A brief Twitter exchange this afternoon got me curious about our recent spike in the murder rate. As you'll recall, the basic mystery is this:
Whatever reason we come up with for the homicide spike, it needs to be something exclusive to murder, not to violent crime in general. What else? I took a look at the FBI's National Incident Based Reporting System to get more details:
NIBRS has increased its coverage over the past few years, which means you have to adjust all the numbers for population, and that makes this whole exercise a little sketchy. That said, it's almost certainly in the right ballpark. What it tells us is this:
- Murder went up more in Black communities than in white communities.
- Both domestic dispute murder and other kinds of murder increased significantly.
- Virtually all of the increase in murder was carried out with guns.
- Every age group was affected about equally.
So what does this tell us about various theories for the rise in murder?
- The policing theory—namely that police have decided to stop patrolling the streets aggressively—strikes me as unlikely. If the spike were due to a police pullback you'd see an increase in violent crime and you wouldn't see an increase in domestic disputes ending in murder.
- The COVID-19 theory is plausible, but I'm not sure it's persuasive. The idea here is that COVID just generally made everyone frustrated and angry, and this led to more homicide. But it's pretty well known that murders driven by anger are much more prevalent among the young, while the murder spike had almost identical effects on all age groups.
- The gun theory looks pretty good. There's been a huge spike in gun ownership over the past couple of years, and sure enough, the murder spike is almost entirely gun driven.
- The George Floyd effect suggests that the events of summer 2020 increased anger in the Black community, and that's where most of the increase in homicide has taken place. There might be something to this, since murder did increase more in Black communities than white communities.
None of this is remotely definitive and I'd say the jury is still out on the cause of the homicide spike. Nonetheless, any good theory has to fit the evidence, so be sure to take all this into consideration if you have a theory of your own.
POSTSCRIPT: For what it's worth, I find the age data especially mystifying. What would cause the murder rate to go up at about the same rate among both teenagers—who are emotional and crime prone—and 40-year-olds—who are neither?
I'd be curious to see the raw numbers, especially by month, and see if there were more murders during COVID outbreak periods (not because of COVID for any particular reason, but quality of emergency health care perhaps resulting in more dead shooting victims).
That's just an outlier thought, but would think it's important to correlate murder increases with shooting increases, as opposed to other types of murders (did these latter change significantly)?
Aside from that, an increase in domestic abuse and violence would not be surprising over the past couple years, and we'd also want to segment out gang-related murders to consider those in their own light.
That's actually a really interesting thought! Because emergency medical services have been getting better and better over time, part of the century-ish trend in declining homicides is that an increasing number of attempted murders fail to actually off the target. Covid-caused stresses on the medical system could have temporarily reversed some of this.
That could help explain, too, why homicides are up, but violent crime not so much--the proportion of murders to attempted murders may have just shifted a bit.
Somewhat related, carjackings:
https://www.civicfed.org/civic-federation/blog/carjacking-incidents-city-chicago
I’ve always said that carjacking is a crime for really stupid/really desperate people. Auto theft isn’t particularly difficult. Why steal cars through force? Really dangerous. If you are caught you will be sentenced to a very long period in prison. Not so with auto theft, particularly with juveniles. You really don’t know who that is in the car. Maybe an off duty cop or some other person who is armed. Auto theft garners no publicity while carjackings are top of the news.
There were a lot of carjackings in the 80’s as I recall it. The crime sort of disappeared for a while. I was surprised that such a low percentage were cleared by arrest. Police Departments are highly motivated to resolve such crimes because they get get good publicity for catching those criminals. Also, stealing a car that becomes a news item makes it that much more difficult to dispose of.
So to me, it’s a crazy crime. Maybe a “fad” crime like the recent smash and grabs.
With keyless entry being almost universal, old-fashioned auto theft is no longer feasible. You need some fairly sophisticated (and expensive) electronic equipment to be able to hack into a car these days. Either that or a flat bed tow truck.
You just steal the fob. The easiest way to steal cars has, for many years, been to steal the keys.
So you mug some pseudo-random person on the street and get their car fob. How long are you going to wander around clicking it looking for flashing lights other than those of a police cruiser? I'd think you'd need to insist the victim take you to their car, which brings us back, at least in effect, towards something rather more like carjacking.
So you mug some pseudo-random person on the street and get their car fob. How long are you going to wander around clicking it looking for flashing lights other than those of a police cruiser?
You tail them after they park.
So, kind of like carjacking, just on foot?
So, kind of like carjacking, just on foot?
Fobjacking. And you may not need to "tail" them at all. Just stick a gun in their face as they exit the vehicle. This technique would be well-suited for stealing the cars of people who get home late.
If you steal the person's wallet too you'll likely have his driver license so you know where he lives. It takes a while to rekey a car so meanwhile the car is just sitting there waiting for the thief.
That would make you think that auto theft would be way down, but that’s not how I understand it.
I suppose auto theft is a bit of a different animal since I worked in Juvenile Hall in LA in the late 70’s. But from what I see online there’s still a lot of auto theft out there.
Also, if someone wants quick cash I notice you don’t have to steal the whole car, just the catalytic converter.
In 1981 there were 1.67 million car thefts while in 2018 there were 749 thousand (with over 200 thousand of those cases where the keys were left in the car).
That sounds like a pretty big reduction.
That was supposed to be: in 1991 there were ...
There's still a fair amount of car theft for the fun of joy riding after which the vehicle is dumped. Happened to both myself and a roommate.
Auto theft is way down from what it used to be. It's been dropping since at least 1990 (the earliest year that I can find statistics) and has plateaued at a much lower level since around 2010.
See https://www.statista.com/statistics/191216/reported-motor-vehicle-theft-rate-in-the-us-since-1990/
I don't have any great insights here, but if I were a betting man, I would bet the entire increase can be explained by increases in gun ownership.
Yeah, that seems like it's worth more consideration. Does anyone have theories about what is causing increases in gun ownership?
Gun ownership is irrelevant. There are enough legal guns by gangs to rid of a small army.
Illegal
Everybody drink
That's not how guns work.
They aren't magical talisman against crime.
One of the things about gun ownership is that not only have Americans been buying more guns, they're buying more handguns.
Handguns are easier to carry around and easier to steal, making them more likely to be used in a murder. This is an even bigger issue now that so many gun owners are new owners who have purchased weapons for "protection." Those buyers are far more likely to have their guns readily available, and not safely locked away in a safe.
Gun culture has also changed, with more people willing to engage in behavior that gun owners of the past would have viewed as reckless.
A day or two ago, Justin posted a link to a story in Florida where sone Republican representative got in to a road rage incident and pulled a gun on some guy. The other guy was also armed and killed him. So, not only are more people armed with handguns, they are also more likely to use them. Places like Florida and Texas keep making laws making concealed handguns legal and use of them to “stand your ground” also legal. It just makes sense that more people are going to be gunned down.
"Gun culture has also changed, with more people willing to engage in behavior that gun owners of the past would have viewed as reckless."
Yep. And that's been the trend even since the NRA stopped being a gun safety group and became a gun manufacturer's shill.
Nope, all covid/lockdown stuff. Mostly black gangs went on turf wars and domestic issues. March is where the spike began and by spring of 2021 was falling. My guess this is back at the trendline in 2022. Lazy analysis.
By now it’s just depressing to hear black politicians and read black pundits complain about everything except this violence among their own. They have all sorts of excuses and very little in the way of influence.
There are plenty of whites killing whites but we don’t really see this being blamed on anyone but the criminals themselves.
Oh well. This is really not my problem. By now the black media and political class have so little credibility that Black Lives Matter is a punch line.
This may be a self selection problem. Few people are deeply interested in hearing people they don't know, talk about problems that don't effect them.
People are really good at finding the things they want to see and missing the things they aren't interested in. This was true before internet cookies and search algorithms....but now its even more prevalent.
I wouldn't say it's the George Floyd effect, I'd say it's the Donald Trump effect, the promotion of insane abusiveness as validating and virtuous.
This obviously has negative outcomes for everyone involved, especially for the helpless victims who end up lashing out wherever they can.
Lol, nope. It's covid lockdowns idiot.
Helpless victim.
What lockdowns?
But the lockdowns are long over with. Why would this effect still be with us?
What would cause the murder rate to go up at about the same rate among both teenagers—who are emotional and crime prone—and 40-year-olds—who are neither?
A sizeable increase in the number of guns.
There's got to be a way to pin it on Fox News! That creates a lot of angry, older guys!
I'm kidding. It's probably just a lot of men of all ages carrying around guns everywhere they go.
I wonder if anyone has solid numbers on actual gun sales. It certainly looks plausible that, say, years of increasingly unhinged right wing efforts to eviscerate firearms regulations had, in recent years, finally begun to result in an increase in the percentage of Americans who own guns (for many years this had been decreasing, probably a function of the decline in rural America's share of the population). Perhaps this effect had begun to be observable by around 2013?
I do think it's possible that the tensions and hysteria related to the pandemic spiked gun sales even higher starting in early 2020, So maybe covid has played some role, too, starting in the first half of 2020.
It is a serious problem - the inchoate rage of the gun-nut types is enough to make this non-gun-owner start reconsidering.
That I don't have the time to put in at a range to become competent with it is one good reason I don't.
This is completely useless, but to me, it seems like the rumble before the boil. I fear this country is headed to a dark place.
In 2020 (the last year for which I can find stats) 32% of Americans owned guns, and 44% lived in a household where someone owned a gun. Neither is abnormally high compared to the historical stats over the last twenty years (2011 is the year with the most gun ownership)
Thanks. I'd guess (based on what you've provided) that gun sales are a more relevant datapoint than percentage of households. It may be the case that a surge in gun sales implies an elevated flow of firearms quickly being resold onto the gray/black market. Indeed, given that this is the USA we're discussing (with its tragicomic lack of anything resembling a coherent national "system" of meaningful controls), I'd be surprised if this weren't the case.
What percentage of murders are/were carried out with guns? If it was already some very significant percentage, it seems unsurprising the greater share of increase would be there.
Doesn't the lead-crime theory hold that younguns these days are supposed to be much less emotional than they used to be?
Doesn't the lead-crime theory hold that younguns these days are supposed to be much less emotional than they used to be?
Sure. And, not surprisingly, they shoot people less frequently than they used to. The fact that gun murders are more common now than in 2015 doesn't mean they're more common than in 1989.
I don't have any profound ideas about why the murder rate is up. Maybe the added stress of life with covid and Trump and Fox and the endless news cycle is making people snap.
There are a few comments I'd like to make.
1. It is entirely possible that most of the increase in gun sales is going to households that already own guns. A lot of the people who own guns are pretty much collectors. Hey, I'm not eating out and I've got stimulus money in my pocket, why not buy that gun I've been hankering for? This suggests that this is not a ridiculous scenario: https://news.gallup.com/poll/264932/percentage-americans-own-guns.aspx It shows that the percentage of households owning guns has been flat for the last 15 or so years.
2. There may be later data, but what I found was 2019 FBI data on weapons used in murders. (Murders is probably a smaller subset of homicides as some homicides are accidental and some are justified.) Anyway, the FBI says that of 13,927 murders in 2019, 10,258 (74%) were done with guns.
It is interesting that rifles of all kind were used for only 364 murders compared to about 600 with hands and feet or 1476 with knives. More people were even killed with clubs and hammers. So our preoccupation with the legality of AR-15s has nothing to do with the murder rate, but it's common use in mass murders. (Which account for a small percentage of murder victims, but are certainly horrible.)
Personally, I don't care if AR-15s are banned, I like rifles with wood stocks. But banning them doesn't really solve any problem. There are plenty of other guns that are as deadly, or more deadly, in the situations where mass murders tend to occur. Why mass murderers tend to go for AR-15s is a question a lot of people should be studying. Maybe we could identify more warning signs.
Of course, with our fractured and dysfunctional healthcare/mental healthcare system we won't be able to do anything even if we know someone has a problem. Medicare For All!!
https://ucr.fbi.gov/crime-in-the-u.s/2019/crime-in-the-u.s.-2019/tables/expanded-homicide-data-table-8.xls
Taking into consideration the age cohort invariability, it has to be a combination of factors that apply fairly equally to all of them. I've definitely noticed a lot of people across the age spectrum very upset about COVID restrictions (masks, getting into places, etc) so it would be really interesting to see a cohort analysis of murders versus anti-vax sentiments (which has a surprisingly large number of both the Dem and GOP sides). Another possible factor is the ever widening divisions / camps that people are falling into which primarily promote anger in order to generate eye-balls and money (not just Fox News!)
Why breakout black/white data? What purpose does it serve?
To look at data and find commonalities?
Why break it out by age, region, gender?
When did video games stop being a reason for all sorts of bad things?
Oh, and don't forget rock and roll.
Short skirts.
OBAMA!!!!!!!!
Guns is way to easy an answer for we simple folk.
I think that there's a possible interaction between guns and policing: the split that would help would be the proportion of victims with/without recent criminal records (as a proxy for the proportion of murders committed by people with criminal records.)
The possible dynamic is this: if you have a felony conviction, you can't legally carry a gun--and a lot of aggressive, problematic policing (Stop and Frisk, for example) is designed to increase the risk of getting caught if you carry a gun illegally. I'm suggesting that if the reduction in aggressive policing made criminals more likely to carry guns, that would make sense of these numbers--and it would also be helpful to know if it's more guns in legal circulation, or more guns being carried illegally, that is the key driver.
The data you use is from incidents reported to police. Homicide is basically always reported, violent crime reporting can change depending on how much trust people have in police. If people don't feel like the police will be effective, don't trust the police, or just want to handle things on their own, then they won't report a crime (even a violent one). This can be consistent with a pull back by police due to the George Floyd protests.
The other problem with non-homicide violent crime is that whether a crime is even counted in your data is dependent on how the police/criminal justice system decides to define it. If it's defined as a simple assault then it doesn't count for the FBI's NIBRS, if it's defined as an aggravated assault it counts and will be reported to the NIBRS.
As evidence for this, see the survey of crime victims which asks people if they were a victim of a crime in the past 6 months. It shows the rate of violent crime (excluding homicide) as about 1600/100,00 (https://bjs.ojp.gov/sites/g/files/xyckuh236/files/media/document/cv20.pdf). The data you use from the FBI shows violent crime ( including homicide) as about 400/100,000. The difference is primarily due to the crime survey, unlike the FBI, including simple assault in its definition of violent crime.
If police are facing a wave of violent crime, and calls from the public to stem the tide, one option is to start downgrading aggravated assaults to simple assaults to relieve this public pressure. Look, they can say, sure murders are up but overall violent crime isn't rising so much; we must be doing something right.
Re: Homicide is basically always reported, violent crime reporting can change depending on how much trust people have in police.
I'm not sure I buy that. For one thing if you present at a hospital with crime-caused injuries, the hospital is required to contact the police. And what about high-dollar thefts? Those are generally reported due to insurance requirements (and with auto theft, issues with the DMV, possible parking and speed camera tickets incurred by the thief, etc.) No, rather than second guessing the stats which strikes me as intellectually dishonest absent very solid proof (not speculation) they might be wrong, I'd suggest we just take them at face value
The post and my response were only about violent crime, not theft.
Not all crime caused injuries require a hospital visit. A punch to the face is assault, however it usually doesn't require a hospital visit.
Using common sense to understand the stats is more intellectually honest than blindly taking them at face value.
Using common sense to understand the stats is more intellectually honest than blindly taking them at face value.
Common sense suggests a sizable surge in murders accompanied by virtually no increase whatsoever in non-homicidal violent crime is not primarily due to crime reporting issues.
The graph clearly shows that since 2000 violent crime and murder don't always track each closely. Look at 2000-2004, and 2014-16 The cause of the difference has to be something that has been true multiple times since the start of the graph
I find it hard to believe that the divergence in violent crime vs. homicide could be explained by reporting issues. Certainly not on a two year timeline.
See here https://www.google.com/amp/s/www.pewresearch.org/fact-tank/2020/11/20/facts-about-crime-in-the-u-s/%3famp=1
The rate of change of crimes from 1993-2019 was different for survey data vs police reported data. Why can't the difference between the rate of change for reported crime and actual crime be variable over time?
I'm puzzled why you don't look more broadly. The difference in trends for murder and violent crime start around 2000 and have been increasing, more in 2014-16 and most dramatically in the last year. What accounts for the difference, especially since better treatment in ERs should mean the violent crime line runs higher than murder, not lower?
"What would cause the murder rate to go up at about the same rate among both teenagers—who are emotional and crime prone—and 40-year-olds—who are neither?" I'll take a crack at that. I'd guess it's because that age group (30 - 50 year olds born 1970 to 1990) failed to ripen, and weren't baked properly. This would be the generation(s) that came of age under Ronald Reagan, Bush I, Clinton, and Bush II. I imagine their cognitive and emotional development were severely damaged by the politics and terrible outcomes of those years, and essentially, as a class, they never grew up.
With regard to crime though as people age they are physically less able to commit violent acts. I'm 54 and no, not a criminal now or ever, but there's no way I could get away with half the physically demanding stuff I could have at 16.
I was also puzzled by the age data, but after looking at it for a while I think I understand it. If that is the increase in each group, then (making up numbers) if you start with 1000 murders by 10-19 year old and 10 murders by 40-49 year old, then a 15% increase in both will be 150 more murders by 10-19 year olds and 1 or 2 more by 40-49 year olds. So an increase, but maybe not the place to look for answers.
I’ve assumed that much of the increase is due to the many neighborhood groups that helped keep violence down being sidelined by Covid. It doesn’t explain why the murder rate started to rise around 2015 or so, but it is part of the puzzle. I have various ideas on why the murder rate is so high, but who knows. I do suspect that more people buying guns is a big part of it.
The NY Times has an op-ed by Spencer Bokat-Lindell called Why Are So Many Americans Killing One Another? He points out various possible causes and several things that could help. One thing that stood out to me was this: “ 77 percent of all killings last year involved a gun — more than in any previous year.”
Why would more guns cause more homicides? The key word being "cause" as in "motivate". I can easily see more guns might result in more murders since guns are generally deadlier than other weapons. It would nice to see a graph that included attempted murders as well to see if assaults that would have previously resulted in an injured, perhaps seriously injured victim now are leaving corpses instead because guns are being used.
Canada, Australia, the UK, and Europe have not experienced similar increases in murder rates (exception: Sweden, where it is attributed to more gang-related gun possession), despite having more severe lockdowns and stricter reactions to covid in general. This has to be a US specific pheomena. What are the differences between the US and other developed countries?