Explaining the Recent Homicide Spikes in U.S. Cities: The “Minneapolis Effect” and the Decline in Proactive Policing

Cory Wilson is a criminal defence lawyer, serving Calgary, Okotoks, Airdrie, Strathmore, Cochrane, Canmore, Didsbury, Medicine Hat, Lethbridge and Turner Valley.

Major cities across the country have recently suffered dramatic increases in homicides. The increases are remarkable, suddenly-appearing, and widespread in cities across the country—although often concentrated in disadvantaged neighborhoods. For example, in mid-August, USA Today reported that major U.S. cities had been plagued by a “horrifying increase in gun violence” this year, with shootings skyrocketing in certain neighborhoods in cities from Philadelphia, New York, and Milwaukee to Los Angeles and Denver. At least 11,047 people have died in gun violence so far in 2020, compared to 15,208 in all of 2019. At this rate, 2020 will easily be the deadliest year for gun-related homicides since at least 1999, all the while other major crimes are trending stable or slightly downward.

While homicides have spiked since late May, many other crime categories have followed a different trajectory. For example, the Rosenfeld-Lopez Report looked at nine other crime categories—gun assault (a subset of aggravated assault), domestic violence, robbery, burglary (including the subsets of commercial and residential burglary), larceny, motor vehicle theft, and drug offenses. None of the other crime categories exhibited a structural break starting in late May. Burglaries, for example, abruptly increased (by 190%) during the last week of May and then, just as abruptly, returned to normal levels the next week. A dramatic increase in commercial burglaries in that week—coinciding with looting associated with the mass protests following George Floyd’s death—explains this pattern. The one crime category that most closely tracked homicides was aggravated assault.

Looking at 17 large cities for which data were available, the Rosenfeld-Lopez Report identifies a structural break near the end of May 2020 for that crime category as well. Aggravated assaults rose by 35% from late May through the end of June 2020. Gun assaults also showed a sharp and sustained increase after late May, although the pattern was not clear enough (at least as of the end of June) to be identified as a structural break. And robberies exhibited a structural break, but the timing was slightly earlier. Robberies exhibited a long-term downward trend, but after March 2020 the robbery rate rose by 27% through the end of June. Other observers have reported roughly the same pattern—increases in homicides and crimes categories associated with gun crimes but not in other categories. For example, a report in the New York Times in early July noted that trend lines for murder and other violent crimes rarely move in oppose directions. But in 25 large American cities in 2020, overall crime was down 5.3% and violent crime was down 2%—but murder was up 16% in relation to the previous year. The homicide increase was widespread, present in 20 of 25 cities at that time. Based on this information, it seems fair to say that most of the largest cities in the U.S. are experiencing a homicide spike and, likely, a shooting spike—but not, generally speaking, a significant increase in other crime categories

The Timing of the Homicide Spikes  

The Timing of the George Floyd Protests

Before turning to the timing of the homicide spikes, it may be useful to set out for reference the timing of the protests surrounding George Floyd’s death. Floyd died on May 25, 2020 while being arrested by officers from the Minneapolis Police Department. Protests began in Minneapolis the next day, May 26, and rapidly escalated on May 27 through 29. Protests outside of the Minneapolis area were first reported on May 27 in Memphis and Los Angeles. By the start of June, protests had been held in all U.S. states. By early June, at least 200 cities in the U.S. had imposed curfews and, by the end of June, more than thirty states had activated over 62,000 National Guard personnel due to mass unrest. Anti-police protests continued through July and beyond.

The Timing of the Minneapolis Homicide Spike

With the timing of the protests in mind, how does it compare to the onset of the homicide spikes that occurred in cities across the country? Of course, in a nation as large as the U.S., it would be unlikely that precisely the same effects would be felt in every city. At the same time, when exploring such crime issues, using “case studies” is a helpful method—a point that the Professor Rosenfeld (among others) has helpfully developed. For purposes of this blog, I selected four cities to investigate. I first identified Minneapolis as the city in which the protests originated. Next, I added Chicago, Philadelphia, and Milwaukee, based on the note in the Rosenfeld-Lopez Report that the recent homicide spikes were led by these four cities. Perhaps exploring these cities, where homicide increases were most evident, would more clearly reveal what has happened.

Turning first to Minneapolis: Minneapolis suffered an incredible homicide spike. As of August 1, 2020, 41 residents of Minneapolis had been murdered, compared to 21 as of the same time in the previous year13—a 95% year-over-year increase. When did the spike start? One convenient way of obtaining data is to look at the expansive website, citycrimestats.com, where Professor David Abrams at the University of Pennsylvania Law School has usefully assembled a valuable trove of empirical data on crime in the United States.

The Timing of the Chicago Homicide Spike

As mentioned above, Chicago is also suffering a horrific homicide spike. Until the last week in May 2020, Chicago’s homicides had not increased compared to the previous year. Between January 1 and May 28, 2020, Chicago had 191 homicides. During the same time frame in 2019, Chicago had an almost identical number of homicides—192. Then, on May 31, 2020, 18 people were murdered, making it the single most violent day in Chicago in six decades. Violence continued through June and July. Nearly 40 people were shot in the city in the last weekend in July, as the most violent month in 28 years in the nation’s “Second City” drew to a close. At least 107 people were killed in July 2020, more than double the number who were killed in July 2019 and the most homicides in Chicago in a single month since September of 1992.

The Timing of the Philadelphia Homicide Spike

Philadelphia, the nation’s sixth largest city, has also been hit hard by a homicide spike. As of July 26, homicides in the city totaled 247, a 32% increase compared to the same period a year earlier. This total puts Philadelphia in second place in the nation for the number of homicides, just behind Chicago. As an example, during the Fourth of July weekend, the city’s “gun violence epidemic reached alarming heights,” according to the Philadelphia Inquirer, as 23 people were killed across the city—the most in a single day since 2013. If these trends continue, the city is projected to finish 2020 with more than 1,700 shooting victims, the highest total since at least 2007.

The Timing of the Milwaukee Homicide Spike

Continuing down the depressing list of cities wracked by gun violence this summer, Milwaukee has also suffered a homicide spike. At the beginning of 2020, the city was touting its five-year reduction in violent crime, including a dramatic decline in homicides. But by the Fourth of July, 86 people had been murdered in Milwaukee in 2020, which was double the number of victims at the same time in 2019.28 By August 17, Milwaukee had seen a 112% increase in homicides from the previous year.

Explanations for the Homicide Spikes: Possibilities Other than Policing

With homicides spiking, what is the cause? Before turning to particular explanations, it is important to recapitulate precisely what the causal factor must explain—i.e., it should explain why: Homicides (and likely related crimes, such as shootings and aggravated assaulted) have recently spiked in many large U.S. cities; Other crime categories, such as property crimes, have not simultaneously increased along with homicides; The homicide spikes began around the last week of May 2020, and have been sustained through (at least) the next two months, June and July; and the spikes may well be a phenomenon confined to U.S. cities. Popular media sources have proposed various possible explanations for the homicide spikes. Broadly speaking, the possible explanations divide into two categories: (1) a grab bag of various ideas unrelated to policing; and (2) changes in policing, such as “de-policing” or “de-legitimizing” of police. For reasons that will be developed in the next section, the explanations associated with changes in policing seem the most plausible. But, as part of a process of elimination, this section will first consider other possible explanations, concluding that none of them appears to be a particularly strong candidate for explaining the spikes.

Homicides During the George Floyd Protests

Given the timing of the homicide spikes, one possibility is that the homicides themselves came as a result of the protests, such as from shootings during the protests or by protesters or counter-protesters. This possibility can be quickly dismissed. By all accounts, the majority of the protests were peaceful. The Rosenfeld-Lopez Study, for example, notes that while some violence has been directly connected to protests, in most cases the shooting crimes appear to have involved perpetrators other than the protesters. Virtually all of the homicides that are occurring are taking place away from the demonstrations. The crime maps published by many cities show homicides are not in city centers where anti-police protests usually happened, but rather in low-income neighborhoods outside of those city centers

It is also important to understand that the homicide spikes are not limited to a single week—such as the week surrounding May 25, when George Floyd died and the protests were most intense—but rather have continued. The homicide and shooting data from Minneapolis, Chicago, Philadelphia, Milwaukee, and New York City recounted above suggests that, in general, homicides (and shootings) continued to be substantially above average throughout the months of June and July.

Seasonal Impacts

Another possible factor to be considered is whether the recent homicide spikes are simply a reflection of crime “seasonality,” that is, the well-known tendency for crimes in some cities to increase in the summer. In Chicago, for example, homicides (and shooting crimes) usually increase in the warmer summer months and decrease in the colder winter months. This seasonality theory is unlikely because the recent homicide spikes are far above ordinary seasonal variation. Seasonal effects are already considered, for example, in the Rosenfeld-Lopez Study, which used a structural break identifier that controlled for crime rates during the same week in the previous year. Thus, the structural break in the homicide data was in excess of what ordinary seasonal effects would predict. The homicide spikes (and shooting spikes) are well above standard seasonal variations. Finally, it is important to understand that the recent homicide spikes have occurred in cities across the country, harming cities with a variety of climates. For example, homicides have spiked not only in cities in the Midwest (such as Chicago) or the Northeast (such as New York City), but also in Austin, Fort Worth, San Antonio, Phoenix, Jacksonville, and Los Angeles. Some studies find that crime seasonality is stronger in cities with colder winters and warmer summers. The existence of homicide spikes in cities with different climates across the country confirms that something other than seasonal variation is at work.

Increase in Firearms Purchases

Another possible explanation for the homicide spikes is that a surge in firearms purchases lead to more gun violence. This issue has been explored in a paper by Professor Julia Schleimer and several colleagues. They report that about 2,109,000 “excess” firearms were purchased in the U.S. during March, April, and May 2020—i.e., substantially more firearms than would have been purchased normally during that time frame. The increase was 947,000 in March 2020, another 550,000 in April 2020, and 610,000 in May 2020. They hypothesize that this increase in firearms lead to an increase in firearms violence during these three months. The authors used multivariate regression to control for possible confounding factors, although the controls they included (COVID-19 cases, stay-at-home orders, social movement, precipitation and temperature) are very limited, entirely ignoring law enforcement and other criminal justice variables—a significant omission, given the potential significance of the law enforcement issues discussed below. The authors found a small (8%) increase in firearms violence, which they attributed to the increase in firearms purchases. Here again, as an explanation for the sharp homicide spikes appearing in late May, the increase in firearms purchases appears to be poor candidate. As the firearms purchase numbers above indicate, the most significant increase was in March 2020—well before the homicide spikes appear. Indeed, on close examination of the study, it is interesting to note that actual firearm violence in March and April 2020 was not higher than what the study’s model predicted. The only month for which actual firearm violence exceeded what the model predicted was May 2020 —when the homicide spikes began. Thus, the model they use is very sensitive to the time periods included

Law Enforcement Related Explanations for the Homicide Spikes

Having eliminated all of the other most plausible (and most commonly mentioned) explanations for the recent homicide spikes, the law enforcement possibility emerges as the most probable candidate. This may seem an unsurprisingly conclusion, since “[i]f we take it as an axiom that the police deter crime, it is natural to think that when the patterns of crime start changing systematically, the police must have played some role.”

This section explains why reduced proactive policing (sometimes referred to in the academic literature “de-policing”) i s the most logical explanation for the recent homicide spikes.

De-Policing and De-Legitimizing of Police as Possible Explanations

The Rosenfeld-Lopez Report raises the possibility that changes in law enforcement in the wake of the George Floyd protests caused the homicide spikes. The Report compares the recent homicide spikes with the homicide increases that occurred about five years ago in some American cities, including Chicago, Baltimore, and Ferguson, Missouri. These homicide increases also occurred in the wake of widespread protests against police violence—protests initially triggered by the police killing of Michael Brown in Ferguson, Missouri. At the time, some observers argued that a “Ferguson Effect” was responsible for the crime increases. The “Ferguson Effect” was a term first coined by St. Louis Police Chief Sam Dotson III in late 2014. The essential idea was that, in the wake of anti-police protests after the events in Ferguson, officers were disengaging from discretionary enforcement activities and criminals felt empowered. Heather MacDonald further popularized the term in a Wall Street Journal op-ed in May 2015.

The theory advanced here is that a reduction in discretionary law enforcement activities—i.e., a “Minneapolis Effect” akin to the earlier Ferguson Effect—explains the abrupt homicide spikes that have occurred in the last several months. The theory is straightforward to articulate: In the wake of the anti-police protests surrounding George Floyd’s death, less policing has occurred. For example, police have been redeployed to manage the protests, diverting them from anti-gun patrols and other activities that deter the carrying of illegal firearms. And even after protests began to wane, police have pulled back from some kinds of proactive policing—that is, self-initiated policing methods designed to reduce crime by using preventive strategies, such as street stops or anti-gun patrols. These reductions have resulted from the protests or other attacks on police, as police have (for various reasons) pulled back from aggressive efforts to combat gun crimes. Likewise, law enforcement capabilities have been diminished by reduced funding and other setbacks (such as increased retirements due to demoralization). The consequence of reducing law enforcement activity directed against gun violence has been, perhaps unsurprisingly, an increase in gun violence. In contrast to reactive policing (that is, simply responding to a crime that has already occurred), proactive policing may be particularly effective in discovering and removing illegal firearms from the hands of criminals and thus in deterring the illegal carrying of firearms in the first place. And proactive policing is one of a relatively small number of police practices that has been shown to be likely to directly reduce gun crimes, as the empirical studies collected in the note below suggest. And some studies also similarly indicate that that proactive policing may specifically reduce firearm-related crimes but no other violent crimes or property crimes.

Thomas Abt recently and thoughtfully discusses a wide range of possible response to homicides. But one point is particularly relevant to the recent developments. Abt discusses “hot spot” policing, explaining that “[w]hether a hot person carries a hot gun in a hot spot depends on, among other things, supply and demand. To reduce the demand for illegal firearms among dangerous people in dangerous places, the risk of apprehension must be high.” If recent developments have reduced the risk of apprehension for carrying illegal firearms or commitment gun crimes, then the likely result is that gun crimes increased—i.e., a Minneapolis Effect occurred. This straightforward argument suggests that a likely cause for recent homicide spikes might be changes in proactive policing in various forms. To explore this hypothesis, it would be ideal to have comprehensive measures of all forms of policing activity, and then subsets of proactive or other the types of policing that might be particularly important in fighting gun violence. Unfortunately, such information is not readily available, although some law enforcement agencies do keep track of some forms of policing (e.g., the number of street stops or vehicle stops). This data is typically only available at a city level, suggesting that cases studies of particular cities may be required to explore the hypothesis. The idea that a new Minneapolis Effect (like the earlier Ferguson Effect) might be causing the homicide spikes in Minneapolis and other cities was thoughtfully discussed by Rosenfeld and Lopez. But rather than immediately endorse the conclusion, they pointed to two possible ways of describing the Ferguson Effect. As they explained in connection with the homicide spikes from several years earlier, crime increases might be caused either by “de-policing” or, alternatively, by “de-legitimizing” of the police: Analysts tied the heightened violence [in 2014-16] to two versions of the so-called Ferguson Effect. The first connects the violence to “de-policing,” a pullback in law enforcement. The second essentially turns this explanation on its head and connects the violence to “de-legitimizing,” positing that communities, disadvantaged communities of color in particular, drew even further away from the police due to breached trust and lost confidence. As a result of diminished police legitimacy, fewer people reported crimes to the police or cooperated in investigations, and more engaged in street justice to settle disputes.

In their July 2020 report, Rosenfeld and Lopez conclude that “[i]t remains unclear whether either of these theories explains the previous rise in violence, much less today’s increase.” They urged further investigation of the topic—and, as they encouraged, this paper will now explore the evidence for a Minneapolis Effect, in either its de-policing or de-legitimizing forms.

De-Policing in Specific Cities

To analyze whether changes in law enforcement can explain what is happening today, one approach is to drill into the data in particular cities. Once again, it may be useful to look specifically at the five cities previously examined, as case studies for possible changes in policing.

Minneapolis

First, consider Minneapolis—the namesake for the Minneapolis Effect and the city of origin for the George Floyd protests. What could explain the pattern of sharp and sustained increases in homicides and shots fired in Minneapolis but no parallel sustained increase in property crimes? The sequence of relevant events strongly suggests the possibility of a decline in policing: After George Floyd’s death, anti-police protests quickly followed, which escalated to looting of businesses. As the protests turned violent, the mayor made a decision to surrender the Minneapolis Police Department’s Third Precinct headquarters to protesters. Crowds broke in and set fire to the building—which left many police officers feeling abandoned. Shortly after the initial riots ended, nine members of the Minneapolis City Council pledged to begin the process of abolishing the Minneapolis Police Department. And the police officers involved in Floyd’s arrest were charged with either committing or aiding and abetting murder. In the wake of these events, some city residents said that officers were pulling back from aggressive police work. Following the defunding efforts in the city council (which the council approved on June 15), residents reported a notable decrease in police presence—“All you see now is them [the police] with their windows up,” one resident told reporters.  In the two months following Floyd’s death, the Minneapolis Police Department lost at least 100 officers, straining department resources. Some of the officers resigned because they felt lack of support from city leaders. Additionally, another 75 took medical leave for post-traumatic stress disorder that they identified coming from the riots following Floyd’s death. Police union leaders reported that Minneapolis police officers “are not going to put themselves out there to get the proactive stops to get the guns off the street … [b]ecause they don’t feel supported, after the fact.” In addition, reports emerged of officer reluctance to enter certain high-crime areas. For example, the Minneapolis Star-Tribune noted that homicides began to increase at the same time as (among other things) “the reluctance of some Minneapolis officers to take initiative amid intense scrutiny.”

One Minneapolis city council member questioned police officers’ apparent reluctance to enter the area surrounding Floyd’s memorial, an area described by the Minneapolis Star-Tribune as “a long troubled corner” that was the site of several shootings after Floyd’s death. The council member said that “[p]eople in this area, they’re not experiencing slow response [to call to police for service], they’re experiencing no response. They’re being told that this is called a no-go zone by MPD.”82 Complaints about slower response times went up, which police attributed to patrol squads throughout the city being increasingly stretched thin. The number of gunfire incidents also appeared to overwhelm available law enforcement units. In a class-action lawsuit filed against the city in late July, a group of neighborhood residents in a high-crime area alleged that it had been deprived of adequate policing, and regularly were told to call 311, instead of 911, whenever there were shots fired but no victims.

Chicago

Chicago is another large American city where homicides and shootings have spiked substantially—and the spikes began immediately following large anti-police protests in Chicago that required a significant police response. As with other major cities, Chicago witnessed substantial anti-police protests following Floyd’s death on May 25, 2020.94 On May 29, demonstrators shut down several downtown streets. On May 30, protests escalated, and one person died and six others were shot. A dozen police officers were injured. Protests “evolved into criminal conduct” (as Mayor Lightfoot put it) and protesters and looters did extensive damage to businesses on Michigan Avenue. The next day, Mayor Lightfoot asked the Illinois Governor to summon the National Guard to Chicago for the first time in 52 years (the last time being the 1968 riots).

The economic costs of the protests and looting in Chicago through June 1 were estimated at around $66 million. And, as noted above, on May 31, 2020, 18 people were murdered in Chicago, making it the single most violent day in six decades. Anti-police protests continued in Chicago, and a curfew was imposed through June 7. Protesters marched through a neighborhood where Chicago police officers lived. Protests continued through June and July. On July 17, 2020, 18 police officers were injured, as police reported “being outnumbered and unprepared.” Sadly, the looting in Chicago did not end in July. On the evening of August 9-10, after police shot a man in Englewood, hundreds of people rioted in Northern Chicago. They targeted Chicago’s famed “Magnificent Mile,” where dozens of stores were looted, at a cost of over $60 million in damages. And at least 13 police officers were injured. In light of these events, the theory advanced here is not that the protests themselves produced murders and shootings. Instead, the theory is that as police have had to respond to the protests and their surrounding violence—and subsequent policing and investigations associated with those protests—day-to-day policing (particularly proactive policing) has suffered. Evidence suggests that ordinary policing in Chicago declined during the month of June, dropping substantially below normal levels. For example, during June 1 through 28, traffic stops dropped by 86%, street stops by 74%, and arrests by 55% compared to the same period in 2019. And murders were up in Chicago by a staggering 83% compared to the same period one year earlier. But simply pointing to these changes in policing, by itself, seems unlikely to explain what is going on in Chicago. From mid-March to mid-April, policing was also down substantially. Traffic stops fell by 84%, street stops by 64%, and arrests by 70% compared to that same period in 2019. And murders fell by 13%.

Precisely why policing activity diminished in Chicago from March through July is a point of contention. Chicago Police Superintendent David Brown attributed the decline to reduced social mobility due to COVID-19—arguing that fewer people were on the streets and thus fewer stops and arrests needed to be made. But social mobility appeared to reach its nadir in late March and significantly increased throughout April, May, and June. Thus, if policing levels dropped to a new low in March and remained low (as the numbers above suggest), then police would have had to gradually increase the number of officers on the streets to maintain constant levels of policing. But these events involve more gradual trends extending over several months. So what caused the sudden spike in homicides beginning in late May? Here the protests seem like the most likely decisive factor—and the mechanism by which the protests caused the homicide increase is de-policing that the protests created in a variety of ways. Superintendent Brown offered a very straightforward explanation of the mechanism for his city, explaining that “[e]very time we have to drain our resources for protests, the people on the West Side and the South Side suffer.” In that sense, what is going on might be described not such much as “de-policing” as “re-deployed policing”—the protests caused the police to be redeployed from their ordinary day-to-day work and were instead put to the task of managing protestors or to being placed in positions to deter the looting of businesses. And while they were working on task associated with the protests, violence spiked in the areas from which they were missing. This pattern is suggested by Chicago arrest data. As shown in Figure 20, arrests declined substantially when the COVID-19 lockdowns began but were returning toward normal levels—when the protest started. After that, through June and July, arrests in Chicago were at roughly half of their historical levels.

To be sure, arrest data by itself is not a perfect measure of policing trends. As Professor Rosenfeld (among others) has observed, other things being equal, police will make more arrests when there are more crimes and fewer arrests when there are fewer crimes.105 Accordingly, using a ratio of arrests to crime is a better indicator of police enforcement. While arrests were declining or remaining stable after the start of the protests, offenses were dramatically increasing, meaning the arrest-to-offense ratio was declining. De-policing could also have been compounded due to other factors. John Catanza, President of the Chicago Fraternal Order of Police, argued that following the protests, the police were moving into a more defensive posture, because they were under attack and not receiving support from the mayor or other leaders. As he put, the criticism of police in the wake of the protests is “going to cause officers to pause and say: ‘I want to go home today safe. I want to make sure I keep my job. And I want to make sure I don’t go to jail.’ It’s not going to be ‘react first’ unless it’s a life-and-death situation.”

Philadelphia

Philadelphia, one of the largest cities in America, is also suffering a terrible homicide and shooting spike. For example, 180 people were shot in May, 201 in June, and 215 in July. The next-highest monthly tally over the previous five years was in August 2015, when 162 people were shot. While homicides and shootings were increasing, Philadelphia police were responding to a series of ongoing protests. On May 30, a few days after the protests started, 13 officers were injured, and the city was forced to institute a curfew. Protests continued. In mid-June, a defund-the-police rally took place near the future headquarters of the Philadelphia Police Department. Tensions rose throughout the month, and, on June 23, demonstrators tried to remove a Christopher Columbus statute—leading to a clash with counter-demonstrators chanting “U.S.A.” In early July, Philadelphia Deputy Police Commissioner Melvin Singleton explained how police operations had been hampered by the on-going protests: “The civil unrest has been constant and ongoing in Philadelphia, and we need heavy deployment to keep opposing groups away from each other. We’ve seen more demonstrations than ever before, and it has lasted from morning into the night. It requires us to consistently pull officers into protest response, and that takes away from the street response.” Commissioner Singleton also noted that the protests had harmed the city’s hot spot policing, observing that protest responses have “contributed to a reduced number of officers on the street in our pinpoint areas, where we are focusing our data-driven, intelligence-led, offender-focused resources. Some of those areas are not as heavily policed as they should be, and that is impacting our violence.”

Milwaukee

As noted above, Milwaukee is one of the cities that has seen a dramatic increase in homicides since late May. And Milwaukee also saw significant anti-police protests in the wake of George Floyd’s death. On May 29, hundreds of protesters blocked city highways. There were reports of looting and what Wikipedia described as “mild arson.” One police officer was, as described by Wikipedia, “mildly injured by gunfire.” As of June 9, protesters had been demonstrating for twelve days. While protests were occurring in Milwaukee, proactive policing was declining—and homicides were spiking. During the first month of protests in Milwaukee, the Police Department saw 26 sworn members either resign or retire—the general sense was that morale in the department was “terrible.”

While I have been unable to locate hard data on Milwaukee proactive policing during June and July, several leaders in the Milwaukee Police Department have discussed the trends. Inspector Leslie Thiele explained that “we’re doing the best we can. Our officers are still out there taking their assignments, and some are still trying to be proactive. But, overall, I think the feeling is that they’ll do what they have to do but proactive policing is minimal right now.” Part of the concern is the fear of officers becoming the next viral incident of alleged police misconduct: “We’ve had younger officers who have been on for a year to 10 years, and have resigned because they’re not willing to put their families through any news stories that would come out if they were involved in anything. It has been difficult, and morale is low.” Similarly, Milwaukee Police Inspector Terrence Gordon noted the role of protests in diverting police: “[T]he fact that police departments in major cities are distracted right now is a contributing factor everywhere. It is definitely an issue here. Our department is distracted with politics, inquiries, demonstrations, everything you can imagine except serving the neighborhood we come to work to serve.” And Gordon also explained that “[w]e haven’t stopped doing our shoot reviews every week, but our detectives were on the demonstration lines with riot helmets and batons for two weeks. Our police officers who would normally be on patrol were also on the line. I really think that if the city got back to doing what we’re elected and appointed to do, we could get a handle on this. I don’t think we need a new strategy. I just think we need time to do our jobs.” I have been unable to immediately locate 911 call data from Milwaukee, which could shed further light on possible causes of the homicide spike. But pending further information, there is at least good anecdotal support for the de-policing hypothesis in Milwaukee during June and July 2020. (Note: On August 23, 2020, an African- American man, Jacob Blake, was shot in Kenosha, Wisconsin, about forty miles from Milwaukee, leading to additional anti-police protests.).

New York City

In New York City, homicides—and, even more dramatically, shootings—have increased in June and July, as noted above. The debate has been over why this is happening has been contentious, and this brief blog does not purport to definitively resolve it. But, at a minimum, New York City’s experience contains substantial support for the de-policing hypothesis. There appears to be no debate that shootings increased substantially in New York beginning shortly after George Floyd’s death in late May and then, after a brief and modest turn toward normal levels, escalated even more dramatically in mid-June (as shown in Figure 13, above). As described by local media, “gun violence exploded across the city,” with three times as many shootings in the last two weeks of June compared to same period in 2019. By the end of June, New York City had suffered 205 shootings, making it was the bloodiest June in 24 years. July was even worse, with 244 shootings—a 177% increase over the previous year. New York City had become the “City of Bullets.” Why the sharp increase in gun violence? The most straightforward answer is that policing declined—and, in particular, proactive policing declined, the very type of policing most needed to fight gun violence.

The decline in arrests would seem to be an obvious explanation for New York City’s shooting spike that began at exactly same time. But NYPD officials have also identified other factors.  Police Commissioner Shea, for example, said “It’s bail reform. It’s COVID. It’s emptying out of the prisons.” Commissioner Shea may well be correct that some of these other things have increased overall crime levels in New York City. There are good reasons for believing that, for example, that bail reform can increase crime. Certainly the possibilities that Commission Shea raises urgently deserve further investigation. But the focus of this article is not to explain general crime rates but rather recent spikes observed following the George Floyd death—i.e., the “structural break” upwards in, for example, homicide rates that the Rosenfeld-Lopez Report identifies. As an explanation for the precipitous spike in shootings beginning in late May and accelerating even more dramatically in mid-June, these other cited reasons do not seem to fit well. Consider the New York bail reform law. It was enacted in 2019, was implemented in many jurisdictions late in 2019, and officially took effect on January 1, 2020—more than five months before the shooting spike. Thus, the increasing pre-trial releases due to the law would have been spread out over months—rather than accelerating in late May and surging in mid-June 2020. And impacts from the bail reform changes were likely already being felt several months earlier.133Similarly, the release of inmates from Rikers due to concerns about COVID-19 also started much earlier than the spike. On March 16, 2020, there were 5,458 inmates in the New York City jail population. By May 7, 2020, it was down to 3,898.

Jail releases beginning three months before the shooting spike might well have significant public safety implications but seems like a poor candidate for triggering the sudden and sharp spike in shootings starting in late May and then surging further beginning in mid-June. Interestingly, according to the New York Times, a confidential analysis of New York City crime data by city officials reached a similar conclusion. The analysis suggested that “the state’s new bail law and the mass release of inmates from city jails in recent months because of the coronavirus outbreak played almost no role in the spike in shootings.” Instead, the analysis identifies the drop in arrests for gun crimes as the explanatory factor—i.e., de-policing in the form of reduced arrests.

Conclusion

Homicides and shootings have spiked across the country, beginning in late May and continuing through June and July 2020. Why this has happened is a vital criminal justice issue with truly life-and-death implications. This article takes advantage of the fact that crime data and other information from public agencies is now readily available in “real time.” The quantitative data and qualitative evidence strongly suggest that a “Minneapolis Effect” has struck—i.e., in the wake of anti-police protests following George Floyd’s death in Minneapolis, police officers are being re-deployed from anti-gun efforts and are retreating from proactive law enforcement tactics. This reduction in law enforcement efforts targeted at firearms crimes has led, perhaps predictably, to an increase in firearms crimes. This article attempts to quantify the size of the Minneapolis Effect, estimating that about 710 more homicides and 2,800 more shootings resulted because of reduced policing in June and July alone. And the victims of these crimes are disproportionately Black and Brown, often living in disadvantaged and low-income neighborhoods.

While these estimates are stated in the cold precision of an economic calculation, it must be remembered that behind these grim numbers lies a tremendous toll in human suffering—lives lost, futures destroyed, and families left grieving. Understanding the nation’s recent—and on-going—homicide spikes requires urgent attention. And even more urgently, the nation needs to consider all possible responses to this tragedy, including responses that involve increased and proactive law enforcement efforts directed at combatting gun violence.

---

Cory Wilson is a criminal defence lawyer based in Calgary. If you have been charged with a criminal offence or are a suspect in a criminal investigation, call today for a free, no obligation consultation.