COVID-19’s Impact on Gun Violence in America

First, it is important to establish whether there was an observed link between COVID-19 and an increase in the U.S. homicide rate. While there is evidence that the U.S. homicide rate was rising prior to the pandemic’s onset,¹⁰ it is undeniable that an immediate, sustained, and significant increase in homicides followed.

On March 11, 2020, the World Health Organization declared COVID-19 a pandemic,¹¹ and in the year that followed, the United States experienced a 31 percent increase in its rolling monthly homicide rate. However, this fact alone is not sufficient to link COVID-19 to a rise in homicides. CDC data show homicides in the United States were already rising before March 2020. As Figure 1 shows, the national rolling average of firearm homicides had been steadily increasing since a June 2014 low. One RAND Corporation study of weekly mortality data found that more than half of the 54 percent increase in firearm homicides between January 2018 and December 2022 occurred before COVID-19 was declared a pandemic.¹² While the study suggests that COVID-19 alone cannot completely explain the increase in homicides, this does not mean that COVID-19 was unrelated to the historic spike in violence.

One of the most conclusive findings related to the spike in violence during the onset of the COVID-19 pandemic is that the spike was entirely due to an increase in firearm-related violence.

During the onset of the COVID-19 pandemic, many Americans sought protection through the purchase of guns. According to data from The Trace and analysis by John Roman, director of the Center on Public Safety and Justice at NORC, annual gun sales in America averaged 13.5 million in the decade before 2020.¹⁵ In 2020, however, this jumped to a record 22.1 million guns sold in that year alone.¹⁶ Survey data from NORC show that this increase in gun sales between March 2020 and March 2022 was “nearly equal parts by people purchasing a gun for the first time and existing gun owners purchasing additional firearms.”¹⁷ This massive surge in gun sales therefore means there was also a massive surge in Americans’ access to guns, which, especially in the homes of first-time gun buyers, greatly increases the risk of someone in the home being fatally injured by a gun.¹⁸ However, additional research found no evidence that excess gun purchasing correlated with the significant increase in nondomestic firearm-related homicides.¹⁹

Still, all available evidence points to firearm-related homicides being the driver of the overall increase in homicides. The same RAND study that looked at weekly mortality data between January 2018 and December 2022 found that while firearm homicides dramatically increased, nonfirearm-related homicides were stable throughout this period.

Furthermore, additional analysis by John Roman supports this finding, showing that the share of all homicides committed with a firearm increased by 8 percentage points between January 2020 and October 2020, while nonfirearm-related homicides remained stable.²⁰ As he concludes, “changes in overall homicide between 2018 and 2023 are almost entirely explained by changes in firearm homicide.”²¹ Similarly, violent crime also remained relatively stable between 2018 and 2023.²² As such, any explanation of why violence spiked during the onset of the COVID-19 pandemic must consider that it was a very particular kind of violence: gun violence.

CAP analysis of CDC firearm homicide data finds that every state with available data²³ saw its firearm homicide rate increase between March 2020 and March 2021.²⁴ Only one state—Oklahoma—did not experience a double-digit percentage increase in the first year of the pandemic. Moreover, while the increase in gun violence and homicides across major cities has been well covered, it is important to recognize that this trend was not isolated to large cities. The CDC categorizes each county in its firearm homicide dataset by level of urbanization.²⁵Aggregating the data by these urbanization categories shows that while the firearm homicide rate increased the most in “large central metro” counties between March 2020 and March 2021, counties with every level of urbanization—including the most rural—experienced a nearly 30 percent increase in firearm homicides.

However, this does not mean that every American city experienced this spike in gun violence in the same way. A study published in the JAMA Network Open models five distinct groups of 88 major U.S. cities based on their gun violence trends during the onset of the COVID-19 pandemic.²⁶ Of the five groups modeled, four groups—representing a combined total of 46 cities—experienced significant increases in their firearm-related homicide rates during the onset of the pandemic.²⁷ The group that experienced the most dramatic increase consists of only two cities: Chicago and New York. These cities saw the largest increases in their rolling gun violence victimization rates between March 15, 2020, and June 30, 2021, with an estimated 3,491 additional firearm-related injuries compared with the period between March 15, 2018, and June 30, 2019.²⁸Meanwhile, based on the same model, the gun violence victimization rate of 42 of the 88 cities only increased by a small and not statistically significant amount over the entire study period.²⁹ These findings underscore that while most American cities experienced some level of increase in their firearm-related homicide rate during the onset of the COVID-19 pandemic, the spike was not uniform, and thus analysis of it cannot be either. Understanding why these groups experienced gun violence differently during the onset of the pandemic, what policy responses cities had and when, and what underlying characteristics may have played a role is key to understanding how to prevent future spikes in gun violence.

Furthermore, while most of the country experienced this troubling increase in gun violence, some communities were particularly affected. Much of the increase was disproportionately concentrated in Black and brown and low-income communities.³⁰ Alex Piquero, acting chair of the Sociology and Criminology Department at the University of Miami, and John K. Roman analyzed CDC WONDER mortality data and found that for all racial groups, firearm homicide rates peaked in 2021. However, while the firearm homicide rate for white individuals peaked at 4.3 per 100,000 individuals, the rate for Black individuals ages 20 to 24 years old was greater than 80 per 100,000 individuals—more than 18 times higher.³¹ Additionally, research by Everytown for Gun Safety found that between 2019 and 2021, racial and gender disparities in gun violence victimization worsened.³² Although every racial group’s homicide rate increased during this period, the gun homicide rate among Black Americans by far grew the fastest, increasing by 48 percent.³³Moreover, exposure to gun violence, especially among children, was concentrated among Black youth, who were exposed to gun violence 4.44 more times than their white peers during the pandemic.³⁴ Analysis by the Brookings Institution supports this finding, showing that during the first year of the COVID-19 pandemic, on average, the lowest-income neighborhoods experienced a homicide rate three to four times higher than that of other residential neighborhoods in the same city.³⁵

Some have suggested⁴¹that the spike in the homicide rate in 2020 was due to the co-occurring events surrounding the killing of George Floyd, the subsequent civil unrest, and the reaction of major police departments to “pull back” from proactive policing.⁴² Importantly, it was not just police practices that changed dramatically in 2020, and other changes may have had a corresponding impact on homicide rates. Many police agencies, and particularly large police agencies, began experiencing significant staffing challenges beginning in 2020.⁴³

However, there is no compelling evidence to suggest these compounding factors played a significant and lasting role. Violence, and in particular gun violence, was already on the rise in the nation before George Floyd was killed on May 25, 2020: During the six weeks prior to his death, homicides increased by an average of 17 murders each week.⁴⁴ While this trend continued after his death, the rate of increase was consistent with the prior six weeks.⁴⁵ If the killing of George Floyd and related theories did account for a substantial portion of the 2020 spike in gun violence, there would likely be a statistically different and more quickly increasing murder trend following his death. While there is some evidence that there was a brief spike following George Floyd’s death,⁴⁶ this was insignificant in comparison with the larger, more consistent trend relating to COVID-19.

CAP analysis similarly finds no statistically significant correlation linking the protests following the killing of George Floyd to a rise in gun violence incidents. While it is difficult to measure civil unrest, CAP’s analysis compares data from the Harvard Kennedy School and the University of Connecticut⁴⁷ on the frequency of protests following his death and the number of population-adjusted gun victimizations during the COVID-19 spike compared with the same time one year before in the 100 most populous U.S. cities. Using protest frequency as a proxy for variation in the levels of unrest across major U.S. cities reveals no significant correlation with an increase in gun violence victimizations.

There also is some evidence⁴⁸ that high-profile incidents of police killings and subsequent increased public scrutiny of policing is related to statistically significant decreases in the number of police stops or low-level arrests, known as the “Ferguson effect.” However, there is no consistent evidence that such a change in policing is correlated with increases in total, violent, or property crime rates.⁴⁹ Additionally, as the Brennan Center for Justice points out, if the policing pullback was responsible for the increase in violent crime and murder during the summer of 2020, there would likely be a proportionate rise in property crime due to police being less proactive in making stops or arrests.⁵⁰ However, overall property crime rates have remained relatively stable,⁵¹ and therefore no such trend exists.

Similarly, while there is a lack of evidence to rule out the possibility that police staffing shortages had an impact on homicide rate changes in 2020, data from the Police Executive Research Forum (PERF) suggest that the staffing shortages were not significantly associated with the overall increase and subsequent decrease in homicides between 2020 and 2024. While the beginning of the spike in homicides coincides with the beginning of the staffing shortage crisis, these trends have significantly diverged in the years since then. According to PERF survey data, compared with January 2020 staffing levels, the largest police agencies continued to experience attrition in every year until 2024, and, even so, remain more than 4 percent below their January 2020 staffing levels.⁵² Despite police staffing levels in the largest agencies being lower than they were in 2020 or 2021, firearm-related homicides in the largest cities have dropped precipitously and returned to at or near prepandemic levels.⁵³ If police staffing shortages were a key driver in the overall spike in homicides, one would expect that the homicide rate would have remained elevated as long as police staffing levels in the largest agencies continued to lag behind pre-2020 levels, but this has not been the case.

While there is no statistically significant evidence that protests or policing changes had a meaningful impact on the larger trend across the nation, this does not rule out that individual cities or neighborhoods within cities may have experienced a stronger impact. Minneapolis, where George Floyd was killed, was seen as a “hotspot”⁵⁴ of protest activity during the summer of 2020. Compared with the summer of 2019, the city experienced more additional gun violence victimizations after adjusting for population during the summer of 2020 than did any of the 100 most populous cities, according to CAP analysis of Gun Violence Archive (GVA) data.⁵⁵ Furthermore, Minneapolis’ gun violence victimization rate increased in the week following George Floyd’s death, during which the city saw 13 more gun violence victimizations than the weekly rolling average of the same week in 2019. In the third week of June, this increased to 35 more gun violence victimizations than the weekly rolling average of the same week in 2019. Additionally, a 2022 study by Kevin Wolff, associate professor at the John Jay College of Criminal Justice, and co-authors analyzing variation in shootings and assaults in New York City boroughs between January 1, 2020, and September 30, 2021, found that, after controlling for various other factors, the period of “social unrest” immediately following George Floyd’s death was associated with a statistically significant increase in both shootings and assaults citywide.⁵⁶However, this increase appears to be almost entirely driven by an increase in assaults in the Bronx.⁵⁷

It goes without saying that the civil and social unrest following the murder of George Floyd, and the subsequent policy and policing response, is essential context for understanding what American cities were facing during the summer of 2020. However, with all available data, it is evident that this moment was not the defining reason why so many American communities experienced a significant and persistent increase in firearm-related violence through the summer of 2021.

One theory about why gun violence increased so dramatically during the COVID-19 pandemic is that increased access to guns in combination with “stay-at-home” orders may have led people to be trapped at home with their abusers, which, when a firearm is present in the home, is five times more likely to lead to death.⁵⁸ One study found that while there was no relationship between “excess” gun purchasing and nondomestic violence during the onset of the COVID-19 pandemic, excess gun purchasing was associated with an increase in domestic violence injuries in April and May of 2020.⁵⁹ Further, a meta-analysis of measuring the effect of lockdown orders on reports of domestic violence finds a statistically significant increase in the percentage of domestic violence reports following a lockdown order.⁶⁰

However, this increase accounts for only a small fraction of the total increase in firearm-related homicides experienced during 2020. CAP analysis of GVA data finds that there were approximately 5,000 additional deaths between March 2020 and March 2021 compared with the previous year. While domestic violence incidents captured by the GVA showed a 10 percent increase in the first year of the pandemic, this only accounted for slightly more than 200 additional homicides—approximately 4 percent of the share of additional homicides between March 2020 and March 2021.

The COVID-19 pandemic disrupted almost every aspect of life. Widespread stay-at-home-orders, mass layoffs, and the closure of schools and other relied-upon local government services were destabilizing for many Americans, the majority of who reported experiencing increased stress during the onset of the pandemic.⁶¹ At the same time that many were facing significant social and economic instability, unfortunately, options to turn to for support became fewer than ever before. Local government employees, such as teachers and criminal legal system workers who play critical prosocial roles, saw significant reductions in their workforce.⁶²More than 1.24 million local government employees lost their jobs in just the first three months of the pandemic.⁶³ The strain caused by the onset of the pandemic—exacerbated by the loss of government and community-based supports—had a disproportionate impact on young men living in neighborhoods where levels of violence were already high, and therefore may have contributed to the overall spike in gun violence.⁶⁴

For example, one study found a significant correlation between unemployment in New York City boroughs and an increase in both fatal and nonfatal shootings, even after accounting for Black Lives Matter protests, daily temperature, and various other controls.⁶⁵ In addition, this instability and loss of typical social supports was combined with disruptions to people’s daily routines. Instead of having normal safe havens away from high-violence areas, such as schools or workplaces, many were constrained near their home where violence is more likely to take place.⁶⁶

This argument is further supported by looking at the characteristics of patients admitted to trauma centers for firearm-related injuries during the onset of the COVID-19 pandemic. One cohort study examining patient characteristics across trauma centers in six states found that there was a significant increase in the number of patients with mental illnesses, alcohol use disorders, and substance use disorders compared with the same period in 2019.⁶⁷ While the study makes no causal claims, this evidence is consistent with theories that the overall spike in gun violence may have been driven by an increase in risk factors associated with violence as people became more disconnected from social or other supports.

Following the summer of 2021, gun violence has been steadily declining across the nation. CAP has been tracking this decline, most recently publishing analysis in September 2024 showing that year-to-date gun violence homicides were down 14.7 percent in the 300 most populous cities compared with 2023.⁶⁸ End-of-year analysis by Jeff Asher similarly shows that the murder rate dropped by 14 percent or more in 2024 compared with 2023.⁶⁹ This would constitute the largest single-year drop in the nation’s murder rate recorded in modern history after 2023, which is also estimated to have been a historic decline.⁷⁰ After back-to-back years of historic declines in the nation’s murder rate, much of the country has returned to prepandemic levels, and some cities have become much safer. Last year, Boston saw its lowest homicide rate in nearly 70 years with 24 homicides,⁷¹ and Detroit experienced its fewest homicides since 1965.⁷² However, there is still work to do across the country. The Real-Time Crime Index shows that the national 12-month rolling sum of murders in December 2024—the most recent available data—is still 4.1 percent higher than at the same point in 2019.⁷³

As with the spike in violence during the onset of the COVID-19 pandemic, it will take years and multiple efforts to disentangle why exactly gun violence rates fell following the summer of 2021 almost as quickly as they increased. Nevertheless, a lot can be learned from looking at the U.S. cities where gun violence has seen the most significant drops. Many of the cities that experienced the largest declines in violence leveraged federal funding from the American Rescue Plan Act (ARPA) of 2021,⁷⁴ the Bipartisan Safer Communities Act (BSCA),⁷⁵ and the Community Based Violence Intervention and Prevention Initiative (CVIPI)⁷⁶ to support community-based strategies, such as community violence intervention (CVI) programs.

CVI programs are unique in their capacity to utilize credible messengers to reach individuals and transform communities.⁷⁷ By connecting individuals at the highest risk of being involved in violence with resources, services, and mentorship, CVI programs address upstream factors that contribute to community violence.⁷⁸ Successful evidence-based models have been shown to reduce gun victimizations by more than 60 percent⁷⁹ and save taxpayers up to $41 for every dollar invested in a CVI program.⁸⁰

As gun violence continues to decrease in major U.S. cities that have made significant investments in CVI programs, it is becoming clear that CVI programs are working. For example, Bakersfield, California, has seen incredible public safety gains after employing its gun violence reduction strategy in 2020 and establishing a central office focused on coordinating violence intervention efforts across hospitals, law enforcement, trauma recovery centers, and community-based organizations.⁸¹

Despite this, sustainable funding for these lifesaving programs is still at risk,⁸² and CVI programs that have successfully contributed to record declines in Detroit⁸³ and Philadelphia⁸⁴ are at risk of layoffs or even closure without increased federal funding. This is especially devastating to CVI work since so much of its success is built on trust from community members which they have earned over years.⁸⁵ Inconsistent funding can lead to disruptions of service provision, thereby undermining their legitimacy.⁸⁶ If there is one lesson to be learned in the five years since COVID-19 was declared a pandemic, it is that more funding is needed for CVI programs, not less.

Five years after COVID-19 was declared a pandemic, Americans are still dealing with the consequences. Fortunately, as CAP has regularly analyzed⁸⁷ and reported,⁸⁸ gun violence across America was falling at historic rates through the end of 2024—and in some places such as Baltimore, Maryland, and Bakersfield, California, gun violence rates are already well below what they were before the pandemic began.⁸⁹ However, while it is worth celebrating that rates of gun violence are continuing to decline, it is important to remember those affected by this violence and the resources that are needed to support comprehensive solutions.

The major federal investments in public safety through the ARPA and the BSCA are set to expire at the end of 2026.⁹⁰ Unless local, state, and federal lawmakers commit now to making additional and sustained investments in prevention, many of the programs credited with helping the nation recover will be forced to shut down, putting the progress made in recent years at risk. Moreover, while gun violence may broadly be returning to or dropping below prepandemic levels, those levels still represent an unacceptable amount of violence in U.S. communities. While the United States cannot return loved ones lost to the epidemic of gun violence, policymakers can continue to honor them by learning the right lessons to prevent another spike. Commonsense gun safety laws and community-based public safety strategies have made communities safer over the past five years. It may be impossible to predict if or when the next pandemic will occur, so it is imperative to invest now in the solutions that will make communities safer no matter what the future holds.