams.

Regarding the sharp decline in domestic violence and rape reports, Abrams concludes that the drops likely overstate true changes in crime because of both limited data and potential complications with reporting brought about by pandemic conditions such as close living quarters with abusers and closed schools.

Abrams then addresses a critical question: “to what extent do changes in observed crime reflect changes in the real level of crime, rather than changes in reporting of crime?” He maintains that “much of the crime change is not simply a reporting artifact.” One approach to the issue looks at changes in the share of reports by police. Only two cities in the data set provide the source of crime incident reports.

“The vast majority of crimes in both cities are reported by individuals other than police, with one main exception, drug crimes,” writes Abrams. “In Nashville, all crime types saw a modest increase in the share reported by police. This suggests that the decline in overall crime reports was likely not predominantly due to reporting changes, or that reporting dropped in very similar proportions by police and individuals.”

A second approach to analyzing the impact of reporting changes is based on geographic location. Many bars and restaurants closed under SAH orders, and if “changes in crime reports reflect an overall reporting decline, then there should be a consistent drop in crime that is independent of distance from bars.”

Instead, data showed that “[s]imple assaults, drug crimes, and robbery all dropped between 30 and 40 percent of prepandemic levels within 25 meters of the establishments.” These drops all fall as the distance from the establishments increase.

Abrams then considers arrests as a measurement of crime changes and finds that results from regressions using arrests reveal an even greater decline than criminal incidents.

“Drug arrests fell a massive 77% and property arrets by 34.8%,” he writes. “Violent crime arrests had the smallest decline, at 16.8%, slightly lower than the drop incidents. Together this evidence suggests that police resources were focused less on arrests for most crime types, except for violent.”

Abrams concludes with a stark reminder that the pandemic rages on and that “political leaders, law enforcement, as well as individuals will need to account for the changed circumstances as they make decisions for some time to come.” He expresses hope that his research can help inform such decisions as well as “help us better understand the factors that impact crime in normal times as well.”

www.davidsabrams.com

Overall Crime Incidents Per Capita

for 25 of the largest u.s. cities, 2015-2020