During COVID researchers have conducted empirical surveys assessing police officer perceptions and agency reactions to the pandemic. While these have been on limited samples or occurred after the beginning months of COVID or the murder of George Floyd,¹⁰ they offer several insights. Early studies, for example, focused on the impact of the pandemic on officer perceptions rather than agency operations. Frenkel et al. (2021) surveyed officers from six agencies in five European countries between March and June to understand officer stress, strain, emotional regulation, and preparedness for the pandemic. They found that officers tolerated stress fairly well (conditioned by other factors), although the risk of officers getting COVID and poor agency communication exacerbated officer stress. Between July and September 2020, Kyprianides et al. (2021) surveyed 325 officers in the United Kingdom to ask about their policing experiences during COVID-19. They found that positive organizational support was associated with use of force restraint, procedural justice policing, and better officer health. However, Kyprianides et al. (2021) also discovered that greater officer self-confidence was associated with poorer health and more support for police use of force during the pandemic. Additionally, in March 2021, Maskály et al. (2022) asked 167 officers and non-executives from seven U.S. agencies about operational and organizational changes due to COVID-19. Their results show substantial heterogeneity in how officers viewed organizational policies about the pandemic, both within the same organization and across different organizations. Overall, officers experienced the pandemic differently, and this experience was likely conditioned by characteristics of their agency and their own health concerns.

At the organizational level, although there have been anecdotal accounts of the impacts of COVID-19 on individual agencies (see, e.g., Jennings & Perez, 2020),¹¹ systematic surveys across agencies were scarce during the first year of the pandemic. We found only three, and none captured changes at the beginning of the pandemic. For example, informed by our surveys, Maskály et al. (2021) queried police executives using international contacts and received responses from people from 27 countries (they did not report how many individuals responded or from what agencies). The authors noted their survey was sent in “the summer of 2020” (p. 271). As with their officer survey mentioned above, their findings were highly heterogeneous, indicating that changes in organizational operations varied across responses. However, they found some consistent changes, including decreases in in-person training and roll calls, restrictions to public access to police agencies, and increases in remote work. It was unclear from the survey how the worldwide protests for police reform during the time of their survey confounded these findings.

Mrozla (2021), also building off of our surveys, queried approximately 2500 rural agencies in the United States (serving populations of 10,000 or less) between May and September 2020, focusing on how rural police agencies responded to the pandemic. Using email addresses obtained from agency websites, they received responses from 312 rural agencies. Mrozla found that agency size may be positively correlated with having policies related to pandemics. They also noted the specific challenges of personal protection equipment (PPE) provisions for rural agencies, officer shortages, and the need for risk management and preparation for these agencies. Mrozla importantly finds that contrary to some beliefs, rural agencies may not be immune to pandemics and should also prepare for them.

A third survey was reported in Security Magazine by Ekici and Alexander (2021). They surveyed agencies post-Floyd starting in June 2020 in Illinois, Missouri, and Ohio (they did not report their survey methodology). They received responses from 73 agencies in Illinois, 30 in Missouri, and 97 in Ohio. Ekici and Alexander (2021) noted significant reductions in these agencies in enforcement, arrest, traffic and pedestrian stops, police training (including academy shut-downs), and community access to police officers.

We believe that our surveys remain the only effort to document—in real time—the immediate (March-May 2020) impacts of COVID-19 at the start of the pandemic and before Floyd’s death, for a large and diverse sample of law enforcement agencies across the U.S. Capturing the immediate impacts of COVID-19 on a wide range of agencies, especially during the first two months of the pandemic, most accurately showcases the initial challenges that U.S. law enforcement agencies face in a rapidly evolving public health crisis with regard to their daily public safety deployment. This understanding is crucial to better planning, preparedness, and response in the future.

The challenges of surveying a representative sample of the over 18,000 U.S. law enforcement agencies have been well-documented in the national policing survey research (see, e.g., challenges of the Bureau of Justice Statistics’ Law Enforcement Management and Administrative Statistics (LEMAS) survey).¹² Further, surveying agencies quickly and during a crisis certainly dims prospects of success. Thus, at the start of the pandemic in March 2020, the first author partnered with the International Association of Chiefs of Police (IACP) leadership to create and implement the surveys. The IACP is the world’s largest professional organization for police leaders, whose tens of thousands of members from around the world include over 6400 chief executives from approximately 5800 agencies in the United States and Canada (with the vast majority from the U.S.). In addition, the chief executive members of IACP are from a wide array of urban, suburban, and rural agencies dispersed across the United States. Using the IACP membership agencies was, therefore, the quickest way to obtain the most immediate estimate of the initial impact of COVID-19 on a large sample of diverse law enforcement agencies in North America.

The IACP agreed to implement two waves of surveys, one from March 25 – April 3, 2020, and the second from May 12 – 25, 2020. The surveys were officially sent by IACP to all of its U.S. and Canadian chief executives by email, and a Qualtrics link was provided for agencies to fill out the survey online. Two additional reminders were sent within the period each survey was open. As more than one individual who received the survey request might have served in an executive leadership role in the same agency, recipients were given explicit instructions that only one survey was to be filled out for each agency by the chief executive with direct knowledge of operational adjustments due to COVID-19. Recipients were also frequently reminded across each survey of the “as of” date for questions, which was two days before each survey’s release. Specifically, for Wave 1 (released on March 25, 2020), respondents were asked to provide answers “as of March 23, 2020.” For Wave 2 (released on May 12, 2020), respondents were asked to respond with answers “as of May 10, 2020.” To ensure that the murder of George Floyd and subsequent protests did not impact agency responses, we did not include any surveys received on or after May 25, 2020, the last day of the second survey implementation. Only two responses were removed because of this reason.

As the survey was implemented at the start of the pandemic, it was difficult to determine precisely the relevant questions to ask. Because of this, the March and May surveys were not identical, although many questions were similar. For example, in the first survey, we focused on changes to agency operations, agency preparedness, changes to the civilian workforce, and the impacts of stay-at-home orders. In the second wave, we continued to ask about these changes but also inquired about trends of specific categories of calls for service and budgetary concerns.¹³

For the first wave, 989 agency representatives completed and returned surveys after two reminders (reflecting a 17% agency-level response rate). For this analysis, we removed the 11 Canadian agency responses given the different country context of those surveys (although they were included in the initial fact sheets distributed to the agencies). For the second wave, 1141 surveys were completed and returned after two reminders (reflecting a 20% agency-level response rate). Ten of these were from Canadian agencies, which, again, were removed for this analysis. While checks across the data (I.P. addresses, location, state, number of officers in the agency) did not indicate multiple responses from potentially the same agency, the anonymous and voluntary nature of the survey made this impossible for us to determine with 100% certainty (for example, in both Wave 1 and 2 surveys, approximately 7% of respondents did not provide the state where their agency was located). However, given our checks, we believe the responses represent unique and individual law enforcement agencies in the United States but note this possible source of error. We also do not know if the same agencies answered both surveys and can only estimate impacts based on a cross-section of the two samples taken from the same population.

Table 1 shows the proportion of agency responses across various characteristics, including the number of sworn officers, civilian personnel, population served, and U.S. region. Approximately three-quarters of agencies in the United States have fewer than 25 sworn officers, and only 5% have over 100 officers (Hyland & Davis, 2019). Thus, agencies with 100 or more officers are overrepresented in both waves of this survey. Nonetheless, we captured a substantial proportion of agencies with 25 officers or less (41% in Wave 1; 32% in Wave 2). Additionally, at least half of the agencies participating in each survey serve populations of 25,000 or less, showing that we captured a large sample of (likely) rural agencies. In the second survey, a larger proportion of agencies with more than 50 officers serving places with a population of more than 25,000 answered the survey compared to smaller agencies with smaller populations. Participating agencies represented every state and Bureau of Economic Analysis region of the United States.Table 1.

U.S. Agencies that Responded to Each Survey.

Given space limitations, we do not present all of the findings from both surveys here. However, we present several main results, focusing on how the pandemic significantly changed the demand and supply for police services. By “demand,” we mean the public’s expectation of safety and police legitimacy, often manifested through the public’s calls for police service but also from other community requests to the police. By “supply,” we mean the resources and activities the police use to meet that demand for public safety. These can include police response to calls for service, enforcement activities, proactivity, and community engagement.

In this study, we report on the findings of the only systematic agency surveys conducted across a large and diverse sample of U.S. law enforcement agencies during the first two months of the COVID-19 pandemic and before the murder of George Floyd. The ability to carry out these two national surveys as the pandemic broke speaks to the partnerships that can be established between researchers and practitioners for a common goal. Given that the surveys were implemented before George Floyd’s murder, they are also the only agency-level surveys with responses that are not confounded by that event and its aftermath. As with other national policing surveys, our surveys somewhat oversample larger agencies (although a substantial proportion of the sample are small agencies) and are samples of convenience (using the IACP membership list). Additionally, given the limitations detailed in our methods section, we could not conduct a longitudinal analysis given the cross-sectional data collected. Despite these shortcomings, our surveys show that COVID’s impacts on policing were dramatic in the first two months of the pandemic, with some lingering effects. Not only were the impacts of COVID-19 on police agency operations substantial, but these impacts were shared across agencies large, small, urban, suburban, and rural.

The surveys reveal that COVID-19 significantly altered both the supply and demand for police services in the first three months of the pandemic. Concerning demand, we find the overall volume of calls for service initially declined, confirming what others have generally found (see, e.g., Abrams, 2021; Ashby, 2020; Campedelli et al., 2020; Langton et al., 2021; Lopez & Rosenfeld, 2021; Piquero et al., 2021). The halting of public life dramatically altered the opportunities, routines, situations, and interpersonal exchanges associated with crime, disorder, conflicts, and traffic accidents. These changes naturally impacted the volume, frequency, and types of calls for emergency services through 911 and other non-emergency public safety numbers. At the same time, studies have revealed that the short and long-term effects of COVID-19 on specific types of calls for service and crime are complex. Certain requests, such as those for traffic issues or minor disorders, decreased early in the pandemic, likely due to changing routines and the ceasing of public life. Some declines in calls for service may have also occurred due to a decline in official reporting but may not reflect actual trends, such as domestic violence (Nix & Richards, 2021; Piquero et al., 2021) or child abuse (U.S. Department of Health & Human Services, 2022). In our surveys, agencies that reported increases in calls related to domestic violence also often reported increases in calls for mental distress. Others have found that firearms and other violence have increased in some cities and at certain times (Beard et al., 2021; Kim & Phillips, 2021; Lopez & Rosenfeld, 2021). Our survey indicated that agencies that experienced increased violence initially in the pandemic may have also experienced increased commercial burglaries.

The overall decline and heterogeneity of trends across call categories suggest that public safety demands may have transformed significantly and quickly, reflecting a new portfolio of community public safety needs for the police. Typical patterns of resource allocation and prioritization that police had been accustomed to changed almost overnight. Critical questions for researchers to explore about these shifts are whether and why certain calls went up or down and whether law enforcement agencies were prepared for these shifts. For example, were agencies prepared to handle the changing needs and shifts in tactics required to respond to (and prevent) increases in specific types of events? When agencies did shift to remote response for certain calls, were they for call types that increased or decreased? What were the consequences of those adjustments for public safety?

At the same time as changing demands, the police adopted policies to modify the supply of police services, likely to protect their officers from contracting COVID and anticipating staffing shortages. A significant majority of the responding agencies from both surveys restricted officers from responding to certain calls for service in person, reduced arrest and proactive enforcement, and stopped community engagement activities. Some of these adjustments may have continued longer than expected. As previously mentioned, Maskály et al. (2021) found similar results in a smaller international sample of agencies during the summer of 2020. Ekici and Alexander (2021) noted these same trends again in the fall of 2020. Our survey also indicated that these reductions in supply were not met by a substantial increase in other forms of communication with the public. If thinking of this as a supply and demand relationship (see Figure 3), the shift in both the supply (S₁ to S₂) and demand (D₁ to D₂) for police services may have significantly moved the point of equilibrium (E₁ to E₂) of police-community interactions. The magnitude and impact of this change depend not only on the elasticity (or slope) of each curve but also the extent of the shifts of both.Figure 3.

Shifting Supply and Demand of Policing during the First Few Months of the COVID-19 Pandemic.