The Numbers Regarding Political Violence in the US
Data? How quaint.
As we find ourselves in a moment in which we are, unfortunately, talking about political violence and the commensurate finger-pointing that this generates, it seems useful to me (you know, a weirdo interested in what evidence and data we have on the subject) to look at the numbers.
Recognizing that any attempt to count and classify these kinds of things is imperfect and open to some level of legitimate debate, as to how to code (i.e., what categories and descriptions should be applied to a given event). However, if done properly, it is possible to see what raw data is used and to understand what definitions were applied so that reasonable people can understand what they are looking at.
One of the claims that has been made by the administration and many of their allies in the media and broader society is that the “radical left” is the main perpetrator of political violence in the United States. I would note, to my point in the paragraph above, they do so without any clear definition of “political violence” or of “the radical left.” Indeed, the accusations are incredibly vague.
One study of such violence was produced by the Cato Institute. I think, given our polarized times, that it is worth noting that Cato is a rightward-leaning think tank.* They are more libertarian, certainly, than the current mainstream American right. I would say that my long-term impression of Cato is that they have a point of view, but do good work in the main. They aren’t, for example, The Heritage Foundation (which has a point of view and did hack work as a result, and then degenerated into the ideological architects of our current doom).
The following was posted on their site the day after Kirk was shot: Politically Motivated Violence Is Rare in the United States.
Here’s their basic structure:
A total of 3,599 people have been murdered in politically motivated terrorist attacks in the United States from January 1, 1975, through September 10, 2025. Murders committed in terrorist attacks account for about 0.35 percent of all murders since 1975. Only 81 happened since 2020, accounting for 0.07 percent of all murders during that time, or 7 out of 10,000. Terrorism is the broadest reasonable definition of a politically motivated murder because it is the threatened or actual use of illegal force and violence by a nonstate actor to attain a political, economic, religious, or social goal through coercion, fear, or intimidation. That excludes individual hate crimes, which are frequently difficult to distinguish from terrorism but are often more personal and spontaneous.
This strikes me as a reasonable set of parameters.
Using these parameters, it is not surprising to find that most of those deaths happened on 9/11, with the Oklahoma City Bombing coming in second.
Eighty-three percent of those murdered since 1975 were killed by the 9/11 terrorists (Figure 1). The Oklahoma City Bombing accounts for about another 5 percent. Those murdered since 2020 account for just 2 percent. Terrorists inspired by Islamist ideology are responsible for 87 percent of those murdered in attacks on US soil since 1975 (Table 1). Right-wingers are the second most common motivating ideology, accounting for 391 murders and 11 percent of the total. The definition here of right-wing terrorists includes those motivated by white supremacy, anti-abortion beliefs, involuntary celibacy (incels), and other right-wing ideologies.
Left-wing terrorists murdered 65 people, or about 2 percent of the total. Left-wing terrorists include those motivated by black nationalism, anti-police sentiment, communism, socialism, animal rights, environmentalism, anti-white ideologies, and other left-wing ideologies. Those murders that are politically motivated by unknown or other ideologies are a vanishingly small percentage, which is unsurprising because terrorists typically want attention for their causes.
The point of this is to simply look at the facts, not to point fingers. To put it another way, it is not appropriate nor is it just to say, if a person from category A commits such an act, then anyone vaguely associated with category A is guilty or culpable.
The data also matter because they suggest that political violence in the US is more likely to be right-wing than left-wing. If that is true, it means that federal law enforcement, if it sees its job as protecting the public, would focus more heavily on groups of that ilk. Indeed, the Department of Justice had issued a report that right-wing extremism was on the rise. The Trump administration removed it from the website (see The Guardian: US justice department removes study finding far-right extremists commit ‘far more’ violence). Using a different methodology from the Cato study, it came to the same conclusion: right-wing extremism has been more likely than left-wing extremism in the US in recent years.**
Of course, if you are in government and aren’t really worried about the best way to serve the public, and are instead interested in finding pretexts for growing your power and going after your enemies, then you don’t care about data and reality. Instead, you say things like this.
The White House has ramped up its vow for vengeance in the wake of conservative activist Charlie Kirk’s assassination, with deputy chief of staff Stephen Miller promising to bring the resources of the federal government to bear against what he described as “terrorist networks.”
Vice President JD Vance, meanwhile, argued that those identifying as liberals were largely to blame for political violence and endorsed efforts to shame and make job trouble for those publicly cheering Kirk’s death.
Their comments came Monday during an episode of Kirk’s namesake podcast, which Vance hosted from his ceremonial office at the Eisenhower Executive Office Building.
“It is a vast domestic terror movement,” said Miller, speaking of left-wing political organizations.
“With God as my witness, we are going to use every resource we have at the Department of Justice, Homeland Security and throughout this government to identify, disrupt, dismantle and destroy these networks and make America safe again for the American people,” he added. “It will happen, and we will do it in Charlie’s name.”
(Source: NBC News, ‘We will do it in Charlie’s name’: Stephen Miller vows vengeance for Kirk’s murder).
So yes, I want the murderer of Kirk to be prosecuted. If there ends up being some evidence of co-conspirators/a network of groups involved, then yes, prosecute them as well. However, there is no evidence of any additional participants in this crime at this point, and experience suggests that there won’t be any.
While yes, Trump has made social media declarations about Antifa, that isn’t some actual group, despite constant usage of the term as representing some bogeyman. As best as I understand it, Antifa (which is short for anti-fascist) is a loose movement, rather than some network.
On The Daily this past Wednesday, it was suggested that the groups that Miller might target are Soros’s Open Society Foundations and the Ford Foundation, because they have provided grants to groups that have engaged in protests. Neither of these is a radical or extreme organization, and targeting them would not be for some righteous crusade against violence, but instead would just be targeting groups the administration doesn’t like.
Meanwhile, I have to note that the only example of domestic political violence in recent memory that actually involved a network of actual extremist organizations was the J6 attack on the Capitol in which groups like the Proud Boys, the Oath Keepers, and the Three Percenters were directly involved, to name only three.
But, as I keep having to point out, Trump liked that political violence because it was done in his name for his goals. And he legally forgave them all for it.
See also, the following from two Sociologists writing for The Conversation: Right-wing extremist violence is more frequent and more deadly than left-wing violence − what the data shows.
*From its “About Us” web page: “Our mission is to keep the principles, ideas, and moral case for liberty alive for future generations, while moving public policy in the direction of individual liberty, limited government, free markets, and peace.”
**Given our history of white supremacist violence, as exemplified by lynching and the general behavior of the KKK, our history does suggest that right-wing extremism has been more of a problem.
Is this meant to be serious?
A tally of political violence in America that starts in 1975? Leaving out the 1960s? Largely skipping over the Weather Underground, the Black Liberation Army, the FALN, and AIM?
And limiting it only to murders? So the billions of dollars in property damage associated with BLM in the summer of 2020 doesn’t count as “political violence?”
And even if we take these numbers seriously, they demonstrate that political violence of any sort is an incredibly tiny threat in the United States, with considerably more people getting murdered every year in Chicago due to crime than “Right ideologies” have killed in the last half-century.
I read Nowrasteh fairly often as he is pretty data oriented, as am I. Sometimes I might not like what the data says but by and large I think he makes a good faith effort. In this study he included every source that i can find that tracks terror deaths by political affiliation. There are 18 of them. This allows for lots of crosschecking and not missing anything. Of note, he has been criticized for including the SLPC (Southern law poverty center) among those 18. I am not sure I would have included them since they have obvious bias, however all of the other 17 groups found the same conclusion, to a greater or lesser degree. Right wing influenced killings outnumber the left.
This has been pretty obvious for a while. The abortion fueled killings have been pretty constant and the sovereign citizens kill regularly, especially police (about 30 since 1980). It’s the right wing that forms the militias, which I will admit is mostly about larping and pretending to be tough guys but they do have real guns and occasionally go off the deep end.
To be fair, I think there is probably more property damage associated with the left. The left is more likely to go out and protest and when that happens sometimes the mob gets angry. Sometimes criminal elements take advantage of the protests to commit crime or just destroy stuff and while they are not politically motivated ie actual leftists they do act under the cover of the protests and the left gets the blame.
Steve
I will again post a quote from the WIKI page on Horst Wessel,
Facts will have little bearing on the actions of the administration or right wing media. But they may prompt enough reaction to trigger TACO.
I make it a point to read some RW sources and read comments for some insight on conservative psychology. At Volokh or NRO Corner any suggestion that RW violence is more common than Leftie is met with scorn and derision, obviously fake news. The usual dodge is they don’t count the burning down of cities in the Antifa riots after George Floyd’s heart gave out during a routine arrest.
I think a part of the problem is the unwillingness of the media to face RW criticism and part of it is that a story like, say, a few nuts failing to blow up an electric distribution center, isn’t regarded as big news, page three, one brief story. Dems simply don’t have the desire or the mechanisms to blow the killing of two people and attempt on two others in MN into an outrage the way GOPs are with Kirk. That, and there’s no video from MN.
I’ll add, in passing that Ezra Klein’s at it again. Today at NYT he has a balloon ball interview with the guv of Utah allowing him to say the shooter had leftist views without pushback. There seems to be evidence of pro-trans views, is there evidence of any other liberal belief?
The “separatist” category caught my eye. Going back to the Cato piece, and then back to the paper it’s based on, it appears that while the deaths occurred in the US, the separatist aspect was related to movements in other countries.
It’s clear to me that the violence is going to be things like advocating for the mentally ill and homeless, supporting Palestinian rights, or calling Trump a fascist. The last two years of screaming and investigations into pro-Palestinian kids camping on campus lawns was followed by a back-page yawn over actual organized violence at UCLA. This is the model of what’s going to happen: hysteria over peaceful protest followed by the police watching right-wing vigilantes attack.
@gVOR10:
I’ll just reiterate my theory until there is substantive evidence to the contrary:
Kirk was killed in a act of passion for his derision of Robinson’s romantic partner. It was a planned revenge for Kirk’s homo and trans phobia. While the act (the murder) is deplorable, Robinson may have acted in the only way he could to silence Kirk.
(I make no excuse for the murder, but only offer a possible motivation)
@PepperPrepper:
Damn straight. And apparently mean tweets and yelling derisively at the VP and his family at Disneyland doesn’t count as “political violence” either.
What a sham. Data isn’t data if it doesn’t count the things I want it to count or counts things I don’t want it to count.
If we’re going to include property damage, how about we include damage from wildfires and hurricanes, caused by ignoring the human contribution to climate change? I mean, that has a political cause/angle.
@Scott F.: Also, too, @PepperPrepper: complains that the data only goes back to 1975, a nice round 50 years. Maybe he thinks we need to go back to firing on Sumter (RW violence) or the Boston Tea Party (leftist?).
Their political bias defines what they include in their statistics. I don’t think there’s any way around that, but I find the inclusion of 9/11 deaths to be kind of ridiculous — that’s American foreign policy coming back to bite us on the ass, and is closer to war casualties than political violence.
So, white supremacist hate crimes are not included. Or the killings of trans and queer people.
Hate is political.
And I’m guessing these numbers on political violence don’t include the unidentified masked thugs claiming to be with the government* who are abducting and assaulting brown people on the streets of America right now.
Or the Philadelphia Police dropping a bomb on the MOVE folks.
Or George Floyd.
I’ll give the Cato Institute credit for being pretty transparent with their biases at least.
——
*: without badges, uniforms and names, we only have their word.
I think my one quibble with this data set is that I probably wouldn’t have included Oklahoma City for the same reasons 9/11 is excluded. It’s also a bit of an outlier (and could have been handled in a note). But that’s more about my approach to statistics.
@PepperPrepper:
First, welcome to OTB. Since that was your first post, it got trapped in our filter. You shouldn’t experience that again.
This strikes me as largely “why didn’t you measure it the way I wanted you to measure it.” They set a 50-year timeframe for the study. I think that might have been in part also to avoid having to figure out how to categorize some of the racial violence that was also happening during the 1960’s like the 16th Street Baptist Church bombing or MLK’s assassination. The political realignment of the South creates a lot of challenges with the classification of a lot of the violence that happened there. For example, many KKK members were often Southern Democrats, but that doesn’t mean that the violence they committed was “left-wing.”
Again, this is “why didn’t they do this the way I wanted them to do this.” In part, a lot of this will come down to what’s the most easy to analyze because criminal legal system data in the US is the liquid-shits worst to work with (especially cross-jurisdictionally). I say that based on experience.
I actually agree on this point. Though rather than just concentrating on Chicago, I’ll say that outside of smaller rural counties and municipalities, just about every municipality and county in the US–regardless of whatever party is in control–has higher rates of violent crime and murder than these.
Perhaps we can then also agree that trying to create new laws or crack down on political parties and speech based on political violence rates are a case of moral panics and not good policy.
@PepperPrepper:
The bombing of the Federal Building in Oklahoma, and the bombing of Abortion clinics, gay bars, the Olympic park, and churches. That would be way more then all the leftwing destruction combined.
But even in the 1960’s no Democratic politician would ever condone any action by the Weather Underground, the Yuppies, or groups like the Symbionese Liberation Army. Contrast that where the current Republican party will cozy up to groups like the Proud Boys and the Oath Keepers, and pardon those who do engage in political violence if they are their side.
The left wing violence in the 60’s and 70′ were not condoned by any mainstream Democratic organization the way these groups are embraced by the current Republican party. Hubert Humphry would have never been caught dead on a stage with Jerry Rubin.
And let’s not forget all the lynching’s in the south. Oh, and refresh my memory … when was the last time a right winger got lynched? Ever?
@PepperPrepper: So, your objection is that you don’t like the variable being analyzed (murders) and the timeframe the study used.
Whatever else those objections may be, they aren’t actually an argument about the actual study nor of the data provided.
Side note: if your objection about the timeframe is that you want it to go back farther, as you think that was an era of more left-ward political violence, I would note that you are helping to undercut the Trump administration’s assertions about the now. (And that is accepting your premise, which may or may not have any weight to it).
Are there any similar data sets for other countries around the world? Would appear this is primarily a U.S. problem when looking at volume alone. A “murders in politically motivated terrorism by country” would be a interesting comparable chart.
Asked AI. What countries have the most politically motivated murders per capita? You get Columbia, Mexico, Greece, Guatemala, Ukraine and Myanmar….and of course the U.S…….and AI summary for us is….
The U.S. has witnessed a rise in political violence, with over 520 terrorist plots and attacks in the first half of 2025, resulting in 96 deaths and 329 injuries—a nearly 40% increase from 2024. Incidents have targeted politicians, racial and religious communities, protest groups, and immigration-related entities. Analysts warn that today’s environment, influenced by social media, conspiracy theories, and gun access, may pose even more danger than past periods of unrest.
Gee….shocking
@PepperPrepper: BTW, a study of property damage would be a perfectly legitimate one.
If you do it, don’t forget to include the J6ers damage to the Capitol.
The obsession with Chicago is such a tell about your media consumption.
There are a number of cities that have higher homicide rates in several cities near and dear to me: NOLA, Birmingham, AL, and even my nearby neighbor, Montomgery, AL.
But these are all in red states, so move along, nothing to see there, I guess.
For the record you are engaging in blatant whataboutism.
@Gustopher: For the purposes of the main point I was trying to make here, I would suggest that it is noteworthy that even if they left out a bunch of things you think should fit that they still came away with the conclusion that right-wing extremism of this type far outstrips that of a leftist counterpart (which comports with the FBI/DOJ finding and that of the two sociologists I also linked.
These facts undercut the Trump administration’s current rhetoric.
I think that is important.
@Steven L. Taylor: Dont forget Memphis and maybe Nashville. Anyway, the topic was homicides. One of the reasons we use those a lot is because it’s harder to hide homicides so we think the numbers are more accurate and lots of agencies track them. It’s much easier to figure out who and why people were killed. Property damage is more difficult. If the Proud Boys protest something but under eh cover of the protest a crime gang busts up a jewelry store and sets it on fire was it just a crime gang or were they driven by their right wing ideology? Probably it was just about the money though much of the time we never catch them so we dont know. Anyway, I love looking at data and if you have a data source please link it. I do find it tiresome when people just make emotionally based arguments free from data. My best guess is that there actually is more property damage associated with left wing protests (notice I dont say caused) than on the right.
A number of the 18 agencies that track politically based killings do go back to the 60s and during that era the left was responsible for more killings, but maybe not as many as you think. For example, you cited the Weather Underground. They killed 3 people (recently looked it up). Three of their own died making bombs.
Steve
@steve222:
100% correct. And even in this case, we know the data are incomplete and at the same time they work well for trend analysis.
This. Race-based violence was far more prevalent. I think some might argue that doesn’t count, but especially in the 1960’s so much of that was definitely also political. There were a LOT of race-related political assassinations at the time (the “Mississsippi Burning” incident took place in 1964, Medgar Evers was assassinated in 1962).
Again, in many cases, those might have been carried out by people at the time who were registered as “Democrats,” but those are all examples of right-wing violence. That’s proving that an arrest and conviction were ever made.
Every time JD Vance speaks about Charlie Kirk, in my mind I hear Chris Sarandon saying “There will be great suffering in Guilder if she dies…”
@Matt Bernius:
Strong disagree. There’s almost no significance in the number of total murders relative to the number of political murders. Political murders are inspired by political leaders, movements, or beliefs; are intended to alter society in some way; and can strike at the heart of how we practice free speech, political discourse, and actual governing.
Despite what one would think from reading the US media, we know very little about the motive/s his killer had for shooting Kirk. From what we do know, however, it appears plausible that Robinson killed Kirk because of the latter’s hatred of trans people. If that is correct, it was only “political” because the MAGA movement has chosen to make it so, just as it has chosen to make a host of other matters political ranging from the threat of global warming to a restaurant chain’s choice of logo.
There is scarcely an aspect of American life that MAGA Republicans have not tried to turn into a front in its relentless Cultural Revolution. When only one side of politics is sending troops into American cities, demanding massive tribute from private firms and murdering foreigners in international waters, statistical arguments about who is committed to violence as a political weapon are academic.