Violence in America 2020-2021

It's not caused by one thing, but there are explanations, and violence should decline in 2022

“The point of writing is to find out what you think.”

- Brett Staples, NYT, December 31, 2021

I was curious to see what the consensus would be about the sharp rise of violence in America once 2021 came to a close. I looked about pretty carefully and was a little surprised to see very few takes. I suppose it shouldn’t come as a shock—those with a policy ax to grind have been flailing away for a year and a half and are probably exhausted. Those who are more curious about honestly untangling this web are somewhat stymied by a great truth about modern social and data science. That is, the available methods for the empirical study of social problems are pretty great at figuring out whether one thing caused one other thing. ‘One’ being the most important word in that sentence.

The truth of the universe is that there are usually many reasons—many causal mechanisms—that explain any single outcome. And, in those rare instances where there is only one causal mechanism explaining an outcome, it is certainly plausible that mechanism could have led to many different kinds of outcomes, and that only chance explains why the thing that actually happened, happened[1]. Looking back at the confluence of a rise of real violence, political upheaval with its own real violence, and a raging pandemic, there is a lot to entangle. I suspect those complications explain at least some of the limited supply of punditry.  

All that said, I do think there is some value to summarizing what we have learned over the last two years about crime and violence in America, as I understand the consensus of thought and evidence. I do so, as always, in the spirit of the Brett Staples quote above. I’m not sure where this essay or any essay of mine will go, because until I write I have only my priors and a loose outline of what I think. So perhaps we can discover some truths together.

What’s the Right Starting Point in Understanding America’s Epidemic of Gun Violence?

The right place to start a complicated journey is to decide where to begin. That turns out to be really critical in understanding the current spike in gun violence. In fact, pretty much any conclusion you draw about the causes of the current violence spike will be determined by your choice of a starting point. So, let me tell you where I start and why, and explain why I didn’t pick the alternatives.   

I think we need to mainly isolate the last two years rather than comparing them to longer trends because the data suggest our recent experiences are discontinuous. The rise in homicide—mainly gun homicide--in 2020 was unprecedented. 2021 has been a little worse. The couplet stands out glaringly in the historical data. As is well documented, there were huge increases in crime and violence in the 1960s and 1970s followed by a decade and a half of relentlessly high levels of violence through both the Carter and Regan administrations. There was a rapid decline in both crime and violence in the 1990s. This was followed by twenty years of relative stability.

And then came 2020, with the largest year-over-year increase in homicide, clocking in at around 30 percent. Looking at year-over-year homicide increases (and homicide decreases) back to 1960, the biggest changes were half as much. The violence spike of 2020 is thus fairly characterized as unprecedented.

2020 was also notable because of the murder of George Floyd. The social unrest and dissent that followed the May 25th homicide led to arguably the broadest popular protest since the 1970s. By comparison, the post-Ferguson riots of 2014 were intense but mainly limited to a few cities. The George Floyd protests were everywhere. And the target of the protests—policing, rather than particular police officers or departments—was unique. The deterioration of police legitimacy was rapid, widespread, and not easily reversed. It has since been swept under the rug, again. But probably only for a while.  

2020 was also notable again for the equally unparalleled rise in civilian gun ownership. Between 2000 and 2009, the average number of guns sold in America per year was 8.17 million. Between 2010 and 2019 the average annual gun sales were 13.52 million. In 2020, there were 22.13 million guns sold. The 2021 number will likely exceed 20 million.  More guns in circulation inevitably mean there are more crime guns. And in 2020 at least, a lot of new guns were being recovered from crime scenes.

So, 2020 and 2021 were just different from the recent past. A clean break. In econometrics jargon, an exogenous discontinuity. Others argue for different starting points in their journey to understand the violence of 2020 and 2021. Yes, crime and violence bottomed out across the United States in 2014, and 2020 and 2021 are to some degree part of the longish upward trend that followed.  Yes, violence remains well below the peak rates of the 80s and 90s in most places, and that difference should be acknowledged and accepted in policymaking.  But while it is important to understand the longer-term trends, it seems tautological to compare the current rates of violence to either the highest point or lowest points is in history. After all, every year looks good compared to the most dangerous year, and every year looks bad compared to the safest year.

But this is a new game, with new rules. Fighting this gun violence epidemic requires acknowledging that these last two years are different.

Gun violence has exploded in America and the crisis is real and it is really bad

I already touched on the essential facts of the matter which are well known, so I won’t spend much time establishing the narrative arc. Gun violence exploded in the spring of 2020. The 30 percent increase in homicide from 2019 to 2020 continued in 2021, which looks on track to end up slightly worse than 2020 overall, although the second half of the year was not as bad as the first half.

Also, and this gets hopscotched right over, but ‘crime’ as a whole continued to follow the now 30+ year downward trajectory in overall criminal offending. Of course, the violence spike was routinely reported in the media as a crime spike, conflating two very different ideas. But while criminologists are quick to say crime and violence are two different things, they often skip over the explanation about why that is important. So let me spend a minute on that key point. And that difference, to put it bluntly, is whether the crime has the potential to kill you or not.

All person crimes—this is a beating, a robbery at gun or knifepoint, a shooting, a forcible sexual encounter—come with a risk of death for the victim. That is, it is appropriate and meaningful to calculate a case fatality rate for each of these ‘person’ crimes. Doing so underscores the terror and trauma that accompany the crime. Those crimes are where our focus should be—our focus as a society and the focus of law enforcement and the criminal justice system.

The rest of ‘crime’ is about our stuff. Our cars, our online identities, our valuables. Having these things taken from you is an outrage with its own trauma, and I don’t mean to minimize any of that. Someone took my bicycle from my yard when I was five and I remember looking at the tree where it had leaned just the day before and knew I would never get it back. A toolmark on a window of my childhood home which explained some missing jewelry and raised some goosebumps, my car helplessly and sadly listing away from the curb toward the road, the brake rotors digging into the pavement where the tires had been. These incidents—and sadly many others come to mind now that I reflect on it—stay with me decades later. And that is real. But they are qualitatively different from being beaten or raped or shot at. An order of magnitude different.

So crime went down, and violence went up and that is really important because violence is worse. Much worse. And the media had trouble understanding the difference. None of that is a surprise.

But what did surprise me this year was how starkly the violence debate exposed some unexpected political faults. The response by the law-and-order proponents was perfectly predictable—crime is up, violence is up, whatever, we need more cops, more prisons, more rollback of even modest reforms.

The progressive response was a lot of silence. And that was a surprise. I guess I do not see what others apparently see—a tension between fighting against violence and fighting for equity. That is, I do not see how an acknowledgment of the violence spike on the one hand and calls for more focus on equity and the tearing down of racist structures, on the other hand, are incompatible. The one is born out of the other.

Philly Inquirer columnist Will Bunch said it better than I can, “People on the left want a world in which human rights are respected and everyone has a shot at happiness — so is there a more progressive cause than curbing gun violence?”  

So, my takeaway is that it is important to acknowledge inconvenient truths even when those truths are actually inconvenient, not just for inconvenient for your enemies. The first step in fighting gun violence is acknowledging the problem.

The SARS COVID-19 epidemic explains a lot of the spike in gun violence in 2020…

Over the last year, I argued multiple times that the pandemic was the main reason for the spike in violence. In “Summer of Rage? Summer of Discontent? Summer at Home” written in September 2020, I argued that the story of violence in the summer of 2020 had nothing to do with partisan politics or even the police.

On spiking homicides and assaults, I think the summer of 2020 is mainly a routine activities story. Young men are stuck at home instead of working or going to school. They are experiencing the same anxiety the rest of us are experiencing, but in neighborhoods with a long history of violence they have the added burden of accumulated traumas, including unresolved disputes. And critically, the people they have serious disputes with are also stuck at home and close by. Imagine someone you loved was killed or shot or beaten and you know (or you think you know) who did it (that is the motivation). You are stuck at home without access to supports, a caring teacher, work for a sense of purpose, or professional supports (that is the lack of guardianship). And the guy who did it is just a few streets away (that’s the target). It’s just a toxic situation.

…And also Explains the Persistence of Gun Violence in 2021

In an essay predicting that the summer of 2021 was likely to be just as bad as 2020, I argued that higher violence was to be expected because:

1. Routine activities will continue to be disrupted in summer 2021. The kinds of institutions that support young people in the most disadvantaged communities will be the last to recover;

2. Hurt people hurt people. The decisive risk factor causing violence in a community is trauma in the community. Loss of family and friends to violence erases hope and opportunity. Time heals, but only in times of peace. And with much more violence and death in 2020, there is no peace on the horizon to end the cycle of retaliation.

3. Illegitimacy. Policing illegitimacy is now a bigger structural problem than in 2020.

4. Disinvestment and intensified disadvantage- the potentially devastating consequences of violence spreading to places causing them to become economically unviable.

5. Guns. Guns turn fist fights into gunfights. It’s inevitable.

I stand by this list, with two caveats. First, thankfully, my assumption that the spike in violence was about both an increase in shootings in already violent places, and, that violence would spread to nearby communities, appears to be only half right. A 2021 paper from five excellent criminologists finds that the ‘contagion’ of violence seems to have been limited. They conclude, “non-contagious (Poisson) events comprised the majority of shootings across time (including 2020). We also find that the spatial location and concentration of shooting hot spots remained stable across all years.” So, the most dangerous places became even more dangerous in 2020 (and presumably 2021) but that violence did not spread extensively.

I also now believe I initially overstated the police legitimacy problem as a cause for violence. In September of 2021, I modified this claim to what I think is a more accurate depiction:

In places with the highest violence and not coincidentally historic mistrust of police, police illegitimacy was already high. In these neighborhoods, the George Floyd murder and the response to the subsequent protests were more of a confirmation of an existing set of beliefs. In places less exposed to violence, the murder was more revelatory and had broader impacts on the public’s sentiment. That would suggest that the role of the police legitimacy crisis was relatively minor.

Drugs and Crime

If I were writing this essay now, I would add another element to the list, which is the effect of increasing gun availability and the COVID epidemic on drug markets and the resulting increase in homicide. Two papers have affected my thinking.

This really interesting paper from Brazil describes the drug selling market in what was a novel way, to me anyway. The central thesis is that there are two types of drug sellers—entrepreneurs who work the party and bar scene, and the second type of drug sellers, who are based in a network of drug selling points. If this model is right, the implications are obvious--COVID severely disrupted drug selling by the entrepreneurs but not those at drug selling points, or what we would call open-air drug markets. In a functioning market, the customers of the entrepreneurs would switch to open-air drug markets, increasing demand. Demand at the open-air markets would further be increased by the toxic stress of the pandemic. And of course, there was an ongoing increase in demand as the opioid crisis finishes the shift from prescription opioids to Fentanyl. A paper by Temple University researchers finds that overdose deaths increasingly cluster around open-air drug markets, supporting this hypothesis. Open-air drug markets are violence incubators, especially when there is no monopoly for sellers and instead, there is competition and conflict. COVID surely worsened these conflicts.

There is a lot more to be said about this, and important papers are coming soon, but the gist may be that we have underestimated the effect of COVID on violence around drug markets.  

The path forward is brighter than the road we have traveled…

So, here is the discovery part of this essay—what does all of this mean for 2022? I think it bodes moderately well. If my list of explanations is right, there is reason to be optimistic about three of the five mechanisms for the violence increase. My Tarot cards say Omicron could well be the last wave before the pandemic becomes endemic and we finally figure out how to live with this ongoing threat. That suggests people’s routine activities should return to more usual patterns—patterns established when there was less gun violence. Social supports should be more readily available and the Biden relief bills have provided a wave of capital for state and local governments to invest in the hardest-hit places—fingers-crossed that they actually use their money in this way. And even the drug market violence should start to wane, as the competitive pressures lessen for similar reasons.

I don’t believe anything will happen with respect to policing. The forces for complacency are strong. One glimmer though is the real problem of recruiting new officers faced by law enforcement. Simply put, the supply of labor is limited where it means working under the old rules of law and order. Young people do not want to work for your Dad’s police department. My guess is that more than any top-down reforms, the basic needs of police to hire and the supply of people who will work for the police only if they think differently will start to subtly shift the policing narrative.

The great unknown is the effect of the tremendous surge in civilian gun ownership. There are roughly twenty million more news guns on the street today than there would have been had there been no pandemic. Some of these inevitably will become crime guns. How many is hard to say. The one new idea I have about guns is that the wave of gun-buying may make gun safety reform harder than ever. If there really are 15 million or so new gun owners, and by buying a gun these people become opposed to gun law reforms, that may be just enough to tip the scales against reform. More on that another time.

Coda

You will note that, as usual, I attribute almost none of the ups and downs of violence in America to anything the police and the criminal justice system do or do not do. That is not by accident. I don’t think it matters very much. I think it matters very much with respect to civil liberties and human rights, but not so much with respect to how many people are hurt by gun violence. That is quite an indictment of our justice system, and we should think hard about how to fix it.

---

[1] There turns out to be a huge literature on ‘stochastic mechanisms’ particularly in the natural sciences where, for instance, your genes burst out a random number of proteins with probabilistic rather than deterministic consequences. See this cool dissertation: https://drum.lib.umd.edu/handle/1903/15748. It also comes up in evolution. I suspect that if I had known about stochastic mechanisms in college, my friends and I would have consumed a great number of beers while considering its implications.