Research: How calls for service changed after COVID-19 lockdowns
Study reviewed how 18 common types of calls changed in 10 large U.S. agencies
By Dr. Matt Ashby
The COVID-19 pandemic has changed policing in many ways. Tragically, the virus has claimed the lives of at least 50 officers across the nation, while cops have faced challenges in enforcing frequently changing lockdown restrictions.
I recently researched how 18 common types of call for service changed in 10 large U.S. agencies during the weeks immediately after cities and states began to lockdown in March 2020: Baltimore, Cincinnati, Los Angeles, New Orleans, Phoenix, San Diego, San Jose, Seattle, Sonoma County (California) and St. Petersburg. These identified changes might help other agencies to develop strategies in case the coronavirus impacts their jurisdictions.
Every city and police agency is different, so it is not surprising that changes in calls varied between the cities in the study. Nevertheless, the total number of calls received from mid-March onward (when the first stay-at-home orders were issued) was lower than would have been expected without COVID-19. This overall drop concealed different trends in different call types, grouped into four categories:
- Order maintenance
Calls to suspected crimes (assault, burglary, domestic violence, robbery, shots fired and vehicle theft) remained largely as expected in most cities, with some exceptions.
Calls to assaults dropped by up to one-third in three cities (Baltimore, Cincinnati and Seattle), while shots-fired calls increased in Cincinnati (but only for the first two weeks of lockdown) and Phoenix.
While decreases in assaults were not seen everywhere, they are consistent with many of the places at which assaults happen (particularly bars) being closed during the lockdown. The possible increase in shots fired is concerning, but it is worth noting those calls did not increase in six of the other cities in the study.
Other research has suggested that stay-at-home orders may have left domestic-violence victims trapped at home with their abusers. In most cities, any increase in domestic assaults did not lead to an increase in calls to those incidents, possibly because some victims were unable to call for help since they could not leave their homes. Agencies might want to consider sending officers to check the welfare of recent repeat domestic-abuse victims during any future lockdowns since we know previous victims are much more likely to be victims again in the future.
Public order calls
As well as crimes, police respond to a variety of calls to maintain public order.
Calls to disturbances increased significantly after lockdowns began in four cities, although in two of those they returned to normal after a few weeks. This is consistent with lockdowns having an initial shock effect on public behavior, which then slowly eroded as people adjusted to the new normality.
One reason for disturbance calls to increase might be businesses calling about customers breaching rules on social distancing, or individuals reporting their neighbors for failing to stay at home or even coughing loudly. Officers will be able to resolve many of these calls quickly, but they still represent a potential extra burden on resources.
Other types of order-maintenance calls (including calls to drug use, suspicious people and trespassing) stayed largely the same as in previous years – even in a national emergency, people still expect the police to deal with everyday problems. Calls to alarms decreased substantially, with (for example) about 1,300 fewer alarm calls than predicted in Los Angeles in the first four weeks after lockdown, possibly due to more people being at home.
Previous research by the Bureau of Justice Assistance found many law-enforcement agencies assumed there would be large increases in medical calls to police during a public-health emergency. However, calls to medical emergencies, mental health incidents and missing persons all fell in some cities after COVID-19 lockdowns began, while remaining as expected in others.
Agency contingency plans may not have accounted for changes in behavior due to business and school closures. Agencies in those cities hit first by COVID-19 had to learn quickly when facts on the ground differed from their planning assumptions, but police in places hit in any future waves will be able to learn from those experiences.
In contrast to other medical emergencies, calls to dead bodies increased substantially in Los Angeles and more than doubled in New Orleans at the height of the epidemic. Chiefs, in particular, should consider how they can support officers through the trauma of dealing with coronavirus deaths as well as protecting them from the danger of infection when dealing with COVID-19 victims.
Patrol officers frequently respond to traffic crashes, but the number of vehicle collisions dropped substantially in almost every city from mid-March. In the four weeks beginning on March 16, police in the cities in this study responded to about 10,400 fewer traffic crashes than expected. This reflects the empty highways seen in many cities as most citizens complied with stay-at-home orders. Similarly, officers conducted far fewer traffic stops after stay-at-home orders were imposed. However, the number of crashes began to slowly rise after the first three weeks of lockdown. Chiefs may be able to re-deploy officers (at least in the short term) who usually focus on traffic matters to support officers in other divisions that are under extra pressure.
Good data and good analysis will be key to effectively responding to an epidemic in your area. If your agency has a crime analyst, ask them to look at how the first wave of COVID-19 affected policing in your area (if you don’t have a crime analyst, think about getting one). Use your relationships with departments in other jurisdictions to learn what worked about their response and just as importantly, what didn’t work as expected. Finally, make sure your decisions are informed by the latest available information, including real-time data where possible.
How this research was done
Reliably identifying patterns in call data is like trying to hear a friend talking in a crowded bar: possible but difficult. Identifying changes in calls associated with COVID-19 is hard because even in normal times call volumes vary substantially from week to week. Many agencies try to identify patterns by comparing (for example) calls this month to last month, but that often produces misleading conclusions.
To deal with these problems, this study applied statistical forecasting techniques to four years of weekly call data to predict how many calls each agency would have expected each week if coronavirus had not hit. We can compare these forecasts to the actual call counts of each type to estimate the difference between actual calls and the number expected without COVID-19. This is much less likely to produce misleading results than, for example, simply comparing calls in a week with the same week last year. Studying multiple agencies also means we can understand how likely it is that the results will be relevant to agencies elsewhere. The methods used are explained in more detail in a paper in the journal Policing.
About the author
Dr Matt Ashby is a lecturer at the Jill Dando Institute of Security and Crime Science at University College London. He is a former police officer and crime analyst who researches crime analysis techniques and how police can effectively prevent crime.