Skip to main content
ABC News
Murders Spiked In 2020. How Will That Change The Politics of Crime?

Welcome to FiveThirtyEight’s politics chat. The transcript below has been lightly edited.

sarah (Sarah Frostenson, politics editor): New data released by the FBI last week showed a spike in homicides in 2020, while major crimes overall in the U.S. fell. This marked one of the biggest one-year increases on record, and in addition to the sheer numbers involved, it was also notable that it wasn’t a regional or big-cities-only story — it was a story that affected nearly every part of the U.S.

And while it’s very unclear what the spike in murder will mean moving forward — there is some evidence that 2021 has already been less deadly, for instance — it does raise larger questions around how we think and talk about violent crime in the U.S., especially as it pertains to our politics. For instance, crime and policing were major issues in the 2020 presidential election, with some evidence that it may have hurt Democrats. It’s already been a key issue in the 2021 elections, too; although it isn’t clear whether Democrats’ messaging on policing and crime has hurt them.

So let’s talk about the politics of crime. First of all, what do we even mean when we talk about crime?

jacob.kaplan (Jacob Kaplan, crime researcher at Princeton’s School of Public and International Affairs): When people talk about crime they’re usually using the FBI’s list of “major crimes” as a guide (whether they realize that or not). The FBI’s list stems back to 1929 and includes eight crimes that are serious and generally well-reported — murder, robbery, aggravated assault, rape, burglary, theft, motor vehicle theft and arson.

These crimes are then often summed up into a single value of “major crimes,” but the bulk of the crimes in this category is often property crimes, specifically theft.

What the latest crime data can and can’t tell us | FiveThirtyEight Politics Podcast

jennifer.doleac (Jennifer Doleac, economics professor at Texas A&M University): And crime means different things to different people. The FBI tracks serious crime, as Jacob said — and that’s the stuff most people worry about. But lower-level crime (trespassing, drug possession, shoplifting) is far more common, yet not consistently tracked.

lisa.miller (Lisa Miller, political science professor at Rutgers University): As Jennifer is getting at, it’s important to distinguish between different types of crime. Violent crime, especially homicide, is particularly worrisome, even if it hasn’t directly affected a given community. People are understandably fearful when they start hearing that murders are increasing. So I think that if it’s violence that is on the rise, that really is troubling.

ameliatd (Amelia Thomson-DeVeaux, senior writer): Right, the distinction between property crime and violent crime seems really important. Property crime has been on the decline; it’s violent crime that’s spiking. This is not to say that people aren’t concerned about property crime — but I think when we’re talking about crime in a political context, violence is what drives a lot of the fears and knee-jerk reactions.

jacob.kaplan: And it wasn’t just murder that went up in 2020, either. Aggravated assault — an assault that involved a weapon or seriously injured the victim — went up too, with assaults involving a gun driving a lot of the uptick.

maggie.koerth (Maggie Koerth, senior science writer): Oh, that’s a really interesting question, Sarah, because it’s bigger than crime statistics. I think it’s about whether people feel safe in their communities. And that can be based on a lot of things. Some real risk to life and limb and some perceptions of risk. 

I remember being really struck by the work of Wesley Skogan, who worked in Chicago in the 1990s, documenting the things that people who requested police assistance wanted solved. And what he found was that people were often thinking about things like litter and loud music as emblematic of safety problems or their fear for their communities — even when those things aren’t, strictly speaking, crime. 

A close-up of a cop’s loop of pull-tie handcuffs

related: Cities That Reduced Arrests For Minor Offenses Also Saw Fewer Police Shootings Read more. »

ameliatd: The terms can get pretty mushy here, too. If we say that violence is rising, are we talking about murders? Aggravated assault? 

As Jennifer was suggesting, “crime” and “violence” can mean different things to different people that don’t map easily onto the data we have. And of course, politicians don’t make clean distinctions when they’re talking about crime and violence.

lisa.miller: I agree with Amelia that violence is what drives fear, though Maggie’s point is really important, too — sometimes a lot of things get conflated. But in my own research, as well as in others’, rising violence is something that really registers for people.

jacob.kaplan: “Crime” is such a vague term and can range from shoplifting to murder that even academics have a hard time agreeing on what offenses should be included.

ameliatd: That’s completely right, Jacob. And it tends to be defined through policing data, too — which is kind of odd in itself, since arrests are just the beginning of a long criminal justice process that doesn’t necessarily end in a conviction.

jacob.kaplan: Yeah, I think if we had much better court/prosecutor data, we’d have completely different conversations about crime. But there’s almost no good public data on that (though Philadelphia’s district attorney’s office is one exception), so as Amelia says, we’re reliant on data that is first reported to or observed by the police and then reported by the police to the FBI.

One thing we do see in the data, though, is that the 2020 murder spike is also relatively broad — affecting a lot more people than just “gang members,” as some organizations have claimed. 

According to my analysis of FBI data for agencies that reported detailed murder info in both 2019 and 2020, nearly every single group of murder victims experienced an uptick last year — though it’s important to stress that some of these groups have really low incident rates to begin with, which can result in large swings between years.

Number of murder victims, by relationship to the suspect/offender, and the percent change between 2019 and 2020

relation to offender 2019 2020 Percent change
Wife 330 294 -10.9%
Brother 93 86 -7.5
Daughter 89 86 -3.4
Son 157 160 +1.9
Father 95 101 +6.3
Boyfriend 164 180 +9.8
Girlfriend 390 431 +10.5
Mother 118 134 +13.6
Acquaintance 1,494 1,729 +15.7
Stranger 1,549 1,886 +21.8
Neighbor 85 105 +23.5
Other family member 209 260 +24.4
Other (known to the victim) 757 955 +26.2
Friend 272 358 +31.6
Husband 58 79 +36.2
Unknown 6,216 8,504 +36.8

Excludes victim groups where fewer than 50 people were murdered in 2019, and only includes murder and nonnegligent manslaughter. Includes only the first victim and first suspects/offender in an incident, even if there were multiple victims and suspects/offenders.

Sources: FBI, Jacob Kaplan

maggie.koerth: Woah, Jacob. Does that mean that violence against husbands is way up and violence against wives is down?

jacob.kaplan: Yes, but remember: Few husbands are murdered. For instance, 58 husbands were killed in 2019, compared with 79 in 2020. Whereas, for wives, 330 were murdered in 2019, compared with 294 in 2020.

maggie.koerth: Gotcha. That makes a lot more sense. So the big takeaway from that chart is less about the different types of victims affected, and more that murder increases in 2020 aren’t something that was happening to a narrow swath of the population?

jacob.kaplan: Right. Nearly every relationship group had an increased number of murders in 2020, regardless of the relationship type, so one takeaway from the uptick in murders in 2020 is that it’s not isolated to certain types of murders.

ameliatd: Jacob, do we have a sense yet for why the murder spike is so broad?

jacob.kaplan: I don’t think we’ll be able to answer why it’s so broad for a while — especially not until the other, more detailed FBI data comes out from the National Incident-Based Reporting System later this year.

But there is evidence at this point that it’s not just an increase in gun violence in 2020; other use of deadly weapons also increased.

Number of murder victims, by weapon type, and the percent change between 2019 and 2020

weapon type 2019 2020 Percent change
Asphyxiation* 68 52 -23.5%
Shotgun 159 154 -3.1
Blunt object 316 312 -1.3
Other/Unknown 733 762 +4.0
Personal weapons† 501 552 +10.2
Knife/Cutting instrument 1,200 1,445 +20.4
Handgun 5,831 7,273 +24.7
Rifle 296 382 +29.1
Firearm (unspecified) 2,977 4,341 +45.8

*Includes death by gas
†Includes beating as well as using their fists, feet, arms, etc., as a weapon

Excludes victim groups where fewer than 50 people were murdered in 2019, and only includes murder and nonnegligent manslaughter. Includes only the first victim and first suspect/offender in an incident, even if there were multiple victims and suspects/offenders.

Sources: FBI, Jacob Kaplan

Notably, however, the quality of the data does seem to have dropped in 2020. There was just a lot more incomplete data. True, the data quality is generally pretty unsatisfactory, given that reporting is voluntary so not all agencies report in the first place, but I’d guess that there was an effect had by the transition to how the FBI is now collecting this data and the resource constraints caused by COVID-19.

ameliatd: Well, that answers one question I had, Jacob. I had seen a theory floated that people were simply carrying more guns in 2020, and that this explains some of the increase. But if use of other weapons increased too, that’s plainly not the answer!

jacob.kaplan: There is good evidence that people were carrying more guns, and gun murders did go up more than other types of murders. But it’s not the only story.

maggie.koerth: That’s interesting. One thought I had prior to you saying that: Research that has suggested more guns in a household or a community is correlated with an increased risk of homicide for the people who live there. Like, it wasn’t shocking to me to see a massive increase in gun purchasing and a big increase in gun murders happening in tandem.

But clearly that’s not the whole story here.

Why serious changes to policing are unlikely to happen

lisa.miller: Is it possible that the broadness of the spike in murder is a function of decline in trust and legitimacy as a result of the pandemic? Or polarization and general political unrest, more generally? 

ameliatd: Lisa, do you mean that people were less likely to call the police or other authorities when they were in situations that could lead to violence, and that perhaps this explains the spike?

lisa.miller: Amelia, I was thinking more about theories of violence self-help — that is, people are less likely to restrain the urge to act out violently when there is less of a sense of shared legitimacy and social trust. Or the idea that criminologist Gary LaFree and historian Randolph Roth have put forth that suggests societies with higher levels of social trust and political legitimacy experience lower rates of homicide. 

But, of course, when those relations break down — or if they are not deeply established in the first place — we see higher levels of homicide. And we certainly seem to be living in such a time period now, where trust and legitimacy were already pretty low even before the pandemic. I wonder if the pandemic may have exacerbated those problems.

maggie.koerth: You guys, why can’t crime patterns neatly stick to theories of crime and human behavior? (I kid.)

sarah: As you all are getting at, there is no one answer for why there was an increase in violent crime, but what are some of the factors we can point to? We’ve talked about the role guns played in this and how that’s not the whole story. But what role did the pandemic play? Or last summer’s protests for racial justice that followed George Floyd’s murder? Is there a “police legitimacy crisis” as political scientist Justin Nix has argued?

jennifer.doleac: I’ve had the chance to preview a few as-of-yet unpublished studies with conflicting answers about what happens to the reporting of crime to police after high-profile police killings. It seems to matter sometimes but not always — perhaps because communities that were most affected didn’t trust the police much in the first place.

There is some evidence of a “Ferguson effect,” where police appear to pull back in the face of criticism/protests. And, of course, it’s hard to isolate the effect of the pandemic as everyone is stressed out and many people lost jobs and income over the past year and a half, so those issues could be contributing to the increase in violence, too.

There are just a bunch of possible explanations and no good way to nail down whether one is better than others. We may never know, given how much was going on during this period.

jacob.kaplan: And when you consider that we still don’t have a consensus for why violent crime increased in the 1980s and then decreased in the 1990s — and even continued its decrease up until now — it’s going to be really hard to figure out what’s happening now, like Jennifer said.

ameliatd: Yeah, the difficulty of untangling different, simultaneous social forces is something that a bunch of experts raised when I was talking to them about why crime rose and fell a few decades ago. You can’t go back and do a lot of controlled experiments to figure it out, unfortunately!

jacob.kaplan: And even if you could, what worked in the 1990s may not work now.

A grid background with a cut out stamp of McGruff the Crime Dog on it, all filtered blue.

related: Americans Are Worried About Crime, But That Doesn’t Mean They’re Blaming Democrats Read more. »

ameliatd: But the political dimensions are really interesting to me. Criminal justice reform takes a loooooong time. So even when politicians do focus on it, the incentive is to do something that has a quick impact they can point to. Which is not always what’s best or most effective.

jennifer.doleac: It’s a challenging conversation to have now, given some of the politics around police reform, but I think police reform is still very important to improving trust between communities and law enforcement. We have lots of evidence that police are very effective at reducing violent crime, like homicide. One paper from a group of economists and public policy scholars finds that hiring more police officers reduces homicide rates and disproportionately benefits Black communities.

And the nonprofit Criminal Justice Expert Panel (which I co-direct) recently surveyed leading crime researchers about popular options for reducing gun violence, such as gang-takedowns/hot-spot policing, community-led violence interrupters and focused deterrence on community members known to be at-risk. We found that researchers were most bullish on the first option (gang-takedowns/hot-spot policing) and somewhat optimistic about the third (focused deterrence), though with lots of caveats. 

The evidence on focused deterrence (which involves police) and community-led violence interrupters (which do not), however, is pretty weak and mixed. That means, despite both being popular among advocates of police reform, we’re not actually sure those policies are proven. It doesn’t mean they’re not promising, but it’s complicated.

lisa.miller: What I find interesting here is that many of the policies and programs — which are evidence-based — are the types of programs that have historically come under fire from groups critical of the police. “Gang takedowns,” hot-spot policing and focused deterrence sound a lot like the kind of aggressive policing — including stop and frisk — that is likely to run into resistance from groups that are seeking to minimize, not expand, the role of policing in confronting crime. While the public would like to see the police involved in fewer social problems, there is widespread support for the police dealing with serious crime.

maggie.koerth: One of the things I thought was interesting when Amelia and I were reporting on police reforms is that all those things COULD represent aggressive policing — or they might not. It depends on how they’re implemented, and they’re implemented in such different ways that it’s often hard to compare data. And it ends up making these conversations harder, too. Because some people might imagine “hot-spot policing” and think of stop and frisk. And some people might imagine something more like a beat cop with a presence in the community. And it could mean either or both.

ameliatd: And one of the challenges for me in thinking about what’s “evidence-based” is that the experiences of the people who are living in high-crime communities can be hard to incorporate in that kind of research. Maybe a particular reform or approach does reduce crime in the short term — but does it make people living in those places feel safer in the long run? That’s obviously harder to measure, but it’s important to explore.

sarah: We’ve talked about some of the reasons why violent crime might have spiked in 2020, as well as possible ways to curb it going forward, but what are some of the risks if elected officials ignore the problem? Is this an issue that is particularly challenging for Democrats because of where some on the political left stand on police reform?

ameliatd: Well, to state the obvious, it’s bad if politicians are ignoring a trend that is making people’s lives worse or more dangerous. So if people are truly feeling less safe, that’s important for politicians to address. The problem is that it’s hard to figure out how much of people’s feelings are being driven by what politicians are saying about crime.

lisa.miller: I think of this in two ways — the short-term and the long-term. In the long-term, if this is a temporary rise in violence, it’s possible it won’t have a big electoral impact. In the short term, I do think it carries risks for some Democrats if they are primarily responsive to the groups demanding that cities divest from policing or shrink police forces. People want more police when violence is rising. That doesn’t mean they want police to be abusive or violent. But reducing police services during such a time has some political peril.

jennifer.doleac: Another possible risk here is that the momentum behind criminal justice reform is lost. There has been broad support for meaningful reforms across the criminal justice system, for many years now. That’s likely in part because people have felt safe enough to be open to scaling back things like policing and incarceration. But if violent crime starts rising, that support might vanish, even if those reforms were moving us in the right direction.

ameliatd: It’s not a good thing that murders are up! But it’s hard to have these discussions in a nuanced way. Nor is it often in politicians’ best interest to do so.

Electorally, though, is it worse for Democrats? I’m not sure. I think the bigger risk is that, as Jennifer and others are saying, reform policies that were getting floated on the left get scaled back because of fears about the electoral consequences.

maggie.koerth: And whether it’s electorally bad or not, it’s bad for the reforms — because the perception of electoral challenges is there.

Do you all see politics forcing a choice between reform and safety? Or rather, safety from police and safety from criminals? 

lisa.miller: I think some politicians will present the choice that way, but I don’t think that is inevitable.

What I can say — from 25 years of researching the politics of violent crime in local, state, national and cross-national contexts — is that a serious crime wave is likely to result in more policing and more imprisonment. That has happened in every context I’ve examined. But policing and prisons don’t have to be the only responses, and the responses don’t have to be on a mass scale. There are many ways to increase policing, and even confinement, that are not as brutal as the system we have in the U.S.

Are Democrats really that far from passing Biden’s agenda?

I do buy that Democrats should be worried about Virginia gubernatorial race: Silver

Sarah Frostenson is FiveThirtyEight’s former politics editor.

Jacob Kaplan is a crime researcher at Princeton’s School of Public and International Affairs.

Jennifer Doleac is an economics professor at Texas A&M University who studies crime and discrimination.

Lisa Miller is a political science professor at Rutgers University who studies the politics of violent crime, political institutions and inequality in local, national, and cross-national contexts.

Amelia Thomson-DeVeaux is a senior editor and senior reporter for FiveThirtyEight.

Maggie Koerth was a senior reporter for FiveThirtyEight.